<%BANNER%>

Thermally Stable Ohmic and Schottky Contacts to GaN

Permanent Link: http://ufdc.ufl.edu/UFE0021980/00001

Material Information

Title: Thermally Stable Ohmic and Schottky Contacts to GaN
Physical Description: 1 online resource (144 p.)
Language: english
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2008

Subjects

Subjects / Keywords: contacts, gan, high, led, ohmic, schottky, semiconductor, stable
Materials Science and Engineering -- Dissertations, Academic -- UF
Genre: Materials Science and Engineering thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: Thermally stable Ohmic and Schottky contacts to GaN by Lars Voss, Materials Science and Engineering, University of Florida, Gainesville, FL. A study of Ohmic and Schottky contacts to Gallium Nitride was conducted in order to improve the thermal stability of said contacts for use in microelectronic and optical devices. Contacts were based on conventional schemes with an added layer in order to minimize intermixing between layers and improve the integrity of the contacts upon high temperature annealing and aging. Contact materials examined were of three types: nitride, boride, and Ir. The most successful schemes were based on borides and provided superior characteristics when subjected to both high temperature anneals as well as long term thermal aging. In addition, light emitting diodes were fabricated with conventional contact schemes and selected high temperature schemes. It was demonstrated that the high temperature schemes provided superior long term thermal aging characteristics compared to conventional schemes.
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: Pearton, Stephen J.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2008
System ID: UFE0021980:00001

Permanent Link: http://ufdc.ufl.edu/UFE0021980/00001

Material Information

Title: Thermally Stable Ohmic and Schottky Contacts to GaN
Physical Description: 1 online resource (144 p.)
Language: english
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2008

Subjects

Subjects / Keywords: contacts, gan, high, led, ohmic, schottky, semiconductor, stable
Materials Science and Engineering -- Dissertations, Academic -- UF
Genre: Materials Science and Engineering thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: Thermally stable Ohmic and Schottky contacts to GaN by Lars Voss, Materials Science and Engineering, University of Florida, Gainesville, FL. A study of Ohmic and Schottky contacts to Gallium Nitride was conducted in order to improve the thermal stability of said contacts for use in microelectronic and optical devices. Contacts were based on conventional schemes with an added layer in order to minimize intermixing between layers and improve the integrity of the contacts upon high temperature annealing and aging. Contact materials examined were of three types: nitride, boride, and Ir. The most successful schemes were based on borides and provided superior characteristics when subjected to both high temperature anneals as well as long term thermal aging. In addition, light emitting diodes were fabricated with conventional contact schemes and selected high temperature schemes. It was demonstrated that the high temperature schemes provided superior long term thermal aging characteristics compared to conventional schemes.
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: Pearton, Stephen J.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2008
System ID: UFE0021980:00001


This item has the following downloads:


Full Text
xml version 1.0 encoding UTF-8
REPORT xmlns http:www.fcla.edudlsmddaitss xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.fcla.edudlsmddaitssdaitssReport.xsd
INGEST IEID E20101111_AAAACA INGEST_TIME 2010-11-11T19:28:52Z PACKAGE UFE0021980_00001
AGREEMENT_INFO ACCOUNT UF PROJECT UFDC
FILES
FILE SIZE 25271604 DFID F20101111_AABLUE ORIGIN DEPOSITOR PATH voss_l_Page_068.tif GLOBAL false PRESERVATION BIT MESSAGE_DIGEST ALGORITHM MD5
444bf719a1fddd40f0f1883b43853b61
SHA-1
b17c5bc1eded8f3e8edb3b77ba1fd59af06930a8
8423998 F20101111_AABLTP voss_l_Page_038.tif
ad653b7211aac953e0754ec3331501a9
7e8e82647a4fc2abefbf5a7f5dfd4a5bd81ed14a
F20101111_AABLUF voss_l_Page_069.tif
dec2899ff6381feec8b2ac4c5847fbff
76b54b11c068c3a2ed71e6adcac67ccc74e86545
F20101111_AABLTQ voss_l_Page_040.tif
538970eb1d9148f32d213b05a79dde51
2dfc49d91ad52c24e91a46c125c7cf5e1e269e62
F20101111_AABLUG voss_l_Page_070.tif
d5ad8afb1e8fce3fc61e8fc68d1458c7
3b3a0d69ce66da503488f81b5808d93fc9c6f439
F20101111_AABLTR voss_l_Page_043.tif
17a704d82ff1bbce18544fa3534b2c0c
88e93bccbaba05655026b1fd43cba906cc8372c1
36937 F20101111_AABMAA voss_l_Page_136.pro
c01b5fbd5ca2a53bc8b59a512cf0f694
e244ca5082d9214db2138ed79f1a32fe81bda97e
F20101111_AABLUH voss_l_Page_071.tif
19a6ca4b61b1f1f11768f5dd6852df8f
791e0bcebd4f23d3b20684dee178e46613fb69af
F20101111_AABLTS voss_l_Page_046.tif
1c0e4aa724be14dedf33640703c62cfa
1c80e0374167a57e91d23ebcb6f6b84f3fdd64c4
52650 F20101111_AABMAB voss_l_Page_137.pro
0b0756720f141847f1e8a0d3bf8ce255
8df8732ccf9cff1ff996467b3540598e23b9cbae
F20101111_AABLUI voss_l_Page_072.tif
f4ce0577dd418857fca0780ea9372ed6
a3a5313c244cb0897fcbfaa3e02e48f662d89c0c
F20101111_AABLTT voss_l_Page_049.tif
a70a7698455456fa828875f39587c548
08544446b4dc44207c26cd3de9915a43340dfda6
49144 F20101111_AABMAC voss_l_Page_139.pro
ee27b42cb61724f6c0a809e5dfa10e2b
74414d083f724c57cf9d202a5300cbadd5cec837
F20101111_AABLUJ voss_l_Page_074.tif
96cf4577759f202f59f235f25b7dc81a
213e2e47425cf2ca9293f5784e4b7cdc6cc1c6b3
F20101111_AABLTU voss_l_Page_050.tif
efdb42a668f788e01a8e02c4bbae3723
910b735130e8aa13cc7275389cfff9ff7c537e3f
52127 F20101111_AABMAD voss_l_Page_140.pro
1b5002215749e1090a667b230e2d9a13
40170f0182343e7cf20b68d07a4fc31ab04f6771
F20101111_AABLUK voss_l_Page_075.tif
f670081b26a2d69791349177a02f31e4
410c89666f61680c0301502215a3e51af30ed1f0
F20101111_AABLTV voss_l_Page_051.tif
c478218daa6d7cdeadfa2e047f48d724
06985b26b8af792f226faa6894628554929e4b52
53321 F20101111_AABMAE voss_l_Page_142.pro
ab99625c04d44fcf7c3ff6e23e89c029
190d9edb784934fb561a18f3e71cdc05422c39d5
1053954 F20101111_AABLUL voss_l_Page_076.tif
b01bacd468bc30648c524baa4c97b772
82c013709ca0c6d71aef9939c4128f6ebb62933b
F20101111_AABLTW voss_l_Page_052.tif
e36f0b367bc0790096141a50f1e96f2d
f18a71363e336f569d4b066badbc0a02d1e8c6bf
27094 F20101111_AABMAF voss_l_Page_143.pro
7376dacd39cee4a142435e735a881486
30f92639609809d7d805d0f1246e2f136825c74c
F20101111_AABLVA voss_l_Page_094.tif
7398c971fb78a07d3ece4e718c2b044c
0bc1a3d3816ead1c89ec40f40a89978968f1aa09
F20101111_AABLUM voss_l_Page_077.tif
1d726ce5a325b4ea21159c0402179e20
366e275306f25097ec513206a2cc8f16551395ed
F20101111_AABLTX voss_l_Page_056.tif
2e839d0efe87880be7d689701df839d1
cd362798fe94759df15fa42dbb5eaa418fee071d
24278 F20101111_AABMAG voss_l_Page_144.pro
fb043a07e9ed095de3e5ae303abf72a5
2f6585d7e6a2b037fcd3e684974f3559ded3f94f
F20101111_AABLVB voss_l_Page_096.tif
410e035e57b897387eec3b1ab66c80f5
77abbf1ce0b8a9ac37b76496c953813991f9cdbb
F20101111_AABLUN voss_l_Page_078.tif
b005c6cf16dd045179cbf7aef5969b71
14082e33744d7a06a3bf2a57c23ff7462b5bedc9
F20101111_AABLTY voss_l_Page_059.tif
f6345f0881c1f95a6247d454e1f3b83b
9becbdffcd2c9fd2adfd855c82a88bf8dccbe2f1
91 F20101111_AABMAH voss_l_Page_002.txt
670b974f4e6a7c4f862b5b566a6726c3
b60419110b2ceea5b1e15100b581c56fb633f52b
F20101111_AABLVC voss_l_Page_097.tif
fe352649b8c5512eed31dfa713eb175d
6c99a54dcbe17686c653564e7813c2ee23cd5868
F20101111_AABLTZ voss_l_Page_061.tif
e00243d55f9d49be32472086b1935197
5dc05b02a77ed9dbca7fbcf180041b56c7608e7e
88 F20101111_AABMAI voss_l_Page_003.txt
8d64415d2ffd6a338b1096341928c0c2
b74ac3975f4fef32b59bab956be47e1069619f49
F20101111_AABLVD voss_l_Page_098.tif
dd54240ab629123cb03a134b587aa437
f70767702facac3cc912f1c34f4c203daae70941
F20101111_AABLUO voss_l_Page_079.tif
76a6042c34628955b0ddcc28cbd994f4
654e049f4e6048e4b7815c6d6417d423293df24f
1594 F20101111_AABMAJ voss_l_Page_004.txt
ef18c0825e3dfd8330c76cf4ef30eb09
433c47bbbca1173081049080ff3ab288ad8cee06
F20101111_AABLVE voss_l_Page_099.tif
2f68cf79448ad7d1026ebb2f0abacd86
f41bd143b0266455086134755c9f3c6333d24c7c
F20101111_AABLUP voss_l_Page_080.tif
0efeeed84b2ad1881b93561b0983960b
256aac6dce0bfee94b15b2c50db994ec90e8b6bd
4125 F20101111_AABMAK voss_l_Page_005.txt
87a9a4686aa6b03890e01d42a540e681
5101c74efa462cf260594ad06de77aafd7c21860
F20101111_AABLVF voss_l_Page_100.tif
1e07d99fa2020340ade41fd9f6540480
5dab869ebe06a1b8ed7ec5a76e66bbae12bb6b98
F20101111_AABLUQ voss_l_Page_081.tif
3c5e3d134d48a68c87278d973430316c
139a9a008235fee2811d3e79cf5f73be62344baf
3337 F20101111_AABMAL voss_l_Page_006.txt
39328cd5b953c00e227082d2bde3d78f
31ddb402c9be3d94658c3c92c4d75931d7c3944f
F20101111_AABLVG voss_l_Page_102.tif
1960e79b189c3f6836c4fe329315af94
4d5ef7248667d6efb52341baa310621c12513488
F20101111_AABLUR voss_l_Page_082.tif
359e1a265633a9e19ccfa7667598b1da
72843d583e25ed3e357cbd70c03012762b2d1022
1408 F20101111_AABMBA voss_l_Page_026.txt
060a742c2e20fab3d25a09d5af41a5e9
31372530eaf9f48b78ca655c20d2a50b0c48b17c
1664 F20101111_AABMAM voss_l_Page_007.txt
38bca64fb6f7ce4704d9ef8c8e75c625
94dfa13368189bb89534b033750034afe89afee7
F20101111_AABLVH voss_l_Page_103.tif
019eed248035cae490410d3aa6e91a72
0f63af2b41ceb7f5b96bb034cef0eb57fa089313
F20101111_AABLUS voss_l_Page_083.tif
fbb83d2ca51741752d1a32d076cb7802
1d6d4265fda1959abb5ae0b59cf07dfabeea5404
1864 F20101111_AABMBB voss_l_Page_027.txt
51fb3e54403bc96c3137da0865ad6246
a2c1d1ba7855cfb5b14ddbc2fd24a719cbafa776
3143 F20101111_AABMAN voss_l_Page_008.txt
acfb1d0743177090ceecdd4de1c14664
54569b3806dd1afc9d9f557ea058782bcac25101
F20101111_AABLVI voss_l_Page_104.tif
ac4e1634bfd85ad305dce2f3022a6660
7bf552ae11f77752a39f096797ebcf3549422eab
F20101111_AABLUT voss_l_Page_084.tif
9755bec640d252de9af12999a6a4f7c2
a2d89a0b5e4c2235686ab8e7655963d7f0d07eaa
2040 F20101111_AABMBC voss_l_Page_028.txt
275126df19501fdf8e5d4c90d1ac0f9b
ded89dd9f720054c5177cc3ac89679a90e994d7c
3227 F20101111_AABMAO voss_l_Page_009.txt
c5583c4109b6007d975b50438febddb8
e184e6cffe32dba609fe6c4ccae737d819d8b2dc
F20101111_AABLVJ voss_l_Page_106.tif
3e074f8ebebf554f0733840fa0625907
d64ab45e1e0f878df47bc4c34be080a5cb148c5f
F20101111_AABLUU voss_l_Page_086.tif
36f0efa23b70b90ac61f7e9f429b85c2
5e5ad84ebf39cf593af0c07c22cc3fa24d0ae63e
1673 F20101111_AABMBD voss_l_Page_030.txt
b347a8155acb9d60500d5e4dcd5dc397
7c0bbc44ab35b10bee671dfb253224d5d817f2e1
3322 F20101111_AABMAP voss_l_Page_010.txt
4c48946ac16130e4225d8302698c31d8
089f2529622a03756d5dc830fdd181c70da21264
F20101111_AABLVK voss_l_Page_107.tif
fef1ee59d7099654d80afd1ac8796959
5b27a216ff45482b420b8d9e84265af0c804f278
F20101111_AABLUV voss_l_Page_088.tif
f5db73ad789f1d8e2e6270d9c42ba28d
13b4a961c077bbaab46be9a324ce47f0bac00cbd
1734 F20101111_AABMBE voss_l_Page_031.txt
b328a74b5dbcee3e45933210b1e92d87
e70c738d1bc7528d7782f69bcac9990910267beb
1869 F20101111_AABMAQ voss_l_Page_011.txt
fe9162efdbc9dbb773913178695ce3b9
0398e45e487491d372e0fb3fbfeb6de38d365348
F20101111_AABLVL voss_l_Page_108.tif
85684531e5927b72dde6df6cc586b5eb
c7d5d2bb6e4b9d6b846e571356aee22f8080a73b
F20101111_AABLUW voss_l_Page_089.tif
c7be990facf8e5152c3c807cb7e28b10
9566aaa453ac6437c255795c88f3f8fc6d312e95
688 F20101111_AABMBF voss_l_Page_032.txt
b2f7a778ba43936ed339d498c95bace4
978599002df6deb4b8b97330b47dfdbca17704b7
1362 F20101111_AABMAR voss_l_Page_012.txt
c054b0ec21d7d0f3b03d7340c6b33ebc
dfd0b3b4a61642440ee53391470e5d4db26e7715
F20101111_AABLVM voss_l_Page_109.tif
bd9e32f0b7dbc8d2b455db9f1ecf2a4e
0c716b6f95c3c9e86554ba14fc6c248504ef8cd9
F20101111_AABLUX voss_l_Page_090.tif
18736b1a336fc484c35e4c807d0a09ea
f403b67a2fd383ca7868d66b7440bf0e6df04021
670 F20101111_AABMBG voss_l_Page_034.txt
867a308e22bde1dfa09711d8cdd390ee
0a70df1d9f106a944082498804bf40cdc5b26c08
F20101111_AABLWA voss_l_Page_127.tif
b69f5573b250b6c1d0fc5968f19a4231
f109d41cf70267d2807eadfda16f18e966a51f54
2211 F20101111_AABMAS voss_l_Page_013.txt
31b6068d8a61e703b1a5b78809d58aa1
a35639a2370878003a2b6a9b561e2e9266c0e3f8
F20101111_AABLVN voss_l_Page_110.tif
6a47263cb24fd0a38dda78a169e32da7
72f6a3d95314160aa316b32b6f78b69c01325772
F20101111_AABLUY voss_l_Page_092.tif
5e1f5cf8990c2c1e49c84e2abe3a612a
94a4826a02ab6000918a5ba36ffffec9684e89a1
586 F20101111_AABMBH voss_l_Page_036.txt
6e1881f66b2f3b165b159c8d82dd08a2
2588718de826993e3fe6a5b086fb9c7d1c98cea3
F20101111_AABLWB voss_l_Page_128.tif
8abdee9731b021c1487af4b5575f3774
9f7d485289e1d15a43e1234333f8c51238bf7163
386 F20101111_AABMAT voss_l_Page_016.txt
4fda82833d26b102a2968e74574a4e6b
1a46e362a32a9d430343faa212ac60aece77f6ea
F20101111_AABLVO voss_l_Page_111.tif
5e36c0ff35a6ee8dccaa0806f400775d
90b146f296e2fb832232ba2e0cba615dccc819dc
F20101111_AABLUZ voss_l_Page_093.tif
7f8d5503e88580ced4ddcbc2bfd611b4
7bf451653290f5c205da785b6be00f1ad62f21f2
639 F20101111_AABMBI voss_l_Page_037.txt
569e5fabaaf14878922022c8e616b959
0e5ed69b62631fa2469d3f2582d23366bd1de6b3
F20101111_AABLWC voss_l_Page_130.tif
21a46c436d320e39638f82bf1ba2c860
37e5abaff0812ba6096271e1ef8ae55288416693
308 F20101111_AABMAU voss_l_Page_017.txt
aa77b9a2db9de9187ecf6af78c3aa2db
2a24266c40d6e4d89b574c6aecb56cedfac9f30e
106 F20101111_AABMBJ voss_l_Page_040.txt
d2fabd12a5dd98be399fcd99e2199b83
d1cbdb56676d4d2038ff64e37a36281a361dbd52
F20101111_AABLWD voss_l_Page_131.tif
cd8f62b399e8660baf8122f68e58e46d
5b68544cf79b2f1f71f3b6250c7ff1ca154d7f4e
2036 F20101111_AABMAV voss_l_Page_019.txt
bf53f33bffe3e8c10870c3a5f4c2a9e6
2833bb7caed26857cb7ed89b4a4c46c2232409a9
F20101111_AABLVP voss_l_Page_112.tif
73fee04c3400bcc8f75e6c92e0cbc406
f718e8cbf4ed5734a0e465de13ccf4378cc0cd71
112 F20101111_AABMBK voss_l_Page_041.txt
77a5db0e22828f14283bb6fadc83f11a
99a7d559f3d78c011f90d49f49e7a6524ee48619
F20101111_AABLWE voss_l_Page_134.tif
074073f30a297a97875100a0c76fbb0b
8f4f1854a3eb0b2597ccf32ba6e43074fb99e1cb
2183 F20101111_AABMAW voss_l_Page_020.txt
44d38d24a636640749d102e0b0347e7e
81db5cab53d75ca9b51444d920cf03e547d15ed6
F20101111_AABLVQ voss_l_Page_113.tif
4bfa5544af3851f6485db9f24a823d41
2d58520da735bdfbb10ab18b669ed8cb835619d3
389 F20101111_AABMBL voss_l_Page_042.txt
d8690e9887dcec57fd6fb8e038c6a4a5
3ea0b48ff4f2b5d656020a8097cde50ed1288b2a
F20101111_AABLWF voss_l_Page_135.tif
22b60fe29f0a7ee9d1c09abb7524aa27
1b9b87651c4f5e0a5adc95f47bd37b6c3f3e7d9a
2302 F20101111_AABMAX voss_l_Page_021.txt
fcd340d8450716722decfeed9189b5ef
3b6606391c952f005602bccafda44783796bc65f
F20101111_AABLVR voss_l_Page_114.tif
6d3b5ad621efed0a76760b2e0ed09f18
9c936f4943a8411a49aa4af66c6cebb152b91e15
591 F20101111_AABMCA voss_l_Page_061.txt
a1c6a82821c3e479f38486be29c7d801
ead1b1ac4b07dcfddf013d557f729e91ce3391f0
122 F20101111_AABMBM voss_l_Page_044.txt
25d9a264ab8af7337a7de69ca0689053
eb24f1a9d100e940b9eb5243370a0eb0e925e93d
F20101111_AABLWG voss_l_Page_138.tif
4c0f324a7e7d1c3d3300c5c6c5402eb4
b9e4979e1d03b0e171d06b22fcec2b54c994cc1f
1901 F20101111_AABMAY voss_l_Page_023.txt
ad1c0c088a8dee80b4b2cb6ba79f1181
35c7feb03f0c50a93ed0546cf275883b2f89cbee
629 F20101111_AABMCB voss_l_Page_063.txt
a8e7aba562cacdb64cb6656ed6c9edfa
ed7e94d3fb532b31031ca4c14a6bd24e5d726302
131 F20101111_AABMBN voss_l_Page_045.txt
35dba1f3a0f13c3e4fd33d840306cea5
530b3a42084d50d0a58a74ae7fa8da5c61034935
F20101111_AABLWH voss_l_Page_143.tif
68eb628452d59ee83a3c0987ee7f644b
cbcc876118ce704762b1bfe50323c10f9c7a6525
F20101111_AABLVS voss_l_Page_115.tif
b0fbeef13f739ff073db846f91bb8af6
2c761fbc1ed3180008d562bfcee72178830ff359
869 F20101111_AABMCC voss_l_Page_064.txt
933f867b5f70353bc9e27ffae9f52ce5
48a84530c9e86920fb2a146aafc7b1b5960f46ed
2073 F20101111_AABMBO voss_l_Page_046.txt
6e552865cd4a1708b4101f16764276f8
3a4a52f2865b4869c2421b586afab9dbf5dbf4ab
F20101111_AABLWI voss_l_Page_144.tif
0e1f9bbbf557eca14e2164fea0ba4cc3
e205e2436b884038498ea6e7869692d300e8bf17
2016 F20101111_AABMAZ voss_l_Page_025.txt
b8e0ac1ca9e64a4a3a910b82b4921d4b
7276f3f81b24b567e831e79cdc841c50e6573169
F20101111_AABLVT voss_l_Page_116.tif
5ef3a103b3fb05a258d9638b3e36ccb8
a03e94a78190114f6839626bc30ca72b3f8986d2
1642 F20101111_AABMCD voss_l_Page_065.txt
6c2920c123f3ba3da2d1789fde0f5b18
fb48364fca6bcc8c1e95b4a33507c89b55b95229
2249 F20101111_AABMBP voss_l_Page_047.txt
174056344c21b1270d3c9e679fac21da
a96fcdf64faa9b32218dc0617cb1387665a81a8d
7502 F20101111_AABLWJ voss_l_Page_001.pro
31eeb4a23f53555f52e1c3edfce6b782
7ea308d737b5952970365be6ba0a7cc39d748fea
F20101111_AABLVU voss_l_Page_118.tif
68794fde0816c43353b25dab8db83f93
6c515b0e8728db6a55bffd085d135d19e93f45e4
323 F20101111_AABMCE voss_l_Page_066.txt
0e7cca4e722f1a9986482e679c19e6d8
d6f226dc64d4af43fd3fa08e8ac63e3be17be924
2155 F20101111_AABMBQ voss_l_Page_048.txt
d916aa14cf7d39b90b8178fc37d9fd84
0d96077be23d63ac0d27e573ed491d82f2ceea86
911 F20101111_AABLWK voss_l_Page_002.pro
cfec4f266943fa39928aeaae0bc6a143
5821fa29b2fcc8ccb32725ecaeaabf9cecc0d595
F20101111_AABLVV voss_l_Page_119.tif
942cbfad2b136ba02e9cbc6a85c852eb
acad76f9931d559f42f3a9f827efdc19b949aa81
893 F20101111_AABMCF voss_l_Page_067.txt
1ed7bf838446099dac9b078f59bd38af
c1fb79bdec2201dd7cef529d80210f5b43b5328e
2191 F20101111_AABMBR voss_l_Page_049.txt
cc4db5e5c38b9a5c0906abc0d6a1b863
5c19627437e7f2c34c493fe7aa018a8cf3f8784c
99879 F20101111_AABLWL voss_l_Page_005.pro
4160d176f7fffc23ede0a38c198e66ea
f3612c0bf2ec4c87fb4397eb70a4cd30e22dfff1
F20101111_AABLVW voss_l_Page_120.tif
e77651d6f1711a0fb44471552e3bb7d3
425a7993e504117bf3b8523207e6f3bf1cf78ed1
558 F20101111_AABMCG voss_l_Page_069.txt
37a99a70092bb37771bb33e3bb09522b
23cf36da2cfc9376d32f9a64b64176c75f53c000
46031 F20101111_AABLXA voss_l_Page_027.pro
7dfc3fefe42a18d21c59080b318ff27f
bb42a44dd8b510625c7b84615322857ccf596fa5
2124 F20101111_AABMBS voss_l_Page_051.txt
6b2d7bdf73e8edde661605ff49b7e9ee
0219e4f07cafa0605ffe9086b49fdc3d03d87df5
80955 F20101111_AABLWM voss_l_Page_006.pro
de1b9388339d91bbd5f933556003ca95
47a83850eef066d91b0d68307c0251e7402a9715
F20101111_AABLVX voss_l_Page_121.tif
1353009c5f3f4a0dbdda2e3264f8f3eb
bb788a1209cc9818a38e30ee5a4cd89e6ccd67d2
566 F20101111_AABMCH voss_l_Page_070.txt
f4e40584a6f0999960138c98ae4a8f14
9f8eb8f0ad86718a6064f6d1e4108ffeb6ffa10a
51817 F20101111_AABLXB voss_l_Page_028.pro
0109fe13a46551a346ed96df10dd2986
3cde5320e4ad8e709b100c25ec47c347a831cd52
2030 F20101111_AABMBT voss_l_Page_052.txt
79c3466d4dc298898b00fdb32f2f9a5d
e3188d6d6c22456e36af7916ddc6d9a9eac451dc
40911 F20101111_AABLWN voss_l_Page_007.pro
1268443e4a039e0a9bf0b68ee8f61c4c
422b90d1c4da3c3fd1ebbb59fe372f2bd8a5e466
F20101111_AABLVY voss_l_Page_124.tif
57d333e76176ce35b855959cc5560084
1b550e22c10878e9aee420c8732c03f8542f362b
1125 F20101111_AABMCI voss_l_Page_071.txt
410dcf08fcbb1ec4fc7b2987edc3b0ba
86d70127b74c956dba8b4b89248156b200440448
43028 F20101111_AABLXC voss_l_Page_029.pro
a797287107f8008c77b49c7f0ea22345
5e2240bf307e0da723806fb29e1127a0c505a1ba
2106 F20101111_AABMBU voss_l_Page_053.txt
7b69e9ad59a9376c0cf532f8b85f0ef7
fcd1db67e31e191ca632d036c9bfd250e9db5eb7
77684 F20101111_AABLWO voss_l_Page_008.pro
a27b2a427ffcdb3873154a0598833247
8029781b3461538e93c614cada4d0bd3e92edae8
F20101111_AABLVZ voss_l_Page_125.tif
903f3c990dc58447bed9b468ca050587
5cfc575df724392a88789ec5a9330476d6f6446a
478 F20101111_AABMCJ voss_l_Page_073.txt
d822ed4514bd9031e79714e257dab448
c4f76e0d9c72780a41d8ef5af6f0d0350556ac2c
41621 F20101111_AABLXD voss_l_Page_030.pro
7e3afdc6921734c5d8e93a1abc4f88f4
f0661148fd30ce02f27b0bc8a2ecbc58316d93e4
2256 F20101111_AABMBV voss_l_Page_055.txt
69b9ff832982c9a20c82a89e5f6d6597
d2b81e8995b58f83bbc8ab0f032625680b403ed1
80585 F20101111_AABLWP voss_l_Page_009.pro
04c0043bf8d94cc373552112533ef2b7
5c18ec0d4ad54dab287cf092ff6cac8d348efbe5
F20101111_AABMCK voss_l_Page_076.txt
1686327b81c9cbd644d1777d6cafdbd0
026c01609fcbba96baf84b489467375f6cf52e7a
43652 F20101111_AABLXE voss_l_Page_031.pro
9d580d2a7078119818adc93ce10a3f6a
7c0c991480b1dceea2bc7ab02884748df85a790e
2079 F20101111_AABMBW voss_l_Page_057.txt
2ea76dca07bea3e1a5eaa28b59fcef31
5035f9b2e09d933b64559f669c79d6712a6091d9
2127 F20101111_AABMCL voss_l_Page_077.txt
0831b54259db8ede79aef9418a11a167
0f2033a19dd154406bed21e5a0fe335f2880ce87
17353 F20101111_AABLXF voss_l_Page_032.pro
76a5764fcac8135b8bf3e4eda640d621
92c160a311753ec2433c29150d8f5d1dc9b39bdb
2120 F20101111_AABMBX voss_l_Page_058.txt
54f17c70f814c95248c41bfd8005c3bb
1647dab21fd9c5bb2233667af635dae63f0fea83
42671 F20101111_AABLWQ voss_l_Page_011.pro
2e4ecec057b70c34e9637e70758c29bc
3a0d8529a960916f79209ea62c59277d37d6da42
286 F20101111_AABMCM voss_l_Page_078.txt
4320df6c173ee02e41129b8e1dc8ba54
bf496a9242fed5f1b97273d38af063d256402c66
13455 F20101111_AABLXG voss_l_Page_034.pro
0a456a67cbc853d7caa7253b15516e2c
e079554c0aacdfb660bd90c1d26ba87bf73bab8f
1933 F20101111_AABMBY voss_l_Page_059.txt
8069af47d3db9f7c3c643a7338e80105
41c44fca8ef779bb955a2ca287339edf53c3ee37
34060 F20101111_AABLWR voss_l_Page_012.pro
0dc9544317fffe86712dd392c4bacf07
a78f130984f59135d437dfb4ff5f7255e93b7339
427 F20101111_AABMDA voss_l_Page_095.txt
1e67a89d851f0303a2f820ae53d1b621
2fc4e14747324275567e775aa806dca085aa3bc5
712 F20101111_AABMCN voss_l_Page_079.txt
952626faa7b11af441b279bf19851071
b24393dc5da2683284d1648587ec9f7e312fe36c
18969 F20101111_AABLXH voss_l_Page_035.pro
94aabcc845a506f2642175b06951a504
5b548854783166da1e5eb6eb3fe8243ac661cbf0
560 F20101111_AABMBZ voss_l_Page_060.txt
73f716c564bb98be0992865bc8ca5100
90d47a43645f2523754c0ea2b6662d079626eef1
9690 F20101111_AABLWS voss_l_Page_016.pro
1da20107c2d5fb366484f7d23a7c335c
f9b15bc86d4633fc86cfd7b186c5ee3dcb41f59e
402 F20101111_AABMDB voss_l_Page_096.txt
e37baec639734b4bb72ae42ff7667259
1534b4b7d108a656eacf16eac4c4a0fd7280a66e
488 F20101111_AABMCO voss_l_Page_080.txt
3d394fde091521de27a930a1bfb1f5b8
9b1a5f58841b7dd2935c918e8fea2724ed6298ce
11575 F20101111_AABLXI voss_l_Page_036.pro
e5cc8728aafb77fbc090b08faa3b7226
a3f9162b0724d244e9f566d3240e3db94d9630c3
5412 F20101111_AABLWT voss_l_Page_017.pro
55e3c678bbbf168c1289c5c54ca61b86
82b4f9c92c5625210b5f0da2451a8c3fdfa1e7b7
F20101111_AABLAA voss_l_Page_085.tif
6db24c7df298c754621aeba5b3af0a83
510412c8031f8e4ffd203f8c9a8789de47080a04
880 F20101111_AABMDC voss_l_Page_097.txt
286298f486a79ad884b72d27827c0407
2160f627e5a75255860f7708d088c38779057d85
915 F20101111_AABMCP voss_l_Page_082.txt
2756283535e401175699122089cdeb1e
beeda45e85391a3a0240746f9e28cc25a15476a9
8151 F20101111_AABLXJ voss_l_Page_038.pro
e2ab85bb124ae2d4fd6a50375f0df156
342813eb33a425cbbb276b83249a0ed048a1627c
51371 F20101111_AABLWU voss_l_Page_019.pro
1fd0dbb2b90760c26ba1196a8bc957a8
4d69427d35565bfcdd2c7c0c2cac78ad42bdcb70
F20101111_AABLAB voss_l_Page_054.tif
2dbe740fddd854f7bf9a453cfa1188b8
27dbbde1de6c79d10c13443ee80bab038a58b5b2
783 F20101111_AABMDD voss_l_Page_098.txt
8b6010ca4155f01d3ceb3fe3c98f8522
a8d86c8f1134eb3fde7710637ddce0ba96fc6057
4388 F20101111_AABLXK voss_l_Page_039.pro
ad4d04f192e251d568aad61268347785
65ad28ecd625ddef6af92c1bb0ed5e0e774baf20
55566 F20101111_AABLWV voss_l_Page_020.pro
942ee6658272633c17bfed2466db208f
487cf8e06e6bee48ea52a5df72f113daaffae5ed
25065 F20101111_AABLAC voss_l_Page_121.jp2
07a23c2548b31d4f6897eb6d4590ba4c
384ff08934679f331c851dcdae8d870735c6d5ae
275 F20101111_AABMDE voss_l_Page_100.txt
1b29b9e9277f69f4bdce22541d16d0aa
b924990263f1484ddd73beb7d61c8530b87bebef
1596 F20101111_AABMCQ voss_l_Page_083.txt
8829450658c02bd87cc11b3fb912b5ae
7f498a866c80b5ada909a7b8a22c610272ad6a37
2539 F20101111_AABLXL voss_l_Page_040.pro
196f11785d943f88be4d48b625e9459c
5417b4a270867f7bb0ffdd38802e813754740cdd
57930 F20101111_AABLWW voss_l_Page_021.pro
2d0b2a2d92f50335bfd20f2b53c8ccda
daa350d66f6c37fb81382ddc03d73c3582b4ab80
2130 F20101111_AABLAD voss_l_Page_103.txt
efa3f5b83c9d633cd230a36dc2d87038
c66adcb0fc31a0131a855e78a087adac92cb7b6c
399 F20101111_AABMDF voss_l_Page_101.txt
96b52b7fa11950a6a2081d61650cd483
5399f4b7a88e7c6a27387278d6c4a630e3d8f3ff
359 F20101111_AABMCR voss_l_Page_085.txt
c9dfb1990688426a6dc880f92e82f047
1461d6dabaee80126e3c9b3519e46a1d39bcddee
6622 F20101111_AABLXM voss_l_Page_042.pro
b642743e531e9d45df86c9c638be894d
9d428ddb32b233a0ed83c995ca69bf5b75ffe899
48167 F20101111_AABLWX voss_l_Page_023.pro
a36a1b3f5b8681136ff8302f9aeecd44
82d6469e2530f13ccbb8cc84397530f2724ab7a5
6764 F20101111_AABLAE voss_l_Page_008thm.jpg
cd4c506d20cb6e21fb66704ffa655962
039e5115215153e72d9a3c0527a82dbea1c78c6d
527 F20101111_AABMDG voss_l_Page_102.txt
8791a21a1ff10b4b71cb73f85e7335a7
7a77d1c94a3bde95b46417b6ee2b04464ef75ef6
15335 F20101111_AABLYA voss_l_Page_063.pro
9f70c290d98372ef99ab7b3e95306745
cd30dd17ea55cb3bda455c9cd7a3792960063e18
843 F20101111_AABMCS voss_l_Page_086.txt
aa350ed9ff72be9f10bb00e38df44209
12ed86556dbf8a390827a985074cca23a5a41bba
2625 F20101111_AABLXN voss_l_Page_043.pro
a4274f44e039d44e4b166c4214aaed61
09b3d7c196fbf6d0205174c720e2614e21fe0ba1
55477 F20101111_AABLWY voss_l_Page_024.pro
392b5022a96e3d59f1e3b8baa50a5c2d
757fda52ca02b40a0d89c6055e59b27980385666
27025 F20101111_AABLAF voss_l_Page_110.QC.jpg
98d119737b9507962b93d6b3c646e207
76fc094dbcb4715608824060d67f1abb743b68b2
1795 F20101111_AABMDH voss_l_Page_105.txt
10858b17ac80c25d5fd656ef985fa903
854f1ebaac786a91454dca56775da4fc6b384605
13999 F20101111_AABLYB voss_l_Page_064.pro
320b2e7c978f511802bf0a83a47b86c9
ccefdae97683d08050e6bce4039dac2815ddb371
1187 F20101111_AABMCT voss_l_Page_087.txt
98e911eff5d8a2a061457679369d07d5
89367cbc6543f91e33c7777c246055e4e14ba1c4
2342 F20101111_AABLXO voss_l_Page_044.pro
13efef0991b38510424df2f341de7383
3c11a4c4f1d3f838210b4080f1dad8b8101a8d3b
31692 F20101111_AABLWZ voss_l_Page_026.pro
4adb8cb78d24f8aa119e9f3477095e9e
c9771f5d197d2b9d926bd59f66cd7c30f41cfaf6
F20101111_AABLAG voss_l_Page_123.tif
610d289754eba56d4da8787ec71323a2
f8982fcae251b86f721711a5b286035540c2967a
2021 F20101111_AABMDI voss_l_Page_106.txt
366a1e8491b0ef382d7ac9bc5056eac2
fc6df0d9a5a05d50daf6c1da9565adcc87815a76
37359 F20101111_AABLYC voss_l_Page_065.pro
47bffdb8c280029093c6b07a7de1dbb9
c124f915209fb4f9a74ab4a1ac70d21148796f2f
452 F20101111_AABMCU voss_l_Page_088.txt
b91e8b4865b6ce8cb1c472f94c3ee9de
0b604f31f4b2f9a9bc0ae58c5e37414d2ced32b7
F20101111_AABMDJ voss_l_Page_107.txt
f7e401f78798948656de89b9cb719cb0
32cc1a15c9cb770f587ccfb2b883f18b2e237b58
50528 F20101111_AABLXP voss_l_Page_046.pro
48e90faba36df7a51795c4a59c50583b
848c889b30840dff8021c6fe6cc720a144def1ba
27101 F20101111_AABLAH voss_l_Page_123.QC.jpg
7ddab5cc3ec5cd229dd50584a79b93b9
2ef08383e98519f33c6ce7ea814a20433096b5a4
11649 F20101111_AABLYD voss_l_Page_067.pro
05e858c6934eee1cfd94ff4eda4267b0
1d3a091850d7f0a854a082b591ceb890f01e2197
1789 F20101111_AABMCV voss_l_Page_089.txt
670b53c54304eda7e27db0367a18826d
ee8cc87751acfeece8b4c003f75bb9e3f57d00e9
2170 F20101111_AABMDK voss_l_Page_109.txt
9fece8a2a97c6d93b661ff74af06e349
ce49658109490f943aab37e6a64a9f2063a495d7
57448 F20101111_AABLXQ voss_l_Page_047.pro
be02c643f5cb467006cf1445c02b1209
4bc947250e291b1902adee7beeeef4e419c6de46
68982 F20101111_AABLAI voss_l_Page_006.jpg
0419d6384fe4df935ca7ea57cd212038
d952c398c8591397094ff66806ad90aa1d188493
8523 F20101111_AABLYE voss_l_Page_068.pro
66b8c01a2c255508166386462795902f
1611b2ed645303a9b1f9a2436847b3c201408e44
1997 F20101111_AABMCW voss_l_Page_090.txt
6cd118d2ea3c3c99526484b1bedf1182
fcee5a780cb5af165fd3b60c43c918683bb326f2
390 F20101111_AABMDL voss_l_Page_111.txt
73843af6a9bd38ce27f7a3d1100542ac
d35d058057765b55eabf1f3bf708260ad3ab2b0d
85371 F20101111_AABLAJ voss_l_Page_123.jpg
62cb2ff26e959aa71d3c4b3a7117c3e7
897dd78dacc69c77ca5210dd0a1ebd3cd7accdd1
11486 F20101111_AABLYF voss_l_Page_069.pro
023b077509c4e2d2d71655e55d99b21b
064acd525724275da6cc4108e6803426569adda6
2200 F20101111_AABMCX voss_l_Page_092.txt
5fa2edb136de4bbe41b974a9ec6530c7
785c643e6337b3fe9ff4da7326bfc842698c2c0b
F20101111_AABMEA voss_l_Page_131.txt
23c257993ca62887db86aecfd3a92254
e4a7f8bf5fb684c59b559cd63877f36232d7f62a
252 F20101111_AABMDM voss_l_Page_112.txt
c5d72c4ba9ed6e2890122f4b7d508c37
3c62fd99499b901589108fc8108fd5f26c6918af
54635 F20101111_AABLXR voss_l_Page_048.pro
152c02b3aca749bccc0c2658f8fbf76b
5d03efd5bcbfbca3affb5d8ffe1c0d87f1e02c67
290 F20101111_AABLAK voss_l_Page_039.txt
54eaeaff4b6471a81059fad398ee05b4
1ca3c87dbabf4b659575fcc412a56ccd1f270c45
13609 F20101111_AABLYG voss_l_Page_072.pro
a2cb089ec86fcffc48fecc453962020b
88334f511ef985ee669430414dd98fd9a1d4e0c1
555 F20101111_AABMCY voss_l_Page_093.txt
a01bfd067563e0473b98ef153a7d3da4
fdf32522b6dcbde60d30513f5146b9826e0da1fd
1920 F20101111_AABMEB voss_l_Page_133.txt
3b0bfcde3872f7661240d1f759ccdff8
4de767720e7192f350180148d3639411a5ca0881
867 F20101111_AABMDN voss_l_Page_113.txt
07a8dbbb126f13bfa88ec3fa36c3227f
f9e4169037d48802016d6ae4735aee4775fd9db8
55711 F20101111_AABLXS voss_l_Page_049.pro
bea140fa87e8f4c9ac10609f5523d260
b23b9c818f5486b8300b6e2549e4db68fd64c375
813 F20101111_AABLAL voss_l_Page_035.txt
3882a02752a015354c23166a74c16a76
12add108da20bee495fef0e7132d5f9c549af771
7646 F20101111_AABLYH voss_l_Page_073.pro
86fa3e0f5db4d8aa919b1933b16db890
4ed48e9e80187df63ee9aaa2367ef374300f01f2
346 F20101111_AABMCZ voss_l_Page_094.txt
ce42889f819259741b805a797e1ab3ec
003e3513d61baee38c32e69a09761297f896e815
2126 F20101111_AABMEC voss_l_Page_134.txt
fc3f01aa6f6554afae36aad0acb88752
f85c10da2af995e9afa743d4623e109723f72881
799 F20101111_AABMDO voss_l_Page_115.txt
7614dadfadff564aa788287b25c54005
4787595cb90e207dc887d8b39b260126bf0b2d7c
53835 F20101111_AABLXT voss_l_Page_051.pro
8a0e0ce3856c21879c1e487abdb41f47
de1ba783a30739eba4b8feba8d47ea5abaa9615b
7110 F20101111_AABLBA voss_l_Page_111.QC.jpg
e07630401726bd5f200048107cc4e8c8
fe271ff90b7d9fb08cffca6a8f5d713245444023
47403 F20101111_AABLAM voss_l_Page_007.jpg
bbb553c5e02b9cf8316c206459261644
2665f26ddcf86ab6720770508a893e59e6b039de
11113 F20101111_AABLYI voss_l_Page_076.pro
a7909f806d24eb36f9edd5e053ac72d3
4ce78572dc679c8646f9761267defa88136802eb
1469 F20101111_AABMED voss_l_Page_136.txt
37c58afc544bf777c957cf8d8e9f9b13
3bfbef785dcf448317c64bc8e3eabbf0def02814
638 F20101111_AABMDP voss_l_Page_116.txt
36f4fe92625f17cba63c3c36424c94ad
f9842a3ea2780d0fc0d43f7f911a4acb0a31ae87
51139 F20101111_AABLXU voss_l_Page_052.pro
234d2867d52fc3ca059767a25757fad5
13379977823dbcfd8e7ebc925d1906aac5872dfe
2196 F20101111_AABLBB voss_l_Page_043thm.jpg
ad1201337295829ee0f4c5783ba085cb
a7263ac92645dba5aa57baa3b264730e16d0cb08
38728 F20101111_AABLAN voss_l_Page_004.pro
7fca7b731eeeaea46402102f83a5a896
4c81738d65f21393599b4466a2db66f147b8cd36
37227 F20101111_AABLYJ voss_l_Page_077.pro
4412960e50edaa7e71edae5be7dced29
697a3211960270749424a137a6259afcbf6147ac
2164 F20101111_AABMEE voss_l_Page_137.txt
a973a1bf1dbd4cdbdb5d4f13e2190364
4b5495ec85cdc4035fb228d8599422029c8233fc
440 F20101111_AABMDQ voss_l_Page_117.txt
5eba3cd44af17e68206da996c5f0157b
3266aadcabf15e7bb4002bb5920600fcc4676cb9
53400 F20101111_AABLXV voss_l_Page_053.pro
96e7ae53c8847646eb115612259dc8b9
db5e57efd1e5f3480aec2a3d1bc898c43ef735d2
48364 F20101111_AABLBC voss_l_Page_143.jpg
ea5cc14aaf7cb8731a2dc62c2e25f19a
d5e1bf12295ff337a39c0f5e69c1f98afb24f525
1051957 F20101111_AABLAO voss_l_Page_054.jp2
a04711d682dcdc0ab2ea3edfe33fa941
8a85004e87e784936320622ee0f69e5ba76c1c16
5063 F20101111_AABLYK voss_l_Page_078.pro
1e2b655d846677d6359d0fd3a9390987
925d0c3f2a193578fd9e1f8022e83a36c70f313e
2358 F20101111_AABMEF voss_l_Page_138.txt
f99e9f702f2f5c202f8d2da22c3362be
e928bbb78891f7bdb79469557ac5e3ed3cbd067a
722 F20101111_AABMDR voss_l_Page_119.txt
e32b0aff8821c6b60f0503ddf07089e6
b3d2971c624df9e08b0990b59ab59e29dd9ce8b4
56181 F20101111_AABLXW voss_l_Page_054.pro
cc6f10be9ce38786042757f8c381cc42
67cf95177a05b6535489a592c146d98a211b0038
7395 F20101111_AABLBD voss_l_Page_135thm.jpg
d8a81ad4c2766c139524d81c643bb70f
19259e0df96814c3a6df87e780270cb7a947e90f
F20101111_AABLAP voss_l_Page_029.tif
fc30fd4ea63bb30dc6b898d339f2c36c
6aeb1b6b52056ceaddecd7eb831b5a90d1845785
14512 F20101111_AABLYL voss_l_Page_079.pro
f72778fadaea07ca4a8b8366e2cb0050
d1eed7511203c0f86c75f1510585cec9387fb860
1953 F20101111_AABMEG voss_l_Page_139.txt
6881776f67e8a98eb9b172842b3ec382
3cceaba123cb568ef378c3e4d92dfe28c22980c0
598 F20101111_AABMDS voss_l_Page_120.txt
4de2a71e7e9e132cf9ec5925e23796a5
96c1e651b3ebefc5f7a91f56b802910c1c447761
52810 F20101111_AABLXX voss_l_Page_057.pro
2c82cb3fb90d0ea1248560e5031e819f
be7985e2d978ce3fe93fffffde06912971ed0cc7
500 F20101111_AABLBE voss_l_Page_084.txt
ae87cd621f7a3d6093449f7bb79c25b6
45faf3346b246496901d4dfa68e4b487d6bc3884
8647 F20101111_AABLAQ voss_l_Page_114.QC.jpg
424385b120bd77accbb2c97723456fd0
fac2d99043e2b9378147702ad7268721a19c13e9
17530 F20101111_AABLZA voss_l_Page_099.pro
ba5477b9efa04dd0a11bca171077543f
a83bcde57b3db5906d5b9d2fe7928f3aa817235b
6583 F20101111_AABLYM voss_l_Page_080.pro
b670ac6b8784bd877b54913fd561c673
ae2fc0c41f7897a1f036a7e5d3ddcb2fb547c654
2067 F20101111_AABMEH voss_l_Page_140.txt
78a093e4fe0380401e9a2c1cf390a765
df468cb3ac284ff82115999daf0bea821aafc739
223 F20101111_AABMDT voss_l_Page_121.txt
749f664a62c0c7ac3d199d0ae4ce7c0b
fe0ac5ac1fac7f4f1457e913201dda32c04197e1
11320 F20101111_AABLXY voss_l_Page_060.pro
715567987c6417c91adf3a6f6be2e12f
c27f438e7d1e22b7ba69c386984968a293b97a6e
2125 F20101111_AABLBF voss_l_Page_126.txt
289f27e8fc458d9fc5a6793b15ec3202
2d632112a015618ca114b83e020dca31a9a22eea
F20101111_AABLAR voss_l_Page_022.tif
6767544ea13e5b14db5bfd2ea5f003c1
a66e93818e57db8de3f589b10debf1339e38ef39
4584 F20101111_AABLZB voss_l_Page_100.pro
4ca13f1a2f056de3edd3777529aba373
36a6f954eff8ae68cabdc0aca2d85545d4e69c16
14165 F20101111_AABLYN voss_l_Page_081.pro
e3012067c32638e60a709729aefed608
fd26835b07f28c74b3a74056c8dc79a665b9fb97
F20101111_AABMEI voss_l_Page_142.txt
96b983b402ebe561012e283833c5b9ca
da716c1319dfc74b3322afabc373c58bd3527559
976 F20101111_AABMDU voss_l_Page_122.txt
3ff1199fc821a52c7f8712981e3b40d3
10bbb7dd5c8ec68cfa4ae2a8a9304194b55d752b
8620 F20101111_AABLXZ voss_l_Page_062.pro
27c59af9276f3351e2927988f14852c5
706030df57dea1c82e4ba85ecda43f6c374319f3
26902 F20101111_AABLBG voss_l_Page_053.QC.jpg
d7c8c510f388d31062177b8e9605572f
1ed00ae76cf67c58f840d9bfe96d8c835924151c
6946 F20101111_AABLAS voss_l_Page_050thm.jpg
f52032f38ff7800447efc615347c5342
de3dbd6a833473f00685d33d7a22d5c2301e6322
51764 F20101111_AABLZC voss_l_Page_103.pro
c1fcd4c8a31875ffd4222b7487bdb3b8
f947dd85e788a84d083562e331d940f1f6e66e93
38769 F20101111_AABLYO voss_l_Page_083.pro
adc066541c81b9d4486cd0da806322cb
2728b789f34338c0b03ea08cec869526c4d04500
1005 F20101111_AABMEJ voss_l_Page_144.txt
5f8b533faff6b7eede39610cd1ba91ab
a2743c9486acf307970289870400ced43a26aa47
2047 F20101111_AABMDV voss_l_Page_124.txt
07dbc1e79fa7529389632da7bab707e7
77bcf590b7c852245ccc03833767a2f61770794e
14771 F20101111_AABLBH voss_l_Page_033.pro
2f026fda02813bfc6af9f90905a7a066
1b94188154f6f2b1efb0957f0dc954e77f90f8b0
6220 F20101111_AABLAT voss_l_Page_029thm.jpg
f9f507346ebaf87412e35159734cf47b
913643adce5ca284fbe5b75bae87cf30428602f5
43721 F20101111_AABLZD voss_l_Page_105.pro
c7db998271182efd786f0af9918907d9
a493c79df4b005cbe1fd6d7fa27828606c86dacf
6499 F20101111_AABLYP voss_l_Page_084.pro
c5afeeabe7bdf3aa97d3ad7b562a28c3
3bfec8ff216ce2dd014e84a39a512ea37d4fc632
F20101111_AABMEK voss_l_Page_001thm.jpg
0e52c850cb8e6e52e281e5c1cada1c1e
a8bd853732485f6cf876712e34bbc22face1af5d
2145 F20101111_AABMDW voss_l_Page_125.txt
5654c68f21da52eb6345daae5885223e
63b81cde44f12faf047280e16e560d76934075a4
26123 F20101111_AABLBI voss_l_Page_028.QC.jpg
b90a010ac1cbc9abf49a766b3820c21a
9aa7bbee7c6fd6ae373a0ead89e8adb952b57859
F20101111_AABLAU voss_l_Page_141.tif
725d542650f78dfe778e4fa958637376
5f85e01574dba536bc0a2343564589cbe047da59
54583 F20101111_AABLZE voss_l_Page_107.pro
b024599c663df5aad4c4a5323d6c40bb
1c741d97947f729414e57bf7f13e05d54dc9b031
5599 F20101111_AABLYQ voss_l_Page_085.pro
4fa0e416e172355dd625d99d8be725a3
2021e3cefe38ca538defc6785a2612234c0d2e60
2404976 F20101111_AABMEL voss_l.pdf
7f06dbe1254d597df8eed4795cd5708b
034cc4c1945fb9cd4f94b0b6de2173ec5493cba7
186 F20101111_AABMDX voss_l_Page_128.txt
debffbbac72e6747d8bf5e1ea42636ba
989b51cd4384014fa15886c4b5dc96c2068a5f27
50483 F20101111_AABLBJ voss_l_Page_025.pro
9642921a412f48314a6802c317d4008c
0fbb9e8fa84715d7c1fb7bf5da9cead1f00bf863
52922 F20101111_AABLAV voss_l_Page_110.pro
44493b98cab411e6000bd8df91a23b6a
8d2ea0a23ab1aab9c5b0ccc355aa24ca6f4a4ebf
54556 F20101111_AABLZF voss_l_Page_108.pro
f857cee7405d1928028db597a4cc67f5
2b98946e5670b0d28bda81449ae520e6d3cd2293
14158 F20101111_AABLYR voss_l_Page_086.pro
4e552d23f86845523a684387563f7bf6
11f4f0f9830704d9f9758f02285da951647f41a0
29323 F20101111_AABMFA voss_l_Page_010.QC.jpg
496584840266782b47f003e53a63cd04
18a665bfcf5346cfa2c29def2b81dd836d8cd71d
164863 F20101111_AABMEM UFE0021980_00001.mets FULL
9fd148b8040e0babaf32e263939061db
38c696e970375bab0c6c3d0d646eaac65f25dcb7
499 F20101111_AABMDY voss_l_Page_129.txt
afbc2f61fe1f58d78ac8c4f37952394d
488272112f2406f3f6e346fca2b7051c6dca151e
6093 F20101111_AABLBK voss_l_Page_066.pro
a48df5e34b2bd98d164771f6376c7f72
a9f24a2e2af2a6efd61035e85714be0308240e50
6598 F20101111_AABLAW voss_l_Page_105thm.jpg
b147bb31ec3be2662e6c34d7c1691b52
a05b13219d4d8ef4f85b1e4d2c503114bf3368d8
55445 F20101111_AABLZG voss_l_Page_109.pro
2bd947bcede3022d6d5b2e73814b7eff
3aee71f8e2d518d6d1897afdd51ebd748d486f53
7568 F20101111_AABMFB voss_l_Page_010thm.jpg
51a57e2c675b5f6a4db0884780011126
ea6dced79e7ceb70b38056d73bcd66bf33b06273
7170 F20101111_AABMEN voss_l_Page_001.QC.jpg
e5a3ff763fedf3680d435f8421a4e034
b96835368cd0da28f4b0f1a1486ab8898e6cd1a7
422 F20101111_AABMDZ voss_l_Page_130.txt
aed223b7b7e1ef8710e8a1190d06bb90
ee686d2670d686058e6bcdad80656e29c5e2e2b6
F20101111_AABLBL voss_l_Page_105.tif
c00412c817068b4859ace80683d8ad79
204267a539821db47366bc6708db381f5a1a43f0
1051971 F20101111_AABLAX voss_l_Page_021.jp2
898627bcc04beaa0c77a93b439909581
8c80402a0275f9882fc818c0c6ac9da5b7c02f9a
9794 F20101111_AABLZH voss_l_Page_111.pro
a1f8287d4c37655fd4021ec9ef1c1d49
ca5cec41bb32a8907cdd7db54dc8d4fe2eadbd09
18188 F20101111_AABLYS voss_l_Page_087.pro
ebcf12de28d2d4daa416f505eff72260
5114c91c63a83f51a2390419a66fbdc4da0bd669
22488 F20101111_AABMFC voss_l_Page_011.QC.jpg
5761a4e4e9110355511ed32aba178a69
ecb0c9a9c833745b7ec1a6471440d3f3ac09d766
3108 F20101111_AABMEO voss_l_Page_002.QC.jpg
9fe043ddae88e45e786e61e6586bfb22
a96030902fde911ff11b0e14ba1d694675cec747
3485 F20101111_AABLBM voss_l_Page_095thm.jpg
14c0e39709039cba8ada8de4b164c2d0
dd1f5c5cc25d6afb77ac0ad00163df242233ab30
F20101111_AABLAY voss_l_Page_041.tif
0c52fad66427fbefb1acbb5f375032e2
25c4527557df5274c83c1679e6ff71c1ef6184c4
4835 F20101111_AABLZI voss_l_Page_112.pro
dd855e57e1d33062bd6bb5b855084af8
a7b0a2b2b37dba9611813dabf738ab93ad7b1529
50282 F20101111_AABLYT voss_l_Page_091.pro
58db1ef4043b3ec6ca94beabc95be721
4d10da361035154533024d7bbeb58bc772502a92
25989 F20101111_AABLCA voss_l_Page_142.QC.jpg
f6369c8ebede48563f8e50dc19f18578
8b2da3e5a884f69b13a549ecac7ab2cf997ab120
6274 F20101111_AABMFD voss_l_Page_011thm.jpg
dee0ae5a29718c86092231bf2b52ffc0
ecdfd703577260dc51fd0ff5cb9b5719f1c6affe
1339 F20101111_AABMEP voss_l_Page_002thm.jpg
d31f239175787a5e1b4caae0c1dce179
a726d4c7bee1d7fbba1e52ec5f39bf7a0f1a9749
F20101111_AABLBN voss_l_Page_009.tif
47bce65968abd6911dde10da378f2e8b
bb07506b923c71c54f94cf22b7cc38a4f1441fc1
53798 F20101111_AABLAZ voss_l_Page_026.jpg
ee5e21596341fbf0f382e487719507da
07b271c9cdb46d1923903f6a3670728075d43190
12857 F20101111_AABLZJ voss_l_Page_115.pro
597a6d24e22cfeff20fb414a5c2c06d8
1ac6967542a516ac438c5b10d230f356478586b5
54991 F20101111_AABLYU voss_l_Page_092.pro
d9d0351ae2fac035014baa39f5593173
f3e8b30b65ac23e3b2a728efbcf60737c807abaa
26429 F20101111_AABLCB voss_l_Page_106.QC.jpg
4182c2adfd75d3da10c2d04906e22505
af9f42c9824acad616b0d69013a11f1e1e06700d
18277 F20101111_AABMFE voss_l_Page_012.QC.jpg
e51458f11b87b686dc8eeec3d986420b
a73c6ab4a5fb0c313ef2dda4f3a4ea124c0f47c2
3051 F20101111_AABMEQ voss_l_Page_003.QC.jpg
59754e053c22ecc648b66649aabc1afc
8bd60c50dbed5c89cab09604268dfcb360247b7b
17542 F20101111_AABLBO voss_l_Page_075.pro
6979bc94bba0d7ea26836b2f6f19f004
aa958fddc1192dfd7c054065817d374470f5600b
16312 F20101111_AABLZK voss_l_Page_116.pro
d2221d28de8e18c3275b9da29b97971a
5b2698c06262bb2b856931043d8422498bb52b01
13958 F20101111_AABLYV voss_l_Page_093.pro
42979b2d2f10aaf478d010d093de4d64
9cd98e507341fe1de3d59a35c84616ff1543827f
108082 F20101111_AABLCC voss_l_Page_112.jp2
08e3ddf8df3d23e605eebc9e0c35daea
cab08c5307ea6afe4d371b9f438f5e41e82aa5a1
5174 F20101111_AABMFF voss_l_Page_012thm.jpg
8c1026dd85419ca58b3fc82f03356e08
c8206e5270fd0fabb24cc4df4143ecec377f87fd
1325 F20101111_AABMER voss_l_Page_003thm.jpg
1daf228d173d8bc973c091d3efd4d667
dffed45fd9afac0b5c265adbd033ac027d2ce85d
26131 F20101111_AABLBP voss_l_Page_050.QC.jpg
be1bf5ce77847a179ca4213ce9989ad9
17c954c2199e9b50378facb6d683a9b6639723dd
8096 F20101111_AABLZL voss_l_Page_117.pro
61910dc8c96f8920daeb1ef7ab55c309
a3b800c3caef06e4c8b88fee89d4128af75bf285
6272 F20101111_AABLYW voss_l_Page_094.pro
2ad0d8f5a92e3f2ffc275085eeb532d1
ca2fb03b07cfb1b52c4a926b143017d3eebe1dba
11553 F20101111_AABLCD voss_l_Page_035.QC.jpg
56da7c2773490da05460770ed7d20ac8
87ae81ebd866f5899df3786bce23257f2b935a1d
27272 F20101111_AABMFG voss_l_Page_013.QC.jpg
4b03cb6736716ff8ed90e4a95370af49
87c402e535714d767a62a2e88b351b4cbecef089
20837 F20101111_AABMES voss_l_Page_004.QC.jpg
3316a0cd7ef8c561373d0c7bed20e34d
46c1b493eede8333fe77c912770fd6993d669fc0
13155 F20101111_AABLBQ voss_l_Page_113.pro
b56056bcf6ca3e42bf502489cb842cbf
8ec2685730fe7c5bb6338687a8f4742fb0f777aa
10319 F20101111_AABLZM voss_l_Page_118.pro
2e64e8aad87e31fe8703a48079da98fa
c8a12486f4bcd4ea1276aa42c37187567c15c540
8844 F20101111_AABLYX voss_l_Page_095.pro
eb32e56d099372aa633cac934aafa7ec
4623a180df7005d295dd1acc72a6a41b1273a64e
106698 F20101111_AABLCE voss_l_Page_010.jpg
2e527df6c600fc4b37e342c3c56b022b
4c61bbe41275858fafbb8e717b7fc00ed4b30141
26529 F20101111_AABMFH voss_l_Page_014.QC.jpg
15e3064c5ee28e4cf157bc051ebd1b1c
ae6b535dc98daf2422ff0d6bf396c001ac2a88a9
5786 F20101111_AABMET voss_l_Page_004thm.jpg
b1ca4d872bb53427bb461b67f5f6b8dd
ec64ad1ede9d6def59940aa45f61d8ea5ad3ade2
F20101111_AABLBR voss_l_Page_014.tif
20d8ec9081e25efab41b0da4917467a8
312b140359717b13e3e3c4249a1223bbfcbbfa7b
F20101111_AABLZN voss_l_Page_120.pro
8c685ea42de2896880e85f333a8c0a9d
02edcd6a04025ccb55bba522afd99b41bf24cc62
7909 F20101111_AABLYY voss_l_Page_096.pro
0c3ae2082ede5b95b24cec4aca7c15ef
ea5273dc483e91a04a4c082c30e370d41320bb55
F20101111_AABLCF voss_l_Page_091.tif
5e6525b257ada0600969620ae7e2baa4
ee228cdaa2771f31ace1922a426a92ea5fc4d203
7197 F20101111_AABMFI voss_l_Page_014thm.jpg
f7e5549c16d281e64981c350659733be
00ec7078ff2535e3ce9fec1a59cfe17036370198
5384 F20101111_AABMEU voss_l_Page_005thm.jpg
c38a06c8a024d89c3254d66fc5294799
c7e36750e45965abfd104bfa0419b92c8f9f427c
2239 F20101111_AABLBS voss_l_Page_056.txt
921d9b9b6a4275995556fcd6378f412d
7e9b1d02f999bb55acad19600cee856f41fd6820
4236 F20101111_AABLZO voss_l_Page_121.pro
0b215f9369c91a567b631092672828f3
5a960b6bf5a7416f7d96a6ddd42dca8235d95844
15098 F20101111_AABLYZ voss_l_Page_097.pro
397d5b4a1e4f9a3120ac527ab4b30619
3457beffe9868ee9cc3cc695a0105bfcce85c4c5
F20101111_AABLCG voss_l_Page_140.tif
22867109c1f609f871a87831e061d566
3191f8b2d6390e43412084b004309ddbc18ed2d8
27212 F20101111_AABMFJ voss_l_Page_015.QC.jpg
2219e56ae726684bcf823479c20f49f6
6a7bcec523ccabc56c3fcd0879906cef6d1c4944
19358 F20101111_AABMEV voss_l_Page_006.QC.jpg
c3ad23ea39ef629be1b16fddcb8ef8e1
d60806a9ca24bfed6d71a54efec714591971d5cc
779 F20101111_AABLBT voss_l_Page_003.pro
da5898422a028297090a055f3c484d75
59fdc196b10bd511e12d3149ed4e32aa569e725d
18454 F20101111_AABLZP voss_l_Page_122.pro
59b2b1dda67223d4e413872225aa4f3a
6aa888de066a0fac548de0ccec2c96867fdfee3a
F20101111_AABLCH voss_l_Page_117.tif
0e4043c928ea4101a62a58c590eceab2
a924d4069a8c993f56114974acebda72235358fa
7324 F20101111_AABMFK voss_l_Page_015thm.jpg
732990591d4168526572ddd4716a0325
2be97ed5f51a5361da8561a75f9c8781a47aee42
4825 F20101111_AABMEW voss_l_Page_006thm.jpg
2e0337a0e4171b21b67f4f5e8c339f79
2970274b685434057ecfd9b73c9c5dde027df710
981020 F20101111_AABLBU voss_l_Page_031.jp2
fd10836a76243e603d13bea93764e5af
ca5f0d74c4893e5dd0ebacb95d307b2e006ffea6
52620 F20101111_AABLZQ voss_l_Page_123.pro
3951daa1f4e4997246ac45ff5821c878
cae2a0586402144eb02767094fe4d391bf3feb37
1088 F20101111_AABLCI voss_l_Page_143.txt
8849a8fdb1f13cd52448e0cae2067181
532e325a131403be5f73fdf10d95a3eada6b5f93
2434 F20101111_AABMFL voss_l_Page_016thm.jpg
242e2fb0d1445da0ee2f6c336d332f78
c7c9c90d7e815a8d1a567b267c124e2c07142cc2
3794 F20101111_AABMEX voss_l_Page_007thm.jpg
b7cbc5a5df7a1befc2c57cff511c234e
f02461d108fd13f8a1315fa4827c92d15be61ef7
54274 F20101111_AABLBV voss_l_Page_126.pro
86d1afdcd23c889ac7d88cd5ad2b1027
b343b8de3283e78b206d7112fd0250ca67c6ff0d
51131 F20101111_AABLZR voss_l_Page_124.pro
11035bff7ffad28cc49037ea5c890b45
8b838ba0cb6fc516e3d908762186f763a8500390
4234 F20101111_AABLCJ voss_l_Page_088thm.jpg
64af72320e8acccb37369fe3904b5c07
c6bbe17febf39b4586995b057735458ef195f136
7078 F20101111_AABMGA voss_l_Page_025thm.jpg
f6aa0576f31ad3291c16f5fbe1be9bbf
540cf1df60900422588601170f4716ac05356094
12237 F20101111_AABMFM voss_l_Page_017.QC.jpg
0618d8bf500b8b59970465cc2de2e75a
bf5d4300605adea5f4af79c94551a5c8c712b2ee
28542 F20101111_AABMEY voss_l_Page_009.QC.jpg
4f8bc45d2d735017ece217203bd997c3
fdd9f7b4be808ed0919ba8bd939389203782c3ba
39877 F20101111_AABLBW voss_l_Page_082.jpg
17fbff8d3cd6965c4a91722e11bb69a1
bd3f3b540bc1e14a8a650995529daebaa6f1f207
53995 F20101111_AABLZS voss_l_Page_125.pro
68a453c09bf261ee33f321f96b77dffc
fa138691bdcbde2145a8be0d21d243db620b475e
F20101111_AABLCK voss_l_Page_139.tif
30d340f690f21035da987fc87089652e
a1b2a9d2b6c0ad87f955ba0c2c14a7cbe1dabba7
17449 F20101111_AABMGB voss_l_Page_026.QC.jpg
0ddbcca702549fb675b152d131b3ac01
3f4f7bf50369a7f6e8ac93b5669cd93f39fc6277
23000 F20101111_AABMFN voss_l_Page_018.QC.jpg
9072a464f467adfb1b44db9065241197
2b56b5c028cab49fa62391f6db19fe9df8cf8bec
7488 F20101111_AABMEZ voss_l_Page_009thm.jpg
441e808b10aa3be8621c7426b249c33d
f1943126d605afad52ca4f37eb9ab57da42c7eaf
385365 F20101111_AABLBX voss_l_Page_086.jp2
e64902fe25cbcb4e80673ec000fe96d0
d352b0d9988939ca4cf847fe28e88a07bef951a8
F20101111_AABLCL voss_l_Page_104.QC.jpg
c452449bfaf41563dc5489f46d202362
38e666fcf51c1c80d3d83dcba30881d7ecd99fa5
5296 F20101111_AABMGC voss_l_Page_026thm.jpg
aba76fc7ef09bdb6155673147b59e5f5
754641b69412742e700f1e135b3ca1087a57e228
6303 F20101111_AABMFO voss_l_Page_018thm.jpg
f7a7b439caa5ff9bd2d85984e5518f92
adbd4b9e545a67517e60f41260e9257e6163a7ec
2136 F20101111_AABLBY voss_l_Page_108.txt
c566b47b7e2a78c07b3650c7744628cc
a65a3b068464445683708be1685c2d283174ed9e
45255 F20101111_AABLZT voss_l_Page_127.pro
f382293b627980b89a4542e9954782e7
6745b3d9009fe5b4bfb97e4e22b1fe6924a5dab5
57038 F20101111_AABLDA voss_l_Page_056.pro
79b3ad86d326a3284925087f338fff64
bee27d2ff8d4b29fdf33b992c705fff6f06cd46f
22866 F20101111_AABLCM voss_l_Page_029.QC.jpg
0283db5f030d4f76cee0bcad872bafcf
ba9f4fbf7a282d842998b57ffce431c7f6726d34
23433 F20101111_AABMGD voss_l_Page_027.QC.jpg
a9e89487117eb08c552be840e9559975
3521ef47fe36307163f766b1b442d99ffd8f3de2
25049 F20101111_AABMFP voss_l_Page_019.QC.jpg
c0157183af82835f5682322bb67f98ea
cdd161d20636b736788c74adb535ff5edb9c7bcb
38644 F20101111_AABLBZ voss_l_Page_064.jpg
7bc056ca20b7b1339d4020ed875d139c
97f6f05ab8ac5f3b9735fc300b3d89c93381c68c
3472 F20101111_AABLZU voss_l_Page_128.pro
59cfa6a6c2710376d002281bb3e2a6f1
024c78b59c3ed57187b9cf0db358e3935adb5ec2
881209 F20101111_AABLDB voss_l_Page_097.jp2
8db397a0e09b8452bb534b49f05c8022
d5869622d479ba39faef6e24abd523396e5cd3b7
411 F20101111_AABLCN voss_l_Page_068.txt
f3d95dd5b56a00c3a2ed95a038aa4e5d
81023c2af9e424cc5465c0ad1d1ef3c02101b372
6434 F20101111_AABMGE voss_l_Page_027thm.jpg
9a0830645f8e3bc7d56802d044bff909
82f16e084c8a5337a0773e61c57f96425f84cc48
7235 F20101111_AABMFQ voss_l_Page_019thm.jpg
d16a1502c6cdf46f4a4e37734099dba3
57060f8318e47decf9377d8f3e5552378ab44a90
7314 F20101111_AABLZV voss_l_Page_130.pro
450228026b8e03018058bf6b096c7df3
46288a17332ecdbde57e8a86aedbb55e3d78601e
9824 F20101111_AABLDC voss_l_Page_033.QC.jpg
bdbd957afcc194185dfd97bf6fa8aa00
6ba8f9798dc515ead08b833d209a31a72ae32695
665571 F20101111_AABLCO voss_l_Page_082.jp2
f1ff5e3d885fb6fedde1bfac10da8f49
4530b45314e27477a9d1b37c0685951763d9137f
7412 F20101111_AABMGF voss_l_Page_028thm.jpg
735eb267354eb0cfdc2b4bc323066097
5cb48e6a50ce3a094467751cbfa5987a4e45cdf1
27506 F20101111_AABMFR voss_l_Page_020.QC.jpg
4f09566e7f94b756416e15905f738d20
6967522f24aff6031106406f5f7fe1b564babc7e
48372 F20101111_AABLZW voss_l_Page_132.pro
e04df20d171e20ebf13e7c3b59cae20a
9f0024e479de06a01be074cabeec8d00b22e853b
3446 F20101111_AABLDD voss_l_Page_130thm.jpg
33d080ac2a00a542e13fb3da72bebaba
d98a47b52cb8d0e43a5d1a6773ee17cab25170c3
27039 F20101111_AABLCP voss_l_Page_038.jpg
5610b4d655871fc6001e2fa27ed701f7
5adb38dd87f1e09f534f049e10979ac90943beb3
22707 F20101111_AABMGG voss_l_Page_030.QC.jpg
65530ef1d712e0fed6fd12417d005dfe
6196d7e1fdb8fafcb0617ce0b364b9f01619f0b5
7491 F20101111_AABMFS voss_l_Page_020thm.jpg
0c3805000693470525322686184d7034
e7111abd747b8f79055266fb91594838366069d3
48346 F20101111_AABLZX voss_l_Page_133.pro
e312ec63e741e8435610cfd2c2368819
aa722154c05bc381b1c4712783bccc2cedf4cee0
82326 F20101111_AABLDE voss_l_Page_010.pro
a8acfb3d6bad5b18f0e5768bc22460c1
a3e9bb274728b0ad0e616fff1b6f1effa6adf6a5
7059 F20101111_AABLCQ voss_l_Page_046thm.jpg
627915822cd1824f15eba982c28cf1c1
17b08c8765a0e1aabf5c4041749a6ad2b4d68dda
6061 F20101111_AABMGH voss_l_Page_030thm.jpg
ae7804b20e927ad2a8663e0e840a4bc0
bdb14394f47a497afcfd284cfe3a21d43e77dd41
27932 F20101111_AABMFT voss_l_Page_021.QC.jpg
241ae683918c874b9abcdeba0db50c92
6275f40a59cd7cb9838e06192c47b38fc882ccd2
54066 F20101111_AABLZY voss_l_Page_134.pro
72f21089d6c2e9470b9664cb791c00d3
9bc359071f14e600ce8bce4dfad35912348c6151
6511 F20101111_AABLDF voss_l_Page_060.QC.jpg
2be9e018020a7974712f3ccbfe1a2688
751ec5c5f294e5df6b70641d4ede599310c90fa7
2553 F20101111_AABLCR voss_l_Page_041.pro
622e1de2f5373bdcc72a9dd1717a43a5
e220a01e88b3bd6dfee6d09cc8a6fd34442daaf2
23746 F20101111_AABMGI voss_l_Page_031.QC.jpg
9c8fa617cca71224517339eeebcad303
614171b9d35af4095e8566be697a3fba582ee861
24467 F20101111_AABMFU voss_l_Page_022.QC.jpg
5fbf35f322568c04e1074c10760080a4
2e335bceeaf447a4cd9ade5a57eae5ef452f0f49
52341 F20101111_AABLZZ voss_l_Page_135.pro
edcb1c10ec77526e200dd65bd970a991
1842ae634379bf242a4b8ccef7f1ee916ef06b8a
7232 F20101111_AABLDG voss_l_Page_013thm.jpg
b1733b20568d8d2e80ee3187fe0d606d
cc649c0b55ce7ad5e1474cbd1bbfa3c7f5e1eb1a
31919 F20101111_AABLCS voss_l_Page_084.jp2
f0a629fb30d17251eaeb642ee2418608
4c010f9d6ca7f582c3c04d883e08d3ce0473d52b
10579 F20101111_AABMGJ voss_l_Page_032.QC.jpg
e1d8869cd39027c24324fae1d1e9deb9
804801612f47c975a4ebecd7075823268c3f6cbe
6820 F20101111_AABMFV voss_l_Page_022thm.jpg
46a3aecb1f597bf75d87e15a9bddcbe8
09787ade837b1fb4fb84dd9b71a1456329dfd621
1995 F20101111_AABLDH voss_l_Page_132.txt
d777772cea7a0d488442242eebc18f8d
fee0fa5dff9a6e88b302a6c37832fab40cd17c59
F20101111_AABLCT voss_l_Page_065.tif
e1b8c9c8fa777e1b1aa9d39920edfbd7
a201090b721555590c417c99cfd5f70cccc26452
3218 F20101111_AABMGK voss_l_Page_032thm.jpg
2b440724c85ad17057e4b305f100e81c
117d4e3fdf64a996f4c7b9f359a6cef8e1f7aded
24419 F20101111_AABMFW voss_l_Page_023.QC.jpg
81ddc6cb45a1d7c11f90ab3de6058866
815c663b5c7a32fd67bb73ca69be9fffd346b37b
13421 F20101111_AABLDI voss_l_Page_098.pro
6fe6ebbdb0b244264be6ad324c1fa569
706befe113927d6ecec0db375ba69e3400bf6548
F20101111_AABLCU voss_l_Page_042.tif
5d2ff59fdce8d6cb168ec949aa017efe
ba06afda5db121335146e6b0888b852e3bc7327a
2986 F20101111_AABMGL voss_l_Page_033thm.jpg
9e92bbffca2257cc808483cfaa12c9da
8768fc036a2148fa586b75560a4491d5bbd75bdb
7176 F20101111_AABMFX voss_l_Page_023thm.jpg
4fc5f43a64c1efaacf17efe1a528fb82
d16747dfc91f119cd6c2bab5454a5aaf4aee2470
7684 F20101111_AABLDJ voss_l_Page_021thm.jpg
3cf46128c147dbd542ac007f779d747b
69df91e3beb94ddfe5f25e040c2859001a465b22
268731 F20101111_AABLCV voss_l_Page_038.jp2
42c0a33e0e925ecab9acd892d1ff3856
2fa93c84bc2c8bf20cc77a3bd48f394535204ca6
F20101111_AABMHA voss_l_Page_044thm.jpg
442eb17d6c7d758365a541f6ae044192
5347a31a6db4e2e152b2a9c157744da14d7796b4
3244 F20101111_AABMGM voss_l_Page_034thm.jpg
fbf9327bf81aefa3cf4b32d3b199ccb5
cccdf780f858d1db9193828f6c05881b9d21d76a
7549 F20101111_AABMFY voss_l_Page_024thm.jpg
3f83eff6f50c67c364dd84bffc995781
69bb01f7b2c6a37724390723f2a8bac29cc07570
12010 F20101111_AABLDK voss_l_Page_115.QC.jpg
d869ffb7cef030322769c5048dafeac1
79434bee95507bf7a46af62cc929b951dd1c4c50
78406 F20101111_AABLCW voss_l_Page_133.jpg
dd83183a5153582c49a1553d161a978b
711854245b2fc2bb2a323ac7a4995afc3f071988
2748 F20101111_AABMHB voss_l_Page_045thm.jpg
3c0760cea9f66d3c8bff7d6182c5035b
d0b203ca4d297aefe5901eadb485359aeb3ffafe
7196 F20101111_AABMGN voss_l_Page_036.QC.jpg
1004e6ea1a5e10950ce00afe5ca5c86e
ad83a33d83c5d3c6e6806e2d84ff807eaeaa6072
24861 F20101111_AABMFZ voss_l_Page_025.QC.jpg
9b4b9eac8f26c50dd0bfbbf6f87abf46
e8c857244e99aa31c2d7b09cc6747c024f5d6e9f
414741 F20101111_AABLDL voss_l_Page_102.jp2
1960f7bd114aa12fb90df707088ce0c5
904c7c3eb3db236cb94cc11cb4e8fa425bdafa35
F20101111_AABLCX voss_l_Page_020.tif
32aa92ec75deb55e931ff3d2a6958d27
245667fd5e97791cebfd3e4a0a5a3eb763467593
28630 F20101111_AABMHC voss_l_Page_047.QC.jpg
c883a36b675ad72ada6bf59884e9dbc1
b030c9bce2088de60c55be9c92ee7d8a928b3ac9
2460 F20101111_AABMGO voss_l_Page_036thm.jpg
87365e9b7972900ca766ab5e3e0a3cf4
5b70ab3f7920d787c8ac52bd0f705ff7b0a7b0d0
3422 F20101111_AABLEA voss_l_Page_131thm.jpg
6166a128c9ca098bacc9ccc2b289f441
b72f5a1a67def8bae9bdefae34895f55f51e0270
930025 F20101111_AABLDM voss_l_Page_083.jp2
63ca7db7344dafcb230673e96bed59e3
0f60409149fbbddb65d8c5f65b4c9abb77f81a9a
2098 F20101111_AABLCY voss_l_Page_014.txt
c920a4bbc60734b287890c4c07d908fe
ad9025e3f8ed69ad9ad48162493157a6a98f35ed
27603 F20101111_AABMHD voss_l_Page_048.QC.jpg
45624cbc0e8f332a4910692544e2634c
f6f747dde0297eb4ae5bafab55c9097814512795
2038 F20101111_AABMGP voss_l_Page_037thm.jpg
89758151eec2274d0ac1b54ddd9c72e1
2e1bf3de8de14f5d7f97b2e20dff81ea50035290
9840 F20101111_AABLEB voss_l_Page_003.jpg
570a9921b90d74d91500d89aa2b22b74
1786beaa47a22a10414abf80bad2b179e485b9f1
9684 F20101111_AABLDN voss_l_Page_063.QC.jpg
a11ee2f6624c5ae9f958244bf02b8f7d
6b184bfa7856c2c95ab54f674161674a1fea218a
F20101111_AABLCZ voss_l_Page_055.tif
e8796b81fc1862526b1e8cfd730cce48
e090b9f8273545ab0c12ff7bb448f8060fa46c9a
7438 F20101111_AABMHE voss_l_Page_048thm.jpg
a80488d3e34c21b9719bc7ae025b3110
e43f5daa2ace56d61f26d1adf077524bcfdf0121
8817 F20101111_AABMGQ voss_l_Page_038.QC.jpg
aaa868e43d07ed9278dbfd9f95dc2fbf
0e420b5f6ad11cadc1ab94f0484834e290253c25
11173 F20101111_AABLDO voss_l_Page_084.QC.jpg
1e6ebfa40a8ff9a80f600c4cf5ddd780
224f0dd4daa5cbc21e16145dd0ada170d6cbff24
3872 F20101111_AABLEC voss_l_Page_080thm.jpg
94724ca1cd621f7e1cf3e2e353501f19
c600837c9d10d92ace9ce8455c61503488ea5873
28044 F20101111_AABMHF voss_l_Page_049.QC.jpg
c5489dfbaf480f3ccbb485b36b8ffaa7
05faed965eb902fcf235225174891f472bac7217
2937 F20101111_AABMGR voss_l_Page_038thm.jpg
200e5d82edf8a7fc8fa6181c6058cca4
d71d2119adff4743f11c72cc283e42930b0fbc53
59738 F20101111_AABLDP voss_l_Page_138.pro
fc641fae946c62eebee4c3a693451d19
b8949e85fa38715203ae4d3041996e36b03cef0f
4146 F20101111_AABLED voss_l_Page_097thm.jpg
28479592aa1ed4514893e65289385171
5c5813261b5013549665ace1afd88403d532c2e4
7252 F20101111_AABMHG voss_l_Page_049thm.jpg
25ff6f8096fabfae2396cf12e77d026e
b62823feef0d79c2dfad399fb10aaed46c679589
10866 F20101111_AABMGS voss_l_Page_039.QC.jpg
c95fca8d9b49dd7082048b2849ffc6df
cd73facaa62cd9998a435d747e7f6f9b9fe79f56
85158 F20101111_AABLDQ voss_l_Page_137.jpg
ddb9833c82661a744636c39835d1f544
01512a41aee62e0fe77560ed039681794a4868c5
54418 F20101111_AABLEE voss_l_Page_013.pro
f207b3c1a15ef792a75064991554fe78
9895e637cda9d828d59b564c8f2eacd1c7714d76
26937 F20101111_AABMHH voss_l_Page_051.QC.jpg
07cdd79c3ea50231c03a8bd97c2393f5
d5043bd5a0a055f5f20eaa5c42b1427e3d98dcbe
3781 F20101111_AABMGT voss_l_Page_039thm.jpg
7bb1649f197fcc119ed6a4e870540b83
f6ffb28d743b8b070e722546bccd114ed4c2f967
F20101111_AABLDR voss_l_Page_142.tif
6b5495dc8650565b419b204368bd1c7e
5a1accec9c1cc934e6cadce17f59b5e37b4080b3
15763 F20101111_AABLEF voss_l_Page_112.jpg
b2ecf75e4f72534604ed0b5aef4df2bd
8e24f81987a4486bfb02a41c20208c57c07ad1e1
7372 F20101111_AABMHI voss_l_Page_051thm.jpg
0e2b0f75e2f9e6e866162565a2360314
4961ffd58428162783d12bd5ef555d71a377bd65
5706 F20101111_AABMGU voss_l_Page_040.QC.jpg
b592e81cad0c8a08476bd3eeaf5036c4
3b8c745ddd51ec558d6f2e78cda65f255e30aaa9
52935 F20101111_AABLDS voss_l_Page_104.pro
e5b3d68d442a686183d713a79f1a033c
32cc7d48f31460eb2f722aec24ead94b9d401305
692 F20101111_AABLEG voss_l_Page_118.txt
305ea71827f3bde0d367906a713860de
c91bc9c9b91ffab6905b3493a3787d0e4ec70e0d
25900 F20101111_AABMHJ voss_l_Page_052.QC.jpg
838f3c6e7f84232908eef513a8ef3100
54c7e700ca1e335dcedb6c31efc3ee7c366e5635
2005 F20101111_AABMGV voss_l_Page_040thm.jpg
d28b0db5a63f44feba2e398f06adc9a3
41d4b23ee9fbbbb5b9b25f641f6470cfbdae815e
5524 F20101111_AABLDT voss_l_Page_136thm.jpg
187faa4d053e22277d0b73d05e7caea6
6135a7be0c855a0d1c48581950e79994802296b1
4479 F20101111_AABLEH voss_l_Page_070thm.jpg
be43fdf766d52757d1bb56926ebe7581
e40f991a61559b782e0380bfa940613516b6a6be
7090 F20101111_AABMHK voss_l_Page_052thm.jpg
cd119172501ecf70b489100ed4a1ead2
3f5718b6db98c1736e556b5527cfb7d1e817834d
8686 F20101111_AABMGW voss_l_Page_042.QC.jpg
3a945444d797b72de52e253c206cd521
a6f17fbe3bd178b0f9b870fb4ec335eccac3f206
5345 F20101111_AABLDU voss_l_Page_112.QC.jpg
74ce1759724e7416c7c481c84e3a73fd
2b3b34ebf8d961adbd5f276fe6f54175c51e53f3
85671 F20101111_AABLEI voss_l_Page_092.jpg
b6547bb73e956ae6d6b3024137cc3884
45ba915c91b29e1d8cde4676eacbc8b04889f15c
2918 F20101111_AABMIA voss_l_Page_063thm.jpg
c5ea76acbe4d389171cabe0ce9e30c19
2415c4a724b1928692ba8d83540c11583e036588
7240 F20101111_AABMHL voss_l_Page_053thm.jpg
89f158ce3d9a3fef92e2af92f637ff10
55cc2275dcd455ed8dc77e31b862186f23fae553
2894 F20101111_AABMGX voss_l_Page_042thm.jpg
ddb02cc25704e4e01e000382e56e6d99
17f263c5b09e4989c7c6cc6622f47374e9dc04cd
1042233 F20101111_AABKZD voss_l_Page_141.jp2
9880fd5df0be5d45bcdf8b164ac23212
2ed04245ce7e5ffaf77e583d89bbeb474d78554d
45712 F20101111_AABLDV voss_l_Page_079.jpg
edd3aa27af26c46dbe87682b641ca7f1
d59e7d8ac06a17b6495595f7a5a8c09860d6b9c9
25941 F20101111_AABLEJ voss_l_Page_135.QC.jpg
65bb97a3b8c10a41681c994db5584e15
732d6168b9a374da0dc7df791ee269a29e9b6a6e
28528 F20101111_AABMHM voss_l_Page_054.QC.jpg
0d8b75d0906aeafdfb4397e72f69c490
01bdf79de5ed76999170c4193ebfd9fefa6f1684
6089 F20101111_AABMGY voss_l_Page_043.QC.jpg
31b5e163151164a7bca6af2a6c89a555
763e08231f33128b0fa12cc07ea42a9333673945
227014 F20101111_AABKZE voss_l_Page_111.jp2
4c72b5475202a1229736289559cffaf2
7a80814e93be14e7d438ff608db242126b973d20
767504 F20101111_AABLDW voss_l_Page_012.jp2
7a293424a0c9d9d0f4ede12e908a4897
6da49dcbbe27a5638519752b762df105a1ba21a8
409 F20101111_AABLEK voss_l_Page_038.txt
f9ff435fa74690cd8bccb4db62a64d71
b86de3f72c3730b0cece42ea0bae3c6adbb5f20d
12202 F20101111_AABMIB voss_l_Page_065.QC.jpg
e215b6da6a486aa9fc9ade0642ff1d41
acc15f1efc7bd4564717f61bf36712a735635e11
7729 F20101111_AABMHN voss_l_Page_054thm.jpg
a476a0d09dbd1ee184555c61d4557e5f
267b313dce42882997c8808e8af809274c826ce3
F20101111_AABMGZ voss_l_Page_044.QC.jpg
a9b6715cb837937d883ee7f617181317
1cdc304fbbdadab45e60b11279c535b3aad11d90
F20101111_AABKZF voss_l_Page_058.tif
812c9268d4746e307993377067d5d8f2
2cd2b0b132ec85307dae70ba22d6783459a31447
F20101111_AABLDX voss_l_Page_073.tif
a9bf47a681988aff515b56c4839581ae
137d3c0cf5a302767a00d58ac5a61636099a2997
2059 F20101111_AABLEL voss_l_Page_091.txt
ca4e6dba64a42fdb588d95a50e23200c
642ef57acfe8811d6be23bae5cc7baeb4103c342
4116 F20101111_AABMIC voss_l_Page_066thm.jpg
b1477d3051ff57157b3f13cf346bb9dc
1b8751cb403818445bcc6bf13218002f3cfe84df
28760 F20101111_AABMHO voss_l_Page_055.QC.jpg
e7578b3d61d76f3073fcc552a8bf8c55
0ed206a87046fd88b4cb60216bf7c1856cb95aca
3866 F20101111_AABKZG voss_l_Page_064thm.jpg
1ea92a8d6f69f2e0a47822a33faca915
f676dbd52095790ec8ec8d39a2dd7851c4ef85f4
12376 F20101111_AABLDY voss_l_Page_064.QC.jpg
506007c67259068457c03a1de407828b
494b35eb1e5d0d1c2940e304ef0f84eee5ae4dfd
10109 F20101111_AABLFA voss_l_Page_002.jpg
28b3a9028add4cf6bd9f09ecdbc080aa
22db941a3df0797be18b7b531250572340d1d0f8
2160 F20101111_AABLEM voss_l_Page_015.txt
cdb10ed5bda935ab272f256f18cd8688
877948c1cc92e8ff90dfdea3868355322b832768
12060 F20101111_AABMID voss_l_Page_067.QC.jpg
9b7d6b7a1d3a5ea2b269d448890876a2
381cf9518dd55c2f4d544ec237c35d4b3afeba44
7868 F20101111_AABMHP voss_l_Page_055thm.jpg
0dbb9f8136ee9cda2e324961244fb433
e8bd9abb4ef53f015ae684550e52f7c1c95aebd5
13093 F20101111_AABKZH voss_l_Page_119.pro
96f83f65050afc7df4562f91fc2b0bee
4e2a3a8d6a42321b1d4cc8772339a83aff9595a8
F20101111_AABLDZ voss_l_Page_045.QC.jpg
44b3d4e1fa6fff81dac2174b42c0bb51
7cdb79f48de245c83315fbb153c5833e01c211f9
F20101111_AABLFB voss_l_Page_048.tif
63801e1b3ccb332e17c34764bde7c770
dd77e43082edbed3ab368fa4264938e75ddf98f7
4035 F20101111_AABLEN voss_l_Page_144thm.jpg
507bc9c6b667bc3194b80f69208d768c
af68c0defd16b6c0c4bd9a6a477b4d3617d22ac7
13358 F20101111_AABMIE voss_l_Page_068.QC.jpg
261935c7dfdc34e6ade6f34251ecdea6
94a92ffe043c73b08d914f90f6ffa377aa6f41e3
28909 F20101111_AABMHQ voss_l_Page_056.QC.jpg
d1fa62db9fcc854212193523dd4a8933
8d6192dc091c403f94c794ffcba0d81ade88b61a
10760 F20101111_AABKZI voss_l_Page_086.QC.jpg
6c39106dedb057f50c8e6d4df48d0b11
a1c63e783554897c80e4a4a2bc8c8187b5c7eeff
1051925 F20101111_AABLFC voss_l_Page_135.jp2
6bb92a86336b2a3b21ad3bde964b683c
ac98f8767daa738a5d0c09eff012ed0b3e579971
88732 F20101111_AABLEO voss_l_Page_021.jpg
278eb988779c2517257f0b2681a63b63
705838071afa4bcf46b1c3075017b94899e9bf07
4600 F20101111_AABMIF voss_l_Page_068thm.jpg
d156e34bcbe3951c2cb078a959d8105a
d871231283baf75e7d579c960496d2cf6c79fbe4
7590 F20101111_AABMHR voss_l_Page_056thm.jpg
d6bc83decf86449e28d39a6937062d29
d11072fc6bb30f9f54472b53786a90b0e89cdb20
624698 F20101111_AABKZJ voss_l_Page_071.jp2
e2a92fe37a87b46517547fdda4966789
50b1e4cb4cc9f1768b7eec4252da429017a7c11f
26419 F20101111_AABLFD voss_l_Page_008.QC.jpg
53b2620fb07210fe9313ec6fe5d525e1
c95d5cb4d5f489579cad50a4b70b9f57d1f4571a
357165 F20101111_AABLEP voss_l_Page_095.jp2
c3eeabfe135b98fc1501e70ea9531885
4c2f3bdb1fa87c69b59c6ce3cbe94be4514ef3cb
11352 F20101111_AABMIG voss_l_Page_069.QC.jpg
8fa90a9e0e45e3e40c60261252e2281a
7924f122e901f290496badc43dd86555fa461f9e
26462 F20101111_AABMHS voss_l_Page_057.QC.jpg
206937c1372b073aecd27477b7d384aa
d6f8313d57daf413eed2ceb99d1adbe732f8de4b
21560 F20101111_AABKZK voss_l_Page_005.QC.jpg
78cac8e2f7c088e67ccd1b26b82bf13a
f10cd006fabe443faeeb3cc674a3ac1ff0cbde0a
717206 F20101111_AABLFE voss_l_Page_065.jp2
862edd4740c081f4c48551a0bab63dc9
f9ca4139aa384b7a659a6d5a732d21f1171a5348
349 F20101111_AABLEQ voss_l_Page_114.txt
cde2914081145ce89f732530d66e2b03
f2ad207199497df855bf179c68facc66eeac402c
3715 F20101111_AABMIH voss_l_Page_069thm.jpg
fb3259ea562797c695a3d12b6cce3b19
cf813d2e0e04594123b4d6a3940dcec08bc93243
7095 F20101111_AABMHT voss_l_Page_057thm.jpg
ee945ebdd2f5a19871c6b33da1a18e0f
0e05756fe4dae5a6272f454bc1d1b82bb044be24
960 F20101111_AABKZL voss_l_Page_072.txt
b5689ac457fad959a6080dc4f18bc457
0d1cdc837ffdcb0a600520622efb68d7e485b2c7
11800 F20101111_AABLFF voss_l_Page_070.pro
625fb66d7e320f8963b2b22b71807e10
7b0ccff965d252fdee02f87d3e7035542935f824
6748 F20101111_AABLER voss_l_Page_133thm.jpg
245579a20b43014761b73ae26e491a47
b37e8f4d56b5c75e75e2982d2f8f2c94165cefdf
12797 F20101111_AABMII voss_l_Page_070.QC.jpg
6dc4e0fd0fc7cc23bc97a94a24d5f553
558a46f0ef27942dff29c3cc4b0185951845dc6e
F20101111_AABMHU voss_l_Page_058thm.jpg
1c0c4728a7a572d6b7572f801ca5d0a9
b62d53c79b9e3a07bd60475c92b11361a5552c16
5939 F20101111_AABKZM voss_l_Page_037.QC.jpg
0ffd631a6288ee3a8214deb438b1abd3
e1e1ce8fe0af042f436b189697c06360e79c4abd
26890 F20101111_AABLFG voss_l_Page_125.QC.jpg
fdbd59e6101eba75c9c15354a3c4639f
38289ab9e1d74a94bb49f0cd6c9465109f435766
11466 F20101111_AABLES voss_l_Page_129.QC.jpg
438ce066cdcd166b568c1d21761e33d2
5838e68253a26a04a49960165fac7b030dc5ee50
3321 F20101111_AABMIJ voss_l_Page_071thm.jpg
c4a0dc01a64e8669438d71370df78a65
9fad2f2f4e1cc85d672ef8052cc2f4d7f69dda36
23991 F20101111_AABMHV voss_l_Page_059.QC.jpg
8f17a33fdd5f95d198078e7b2cf134eb
5075aa0a74faa6d54fcb4dfb96553ed2bd8ac3c7
F20101111_AABKZN voss_l_Page_122.tif
8be1c189808b609627178f25c829ebd8
67017bf7ca22e27f4a8e096f5336ec3ec8af53a0
1051966 F20101111_AABLFH voss_l_Page_049.jp2
43578f518a085165bf6dcf2e92d0ce23
3f759de01748a2c40f0b7ad21d2e5799f6e667a5
129 F20101111_AABLET voss_l_Page_043.txt
b3efee097b4b6aac54f335a541f0b99b
98194fc731da6210e9344766418fa793fc296b3d
10641 F20101111_AABMIK voss_l_Page_072.QC.jpg
68d876642f8a8c2aa180252dc05f83e0
a7ae10d46628ca3037357e02af5cbdb6ddd6ec3c
6442 F20101111_AABMHW voss_l_Page_059thm.jpg
6b804ee1af8d6c0a8b8e396015deb49f
68e97d9358ecc192191c1e28b08aab1075be5c07
26250 F20101111_AABKZO voss_l_Page_046.QC.jpg
b7ff4100bada2e14811f9f8335bc3be1
b6b97df37426fc43a3c8f76e95a94a6134c576ea
16050 F20101111_AABLFI voss_l_Page_071.pro
1c799427b9a087326346d4be4fe4e162
7eff7f0a56c721ae895368ad15695c763f7fe724
11051 F20101111_AABLEU voss_l_Page_102.QC.jpg
c1da2fc7026a8fde442f84925814e9fc
30829ea68d41094bf2bdac00e5dd16f49c6d24b5
4252 F20101111_AABMJA voss_l_Page_082thm.jpg
b509c445858b183b9fbb58e09f9aca1b
55df0765d763873016953c14faa43cc48be2657d
3747 F20101111_AABMIL voss_l_Page_072thm.jpg
d895cc5a147ddbcc6140561a588b23f0
223cc561cf73024219e82c15d17e81fb8eeeb743
2193 F20101111_AABMHX voss_l_Page_060thm.jpg
32fd3aca4aee184da7379aa5584780a1
7f9b02d7dca59eabff86dbf3f71d8aa81ed64b82
1958 F20101111_AABKZP voss_l_Page_141.txt
882831e46198b0e5f1ef06eaeb8ea8ce
0ed31d95c183c302d19bf665e08a6bfeab8918cb
F20101111_AABLFJ voss_l_Page_129.tif
d305547b1e85a448601a74795d34788e
afabff27c3a46cb124ee504b2e40adc16c3a5fb3
F20101111_AABLEV voss_l_Page_101.tif
2e02dfacbd54278af2f505c89a32819e
456e1b17ca54e7820dc168f3b307210291fa4b56
12783 F20101111_AABMJB voss_l_Page_083.QC.jpg
107437ed6b6cd6ff712f658d637eb103
c4338c37595c2384d096225cfe1d82797ed6c7f8
3451 F20101111_AABMIM voss_l_Page_073thm.jpg
dd9123d2890c624e955f06dcf142ad59
9d04d4f1d3ac7b3b1744ad96a5619c0ab600b854
8287 F20101111_AABMHY voss_l_Page_061.QC.jpg
296804a32ccedb2902b8fb75805b3a5e
7e2991d6281c4d54729a39cd5c2b689d60b05cf8
9629 F20101111_AABKZQ voss_l_Page_095.QC.jpg
bbe87c9cac094dc9aaca6def1094db98
1fa2cb3f6770dcc4b7a1e03ee9885caa476fb2b9
715401 F20101111_AABLFK voss_l_Page_128.jp2
cc17a962e848814096536839e44c77dc
35ee4f798c6b85ffc06367b001b6f8d38dccf01e
908 F20101111_AABLEW voss_l_Page_099.txt
4af3bec9a2162a457b8d2673a2e3de3d
6c877d23b85cb6ab61539b9500fd5f21b6f7dad7
9296 F20101111_AABMIN voss_l_Page_074.QC.jpg
d2f1b22216b52465060601094ccc6bf6
97c76c7d3d3b574b603323b73067d94238909620
2629 F20101111_AABMHZ voss_l_Page_061thm.jpg
a900c8418649003c5f8ee1b8cfe66ec3
5b7e47eba039b912933d41b84c2d29d322782242
77412 F20101111_AABKZR voss_l_Page_132.jpg
9b8e376d0d5b5a8efa26df8e8372b20b
89d6346588706e0caec1622d99b93a1a6be7514a
28343 F20101111_AABLFL voss_l_Page_024.QC.jpg
ba829f176606ddde4de35049f0dac8d5
f4ca1bdea01cdd451875356907c5108ebc07f279
F20101111_AABLEX voss_l_Page_126.tif
32b27372a72a1445023294c544dfc1bf
f0d171a8ddfe2b9b9aed7458f40c6360f99c9dce
4030 F20101111_AABMJC voss_l_Page_083thm.jpg
193a27a06f8bcb43dd51a4b759119070
d1308781fd9bae3cb0d09aace370e2cc1cc26308
3005 F20101111_AABMIO voss_l_Page_074thm.jpg
468499cfe444cfbeed953da9150e7cd5
a5d3ebef89ccb8e100e5e0ef97259be036249620
21490 F20101111_AABKZS voss_l_Page_042.jp2
9894706f469322a1fd607a9815ca0fe6
ac4f0b645a14a400578915e528ea46ae05f1c9fb
2154 F20101111_AABLGA voss_l_Page_123.txt
fe04c9b26f259a17b310b82e5d207977
cccabeefef919bdc9660c38d4ad24c191213d89b
983400 F20101111_AABLFM voss_l_Page_127.jp2
da1e7aed2ccfeaefe82311fa4882131f
66633b063c00b4fb7237b07efee9971104ecdf4a
F20101111_AABLEY voss_l_Page_095.tif
ac429d3e6035f2f33120d823136aed99
5625f2d03cb0a28a654a569d3812f34e36db46db
3671 F20101111_AABMJD voss_l_Page_086thm.jpg
b0c172ab2803020c5a23386707aae84f
5775413a87d68c5370cc4ba552403e2a293a92fc
3125 F20101111_AABMIP voss_l_Page_075thm.jpg
1b948da5ca6b4d350fb0576c78f0f6b0
64efc8f736281bb61f7500991e1b773a719f7243
7404 F20101111_AABKZT voss_l_Page_107thm.jpg
e6fb671fdca779367300590a5a25aa69
0d05745afe7e9555f229899d1e45f64afbdb10ab
2055 F20101111_AABLGB voss_l_Page_135.txt
922db8a399d0df47ccb506ae67beac0b
86e33d7726d127eeba28a270b3fe4132a5a9b8d1
26607 F20101111_AABLFN voss_l_Page_058.QC.jpg
87c09e1cb940123703c7d77690eba8a0
3697ce8fcce4b35fd4e4b190d158360c4ab8edc7
13626 F20101111_AABLEZ voss_l_Page_082.pro
eb1976483f298a71c8f19154316083be
ab1864509645369e22d2fec8a7cbd8e28bca97ae
11026 F20101111_AABMJE voss_l_Page_087.QC.jpg
bbacab54ad009c4f0a3ea44a9e75f9b8
55f5dbb34feacc5037b8816f18b0ae12476aaf15
11076 F20101111_AABMIQ voss_l_Page_076.QC.jpg
865a3b7518068694918aafabd47a39ff
3f0013c5ee72ae3f3b92fbb8fb2c0287af3b2020
972 F20101111_AABLGC voss_l_Page_074.txt
cc6ddbc2563b6e80ae00755997eeecf3
f8209f9381ed231852877cdcc4a568d6fe51cfd4
F20101111_AABLFO voss_l_Page_136.tif
a1ce561db3d6d070ca948e501ba9d759
613e03c5e268fb2ac1ae1f4398a3cdd7ec7dcbea
3492 F20101111_AABMJF voss_l_Page_087thm.jpg
73fe7eb97f004758200666864b0c666c
5659806df1b6921651f1cdb23f9d007c287f70f7
3816 F20101111_AABMIR voss_l_Page_076thm.jpg
d8bb799c4be9172ed43ac15b819468de
5b1219d2e6b234c21a791e30825cb06fee52ab22
14025 F20101111_AABKZU voss_l_Page_144.QC.jpg
51d036848dddf32b91e3a401c13302a6
b242a59eb6f4cd769f87872b34dba8f70640fb45
22819 F20101111_AABLGD voss_l_Page_036.jpg
2e185a844a54181f77830a1310ed4d98
768016ced4aa8b64ccd0451768115434796e92a2
F20101111_AABLFP voss_l_Page_015.tif
03e0eaa7caeb90e0b37c32bffe7e70af
1632c23df7ae77b78b5a0a79be4be98a3911dd3d
23271 F20101111_AABMJG voss_l_Page_089.QC.jpg
696df838b8779fb91816499bb9ba7f80
b89af828a451670b338040d2046e3446352adc64
17268 F20101111_AABMIS voss_l_Page_077.QC.jpg
069aae9153939c427d89cb8576e1be5f
9f7a6e67fd1d5046262de5e1cae18f552e7c775d
F20101111_AABKZV voss_l_Page_132.tif
4b1d6c65a17c539f9828657211b6036a
56a1a5cb3e942ddaf46f42fde86c145d28021bf0
49969 F20101111_AABLGE voss_l_Page_090.pro
ba2252d80a41d7bd72cde7405f72a24e
17a28958bc45a9972e3be46cb5f9dfbed0d66b23
81681 F20101111_AABLFQ voss_l_Page_019.jpg
37d55de42e38054518a3f6735d7b1cd2
e0ee82c56bc4ccb02f1b99502e311cc8abb9ffdf
6345 F20101111_AABMJH voss_l_Page_089thm.jpg
b172d4011b1c1ff5f050c2b2eb7580c9
2f38e643ff66d6b02439e0dd9d932865d8b58a4d
5067 F20101111_AABMIT voss_l_Page_077thm.jpg
a418a12c1af1ecd9353c302608145afb
50def9562ad1f323408b81612bc4f092a9bdb966
50564 F20101111_AABKZW voss_l_Page_106.pro
ae8da21041b4e28c2b425199012c1870
443e22e31a01beba9e62e0f7ea17dff85bf12e1d
35963 F20101111_AABLGF voss_l_Page_070.jpg
51a5dee97d8cf1f5a2c94132b790e0a2
6c7324f407bb29c09bcb7ff53c77d3b5be6f4d4f
13397 F20101111_AABLFR voss_l_Page_066.QC.jpg
abcc809699956a08196afd34b1b625b5
70e117d9a0a958d0fb967817c90da57ae3155889
25762 F20101111_AABMJI voss_l_Page_090.QC.jpg
c7cb1a20999ebfbddcd8d78d3792a0d5
7f5e8a449ce2e46ed24ca3e4f6615302f9871d66
9173 F20101111_AABMIU voss_l_Page_078.QC.jpg
2ea5b84efed328ae844555cf27f221b8
9dcd6a18f73fead2eda5b3edcbb42ba6d54e9baf
9029 F20101111_AABKZX voss_l_Page_118.QC.jpg
d188873f41e7a27833ba1d581248743e
dd9dddf5b494605b5386816e1ae7d7f5b4128e6c
443453 F20101111_AABLGG voss_l_Page_064.jp2
eb283e78e10dc42c5faaea814c4d43ec
648eef750ddb36860f4a36d8fb865ff9def73236
11359 F20101111_AABLFS voss_l_Page_071.QC.jpg
bde884b220466037620edec7588cffb0
c2fde868f7db85bf4871c2e1fc7d7e207770a172
7097 F20101111_AABMJJ voss_l_Page_090thm.jpg
d3aacfe73274644d4899699f95130aac
4e35c42b3676914f4e217e8331d2c519e082341b
3121 F20101111_AABMIV voss_l_Page_078thm.jpg
91a12a5d54cc33498c5e5c3a3f4b66dd
3b89cdc9a1603107a57bd38a0efb8344e4185cc5
11443 F20101111_AABKZY voss_l_Page_129.pro
5ec2b47157ebfab96913d1b4772b034a
d95b132c5dd929e8b34e368f0229adfa094617b7
F20101111_AABLGH voss_l_Page_044.tif
0cdd3e972e2386077bdb34b92c7ba356
49624aeab33f0ee7a9a8eb993feee343f19dbe92
F20101111_AABLFT voss_l_Page_047.tif
9f58875bb8f27dbd64adbe86d52b3b38
f68a416a5bed10b9cde3a84266e76cac856a88dc
25819 F20101111_AABMJK voss_l_Page_091.QC.jpg
1cd5ec32f8af7995baa7bdb074418fcd
e30eda0028415f119a894fe77650e659cf47ec82
14726 F20101111_AABMIW voss_l_Page_079.QC.jpg
fe1b493688ca1d1e305fe8ff5bc2de10
1f7aef392e0ef004747d3262cb64e93c8cbd0738
7791 F20101111_AABKZZ voss_l_Page_037.pro
d682705d50d57355bc978945023d70ca
e4e3b89eb46628c98cc53f674237ef7366263bba
18218 F20101111_AABLGI voss_l_Page_041.jpg
34b504e829d6899fc0c93a2e4fbf4cf8
05b48e238f51c18a565027a4f2b4738998b75d9a
1851 F20101111_AABLFU voss_l_Page_018.txt
b86363fe20ed0bacc5aba9dd7a6e30c1
751d8cf3378256a3ea4890c68e2ab7f41a61dab0
3185 F20101111_AABMKA voss_l_Page_101thm.jpg
3cac8e6aed4f208b7068dfb2af467ec8
1dbb77eedebf55f21641b721a9836cbf7c8f1924
7219 F20101111_AABMJL voss_l_Page_091thm.jpg
b731b15fbfd5405412d1b17ea514e050
884ba6374849d2bdaa9cf8ed5ae16bce8d0d9f51
11311 F20101111_AABMIX voss_l_Page_081.QC.jpg
48f24f3eaf3d7869c1cf2ae23d8704c5
b408abc89ee8cac1525cd5b689903d7a88d3c92d
1051985 F20101111_AABLGJ voss_l_Page_007.jp2
f1aa808ab0a73deef70d17c54966fbac
59ea435d74564fa180d367b8122596e1b3e24245
F20101111_AABLFV voss_l_Page_039.tif
42730315dd5f32d6f0c63949478cb0d7
431e626828bbf3365988f0832643f160fc1f57c2
3411 F20101111_AABMKB voss_l_Page_102thm.jpg
e9599685a8ec56be4a16579a08504f38
0dddda21e31c31506842469a6d9272299b0b46ab
27433 F20101111_AABMJM voss_l_Page_092.QC.jpg
fc3f031b35e294fd04c766086ccbddc4
a73b45eb7f4cd067e1f601fe709d248d6ff96707
3710 F20101111_AABMIY voss_l_Page_081thm.jpg
ea75340a2b082d14c1e69fc9cb910050
abd79131a3b7b26026a4a2797242df9380d97569
918 F20101111_AABLGK voss_l_Page_075.txt
456be01a6575186735ae433b61fb82cb
70b4a0baf7dbc28d3f1f9d2c62bbda1ad17b407d
72883 F20101111_AABLFW voss_l_Page_027.jpg
2bec5ea1053911fa8a7919533ecc1109
44d57c644b967e0f247bed3728bed461d6711d29
27241 F20101111_AABMKC voss_l_Page_103.QC.jpg
371867eb7f458903bda136c23191bf7a
b76a0d6d2a27fc7372691c934a3df55ee11337ab
7403 F20101111_AABMJN voss_l_Page_092thm.jpg
6f85548fb240756d8bc93562bd2cda09
ec7af856409b1276fc1769e2e3d3cd7f2fa355ba
13112 F20101111_AABMIZ voss_l_Page_082.QC.jpg
676c1124b68d2112e4e617c5575f4968
6fdf8d7a8abd5f3b7b23e366b672dc77f68dd940
7053 F20101111_AABLGL voss_l_Page_139thm.jpg
27a5d5996fddb50f8d42dd22311a7a42
66d630a05d0ba25fda5deee220be2c591ae8da15
10570 F20101111_AABLFX voss_l_Page_085.QC.jpg
6d2ae5df07c92f95b1271607118a281b
d9a50dfe56a78ee7e34569c2c7cf698fc19d0e90
1051972 F20101111_AABLHA voss_l_Page_015.jp2
81949df295e9e8ebbd1dc9ca284adf94
e274c33a9a5e629245409cf0f76acbcfec677ab0
9043 F20101111_AABMJO voss_l_Page_093.QC.jpg
c058d4bb3f948350c7606793539b2b06
fa97afc587863bebf8b49218f1c527ab55141563
F20101111_AABLGM voss_l_Page_024.tif
d687fb37003f285194b28d82b23b0120
5cb7fe6e7fe4932617eaaf65ac92760bd419e54c
2099 F20101111_AABLFY voss_l_Page_041thm.jpg
ba7e07a79e3551b41c7f04b5b92d3343
c9f04f40dd92328c91c65706cb815c1e804757af
7284 F20101111_AABMKD voss_l_Page_103thm.jpg
755e4fdd0a2987ff36e25b5ea98f62d4
387209b51821e4114e490aa39ffacc181290acd0
2730 F20101111_AABMJP voss_l_Page_093thm.jpg
21e1c32cb3552086b83eebde14806bfb
10d85853f5d951b2211cd6970f4bf2b6fe024c73
4561 F20101111_AABLGN voss_l_Page_113thm.jpg
72cf86fa864cbc13e374f0cfb23b5be0
4a09393bddeb4ce25e4c62244c3c7ef049e65401
80031 F20101111_AABLFZ voss_l_Page_091.jpg
1d1ed4c3cfcfdd3e98464093efff26f9
d10503fa832e0e5e4ecd4173613d01b2920e44be
53259 F20101111_AABLHB voss_l_Page_014.pro
c33f8c63e5b08cebf6e84d0d75afe093
365653d0f526db6b22ecb65d4ed3a83942ba40e7
22985 F20101111_AABMKE voss_l_Page_105.QC.jpg
a3cbb50b9207b48e3d0d746fbfc457cd
dae1cbe6175bbfd381c20263ce7a97582ae058bb
5213 F20101111_AABMJQ voss_l_Page_094.QC.jpg
35c1ab32e5ac337c0554b98cd5adc3f5
2ba9df07c1b749f47b476aa99c52176724ef1d22
420 F20101111_AABLGO voss_l_Page_001.txt
ac7bedfe307939b13dcb231fc7847310
7e5e2c140b63413f4d4f026d3c678e4c1b32523a
9286 F20101111_AABLHC voss_l_Page_075.QC.jpg
b5849058050b5f8cb60653c508fbc9dc
a4877a16ceea48f67512347a6d0edc1c21ab5275
7424 F20101111_AABMKF voss_l_Page_106thm.jpg
62c3d55efbac52a453b63cdec8a9e85e
21827c12d9a3f65bf5da7b57d6250dc1a0880af2
1896 F20101111_AABMJR voss_l_Page_094thm.jpg
6601f7d0c24d2cb2aed5f2cbd18751de
8566924f7daf655176d657ebc8789c711ae0f0ac
504804 F20101111_AABLGP voss_l_Page_131.jp2
6268ffbaa9cc3596ee31ee1766ba5f2c
2aef99389fc28876fd733612dc57efd7a8f4a174
F20101111_AABLHD voss_l_Page_005.tif
721c66194b5afbe2a22d5c2601a8336c
134545ffc282a68dc6ca1746cb92893c34b09345
27331 F20101111_AABMKG voss_l_Page_107.QC.jpg
57ba0dbda29d7f9702861beb37f6dc18
252fbdc57c34ca22ef638217643bb5c06672b6f4
19107 F20101111_AABMJS voss_l_Page_096.QC.jpg
b36cb5a2af90c39686e095433bf306fb
a34e28aa60fd826b56c14aeb0fd8d497f1dadc71
856 F20101111_AABLGQ voss_l_Page_081.txt
c2355b24305df6a750977cd91efb19f4
9ea6cd91276f70ad83a89cbc5d88ee9fda6e2d91
2113 F20101111_AABLHE voss_l_Page_104.txt
4777beb485b825e8ed5f79b043c99cc9
29e658d1f6b23000c75dee30f9fe3eb84234c2f2
26938 F20101111_AABMKH voss_l_Page_108.QC.jpg
0ac0f2af27459190c7dd6af8e9ccc0e3
330a92726bc7fb91c0203e1eec771805964bd505
5962 F20101111_AABMJT voss_l_Page_096thm.jpg
58b207af17faf4e43b9936585aef454f
c82cc5f1fa85040d135b297b8b5c01cb4134ab37
4151 F20101111_AABLGR voss_l_Page_099thm.jpg
2eec156d278c3c36c24e6b256908971e
39e7e21f06e29d48e5390fbc18ab41f582ed3834
48854 F20101111_AABLHF voss_l_Page_059.pro
65ad5fed218bfcc462271c21532c3e58
c38e573518afba5553bdcb60ea821aa26df7179a
7513 F20101111_AABMKI voss_l_Page_108thm.jpg
377d3c3c85a94655a38ec29f947127fd
fc9d3fc49015dbd1783f12a76a8e34156718f2be
13570 F20101111_AABMJU voss_l_Page_097.QC.jpg
ce7660c49021a40d4480025ff5e74617
a012f318ad0725089b8b8aee2d2ed3dafef3df94
27540 F20101111_AABLGS voss_l_Page_126.QC.jpg
057a8824627ae0776552bc690704caa8
ce54db93a3a512bb5e7a3e35c61530bb78d9634f
428 F20101111_AABLHG voss_l_Page_062.txt
8d9640586bf5625d8969c915285fdde9
e77e761a68624dc272bef864cc6e43f042ab745e
28158 F20101111_AABMKJ voss_l_Page_109.QC.jpg
8d8dd421e4f0877ff5c2f828f042c8a2
c52c68d0ac7f61221246ee718bc5722b4f347457
14406 F20101111_AABMJV voss_l_Page_098.QC.jpg
761c97c68252b4afd5c6146ccf29045c
3bcc630cdc340f61a80e01ffcc7f1f1b15061c16
49316 F20101111_AABLGT voss_l_Page_141.pro
9febcc1440865ff971fdc62475ef063f
2585ab397bca91a7baef1f6b44b095fbf52f912e
78640 F20101111_AABLHH voss_l_Page_141.jpg
e8425d50e1c6b9a2008c286d5a76ef01
b26c5580fe8f41fa366852959a4323baa4c7bdae
7809 F20101111_AABMKK voss_l_Page_109thm.jpg
5ef6f229e937866e2a853c43b96b7736
8925d29afabf01acd7e249f8db9cd55848e89ad1
4122 F20101111_AABMJW voss_l_Page_098thm.jpg
3e9cd431d15c3e19e94dbde51d360671
88b725625c7a270fb8f1cb2b356618788fc87b2c
5905 F20101111_AABLGU voss_l_Page_131.pro
9ff55c8c1b400c511f239b35cb84824e
9e980308bc35a04c08c91a31e8d1f1ce8a03ec37
2064 F20101111_AABLHI voss_l_Page_062thm.jpg
c71510cf03d1827a9cd080c03198a176
fbf240c467ebb4887e9f75754a0750cea83bf947
2850 F20101111_AABMLA voss_l_Page_121thm.jpg
6cc6f0d38f48a0c81621b486fc3362e8
44852f949e7b3bcbdd6cdc3c862b30add36d6e32
7315 F20101111_AABMKL voss_l_Page_110thm.jpg
42cf5a8761f3133663153302eec60821
9cf394cea30404932490cbe69ec53fe61e9c05a1
14277 F20101111_AABMJX voss_l_Page_099.QC.jpg
5926ad45126251679743dd65caa63d00
e7add9ab25ec5df5585069d27af6599b084ca890
49055 F20101111_AABLGV voss_l_Page_022.pro
7664174ddf1ca5b832211276b4da6f4f
18932ff9d40d462862ceedcd3c34d0589fd37aab
12425 F20101111_AABLHJ voss_l_Page_088.QC.jpg
ee82b0cf28a3469717d1d40a17878558
e1174cde0b25cdf17fc9be43cd04a9cffdcbefd4
14501 F20101111_AABMLB voss_l_Page_122.QC.jpg
69f8c7205dd40d366c98de20976c5de4
1f6fc1d59e3f20e4bbb4772dc1c63e1ce6f1b939
2417 F20101111_AABMKM voss_l_Page_111thm.jpg
2fbc57dc35f1c694a0b406ccd263d535
116df48ee2bc33ba47b1b8582e9a76594cdf0ca7
11254 F20101111_AABMJY voss_l_Page_100.QC.jpg
f27ac0ee5fa92563420e32031918d2fc
bf4a1dd61560c321b00175ed03bc90f614ceffb1
3593 F20101111_AABLGW voss_l_Page_085thm.jpg
439d7885110e7b529d8e6b8c6d897135
7881f68faa84d83ca3909c0934ea4a2e6a71c1c7
1051986 F20101111_AABLHK voss_l_Page_052.jp2
2b67ebaa119e410b0d95eec28967bfdf
ac2d470c75b1529f5b98c029ce54beab51ec92ee
4183 F20101111_AABMLC voss_l_Page_122thm.jpg
8300ba79643a3234d6425109dec55c91
64f0de7323046ba371cf5482939720020bb30a02
1914 F20101111_AABMKN voss_l_Page_112thm.jpg
b63d503ea6c5ca87680633d31af0f2ea
da04fcab305a055f09d97248903dc3978f35d4e8
9975 F20101111_AABMJZ voss_l_Page_101.QC.jpg
a109b1814423832bed82465ecb8e873e
217f629a354dd6b8b981dc0331757924a3060a4f
7266 F20101111_AABLGX voss_l_Page_126thm.jpg
9f43697ab954e26e031a457d33e9db07
d6256ffe383c2bbad5dc1af07c127203559f16c8
14020 F20101111_AABLIA voss_l_Page_007.QC.jpg
769d977218f8abf878e95148a1dcc4d6
63dbf398cb90c32cb468cb6b2face557e5689f41
1976 F20101111_AABLHL voss_l_Page_127.txt
7cde2ca62f24327e1da709fd07ad5c32
023ecba70d3a3f81c627c7fd4d585bf6ba004d9c
7380 F20101111_AABMLD voss_l_Page_123thm.jpg
ec08522ad238ca3ccf70846a91e7e57c
a3da46eb75af5da31cdd3280b1047f4aa3966e98
13735 F20101111_AABMKO voss_l_Page_113.QC.jpg
5b6016dd4f8ea566483c7415f23b3153
6af48efa01153b483403f422337c5e21f44ed4d5
24361 F20101111_AABLGY voss_l_Page_139.QC.jpg
c3b1c4575fabf607958009287b488c87
8fc7d833a09413344fd8509f48d48287a955f432
F20101111_AABLIB voss_l_Page_045.tif
60ed5ef4d71f50b7b2c197b012a54d9d
131590f9d800578c1cff4e66ed6896c869b9ab87
10992 F20101111_AABLHM voss_l_Page_119.QC.jpg
bf02cd07ff75caa8ffc71bd50048ebec
7e57a3d1b40588a3371d51d8fcc1ef54ca8c21ce
3048 F20101111_AABMKP voss_l_Page_114thm.jpg
f648aa5f94074924702dc556ce75da69
b462ca9d48d6518ef96d9550aafe3fea86a4ac0f
F20101111_AABLGZ voss_l_Page_123.jp2
315078f21ea0715e128e98e16548bb97
d682c3f9b29e6654f7e9048850a9fcba182c65f6
3722 F20101111_AABLHN voss_l_Page_067thm.jpg
b5335c646a66f4ee9acb3f7839ea457f
60dc8b80d5bf9f5ea1877273403bac8df8dcf055
26430 F20101111_AABMLE voss_l_Page_124.QC.jpg
7807d41b9adc144a1abedd6d6c52a2b5
3783492a56141f9819760123c40ca31b2769a9aa
3701 F20101111_AABMKQ voss_l_Page_115thm.jpg
6e0768d6fadbcb17e38a537c4dd21dde
6db90ebf362a473261efb1d2914fc4ca2ee8ba81
10800 F20101111_AABLIC voss_l_Page_073.QC.jpg
b49184472cde69483d032b4a9d50ae79
75cd4bf25cd5be1ce9187216d8bb688b9b66c38d
7660 F20101111_AABLHO voss_l_Page_102.pro
83ed254b868191449b3562f031d30942
ae1c8ce5b082cc787fd699fd3003e95533b4e2b8
7278 F20101111_AABMLF voss_l_Page_125thm.jpg
38485f4077491e579f9eec8116ffe761
ec51451926c12288fd70043dbeef83d09037c14e
10225 F20101111_AABMKR voss_l_Page_116.QC.jpg
0900215320d94693ce1a0299d1d0a029
ceb1b5becc0781819c0fd49f4354c094a7554cb4
F20101111_AABLID voss_l_Page_036.tif
7c9e797a7ae47a199f56149b1c9dbf92
15d7789e9cea5873d7b0f36d9771644e12086f1c
55047 F20101111_AABLHP voss_l_Page_015.pro
797433b00fb07c48eaf4e98aa32c8e6c
136f2b5070baf08e320f2faa497f8138987ee02e
22385 F20101111_AABMLG voss_l_Page_127.QC.jpg
79be5b2c484d51f952ce12f99c769b03
a85d34d67a2f793b716bf2ae39a4f2bcefe614a7
3512 F20101111_AABMKS voss_l_Page_116thm.jpg
b03ce8e1fb52b7cecc093ff5a9cfad0d
7173779a49a7e3127be2305a604933543f3afd1b
34652 F20101111_AABLIE voss_l_Page_115.jpg
1e7b7da834a8c93318de71e36d3e112c
3ebf0260e7277def198fa97342e7d1185fa460fc
42021 F20101111_AABLHQ voss_l_Page_089.pro
a538e6098fcc6b55317f102cd5c35c54
bb5ca52209866466726292ac5a20c0207459392f
6437 F20101111_AABMLH voss_l_Page_127thm.jpg
50686a0341ed4d7e5ad592fb9435a4ac
7d0aa68ec789858c335d2e78f59f3563a2924dd8
8156 F20101111_AABMKT voss_l_Page_117.QC.jpg
590587ca83d1974cabe0cdabd3f92284
622d6861a84b67b70a770efc2a00aadb23e52faa
84102 F20101111_AABLIF voss_l_Page_108.jpg
709d482e74f92f9ec7d439462002a2a9
5f74620ede51aed0394ef53f77307a7331598c1e
F20101111_AABLHR voss_l_Page_010.tif
59af2ec5484e2a35c0f2852428332196
155b745a4ef93ff9b7e1a9616a4081c09663a192
7006 F20101111_AABMLI voss_l_Page_128.QC.jpg
b31dbda30bce8b6115cc50bab8a64c9c
7abf0736d4591b92ed974bec2ca0bfbcac913b79
3119 F20101111_AABMKU voss_l_Page_117thm.jpg
a1b380d75dab472419660dab847f8805
b3f65ff58ee5c020544483e832944c34744253d2
16536 F20101111_AABLIG voss_l_Page_074.pro
5c977997f0883444670fb0fb22e8dc08
c32c3cbb49b704aa319840be135a44fddf28a3b3
F20101111_AABLHS voss_l_Page_060.tif
f8b408bec518909db514c0e8b09f844b
b5dccfe874b82d654ff24c2f6926e80c1822cff0
2375 F20101111_AABMLJ voss_l_Page_128thm.jpg
a5053dda8e4b2aadce30ee5be4c28455
d85d9d817d16a730c1af5587c5d8d1efeed01db1
3154 F20101111_AABMKV voss_l_Page_118thm.jpg
b9682fcc39d344a62d84bb9c801cf76e
4fdd31513ab75fe9f828daa8809d030c5d8595f6
43891 F20101111_AABLIH voss_l_Page_018.pro
4064486bf56cbf9e6b5d572d83b3e6ac
c9ced5f9b97395500ede1e66369f216ebb04fab7
F20101111_AABLHT voss_l_Page_003.tif
d4469d79dcfe6cc59ba1c53cb3b35969
c621ab08119dcd6960c5d07f6c98f9fddb9f2041
3480 F20101111_AABMLK voss_l_Page_129thm.jpg
0936a9930e36ee9b84543fd7bcfe4813
34e79f4907fd36f85c977027ec4f223eaec4545b
3682 F20101111_AABMKW voss_l_Page_119thm.jpg
897bd4f4c9a831d727c277c191d3179b
123aa514145aae0fa985d561963a367e097452b6
12488 F20101111_AABLII voss_l_Page_061.pro
4d41bdfe131baf018aa76da6138eb55b
e908e6d6cd1328bbb50eb2eb02e9f3f5c5605c86
7246 F20101111_AABLHU voss_l_Page_104thm.jpg
261d85da0e4f89acc324db075cfbaca7
9d446dc97f2adc64a6b74897fb0504cedb70ba10
14518 F20101111_AABMMA voss_l_Page_143.QC.jpg
e8055b8550f05a5955c1ffc8be3384a1
cb80bd2930d1b50d01c59ef9d25ef55db163bbe3
9298 F20101111_AABMLL voss_l_Page_131.QC.jpg
79d45a9832fe719ffad1a2c6814a02ba
0ed89028f9e3655363ed56a62416d92faddf2e9f
8741 F20101111_AABMKX voss_l_Page_120.QC.jpg
b3949e759dbb27d310cbca7686367d67
b523dd5df0b762e1e5aacf7640076aded578556f
7140 F20101111_AABLIJ voss_l_Page_016.QC.jpg
712501517556a6a00ee4b74b0e5c5b62
5f9342b7c2c2a7d9e58a4a2b5dbee825f5fe8668
595 F20101111_AABLHV voss_l_Page_033.txt
50b4b5d019ea6c9372069bf31ed7c59c
bea30fb602e4e389cad724b9d35d514e387a5c45
4208 F20101111_AABMMB voss_l_Page_143thm.jpg
aa592552d8ae2d73eae9d13e883b8e60
14e4996499fe90bebb6fd2a6face20056f115cb3
23527 F20101111_AABMLM voss_l_Page_132.QC.jpg
4fdda215c48d10b2d6dee1d43265c30a
287279e5ce38d3649a1add3f5e43f3540c9d6273
2873 F20101111_AABMKY voss_l_Page_120thm.jpg
7812ee864c46cc4a413f44bcea63fcb0
80cf75be9437fbd357e04c8aeea9b7ecdcce73b0
F20101111_AABLIK voss_l_Page_053.tif
3b7ff1a544c7e9c3009b67230a0832ae
fda37f9278e32aaa6d0e612a9209e1d39733eda5
157181 F20101111_AABLHW voss_l_Page_062.jp2
50977a3f76db888e4db707a8aa6efd3a
b96518e02c82e328a929ede699602d1ea7529a37
6568 F20101111_AABMLN voss_l_Page_132thm.jpg
f2f758f8132597061dd3ed5ccdbeca6c
bf3633a5c2097a880ca7c4a2f0a7bf804776bcf9
8144 F20101111_AABMKZ voss_l_Page_121.QC.jpg
8e868b642f6e27eb5c9c28e66487b003
036a5537e993e812c35b20e8971852c296c9f36f
1051975 F20101111_AABLIL voss_l_Page_014.jp2
e8a46a441b916d6cb502353df7c0abde
db274b260854893d99da295faa64efef639fa533
F20101111_AABLHX voss_l_Page_133.jp2
a84b712b0be396d630e1033100318b1c
5aab30553adf691674fa877df35709902b475f5f
F20101111_AABLJA voss_l_Page_137.tif
784ef89939ec931a014b7e0a68a8c794
c7c1bf9f5558911ab82e78eb0d8103e12d0c9d27
24703 F20101111_AABMLO voss_l_Page_133.QC.jpg
8e9b25a37f4b2967d9af6030d10d595f
102c208abc08060ee6d47431163f1bf5d1dbc507
9354 F20101111_AABLIM voss_l_Page_080.QC.jpg
b5a0694857b8f04feb5b669252bdf3ca
29ba31da4092ba7006ecb73c7b5ba79617281133
87963 F20101111_AABLHY voss_l_Page_109.jpg
8d225e0197142c1d6b57c6aabb9b900d
b5b1412398116ea8bea0b86e435cb45b18b0c19c
F20101111_AABLJB voss_l_Page_054.txt
21dc8af86c7ca9f7d3a0fdb0c2614c62
581001c6ff47016f452349e8f54ac704e4b79785
27422 F20101111_AABMLP voss_l_Page_134.QC.jpg
08366df35be2eab6fcd2aace558ff7a9
41febf96cc848f7bc5b964c574f3769d6a6932f8
F20101111_AABLIN voss_l_Page_057.tif
f0dece4d4e629e78225d8158f26bfc9c
c68c455c2919b35a36761111f146cba341dd927b
4554 F20101111_AABLHZ voss_l_Page_079thm.jpg
21e01bbb5bdf178ce743d0c62fd82f54
70a482b7231f6e681e4d48382570e5e909af82bc
78546 F20101111_AABLJC voss_l_Page_025.jpg
b053934f4da5707ec61cb3ac82629837
79549ddfded2347c3a28059cb52bd5f77fd6df26
7133 F20101111_AABMLQ voss_l_Page_134thm.jpg
50c17e61e5cb7809eabbcd32e5677262
e010e91ac9a59703c5f1d96c803338b3c91108ad
45381 F20101111_AABLIO voss_l_Page_122.jpg
ea054c4fb697bae0482c8f1f5ffa7dd6
b251c54145c7e08ea3adf2cf8bdca85c80ffc2da
25635 F20101111_AABMLR voss_l_Page_137.QC.jpg
b704b756b9e23cffcf0390d9a1255f14
9bae396e26c7b5314e4c18adb4727adaf5ea3210
F20101111_AABLIP voss_l_Page_063.tif
476ea17ea3132a3ce089b265b4ab3f97
daa248599fa527e05d9c44d9cf7386ad37a51ea1
F20101111_AABLJD voss_l_Page_110.txt
6441d29d8bc2e207f52b18660ec17bba
5775c8302f43e4a03c4f7cdb75a7aa456e4dc30f
7261 F20101111_AABMLS voss_l_Page_137thm.jpg
79f1d0b9b0a25fe494ab78b5cc14ad44
0d15d64db8ad77416aaf7962397e465493315d63
F20101111_AABLIQ voss_l_Page_016.tif
6d157752b2fe249a1bbb7b84131681a3
e26ddbefbd925e1b8a9927968fafcf522ac51734
345666 F20101111_AABLJE voss_l_Page_033.jp2
1a610c6fbb7d4e29a8dddfe71d1bd76e
4a7be3a099503c4e11d7dcd1e0230ccd8ffe5ce2
27828 F20101111_AABMLT voss_l_Page_138.QC.jpg
e927aadcc87e0e2cdf9feaa1f6bc36d4
0a5bdde7b9df1454d8d9c9498cdc5f84b5f3400e
3518 F20101111_AABLIR voss_l_Page_100thm.jpg
ae0705f385cb876fa47515febf2fb124
f2ea4ed7438276263d637d29c6a9d616531bfeb3
10928 F20101111_AABLJF voss_l_Page_130.QC.jpg
3edfbad8752d179a62eab37a18487366
9576365ac2ab4413bb47d7a61d77edceb360918f
7535 F20101111_AABMLU voss_l_Page_138thm.jpg
cb4d48f4dbd34c4684696575a095cbc0
7fe79974d7d2a64012ffade673afb83cb7ff0697
53488 F20101111_AABLIS voss_l_Page_058.pro
244286dbf344f8094f3fe4ca186a29d0
dcfbc6923c9a815c69d62df311ee875587b230d4
F20101111_AABLJG voss_l_Page_002.tif
fd29c35de082c37272e6e845a2eaddb6
d9ba098f90aea7884e266eb4fa91a5c407e6cf88
25747 F20101111_AABMLV voss_l_Page_140.QC.jpg
51fb0a24d0256721436d5b4d041c62eb
ddfa35517e13f1a7997a61b5df6856d2ab7fd1ec
356899 F20101111_AABLIT voss_l_Page_063.jp2
30b8beb68a8cdfae1abdb43f0fad3ede
0dbb6467cffa42ca94929d32b6ebe4adb32aab5b
225889 F20101111_AABLJH voss_l_Page_001.jp2
e60bf92ef2b22691e8ab17bfcf7e59a6
251c36897209b568130ec30681605e60e841a419
7111 F20101111_AABMLW voss_l_Page_140thm.jpg
fda31e74c1b975f99d49bcc22ac831ff
e8de1f037eae5c629c18cffabbe13316270e1fca
2968 F20101111_AABLIU voss_l_Page_045.pro
55f5a3dd6f03306ea3dad74102106dc3
35217dc47ae883118c38008320e499e74ef2b236
78414 F20101111_AABLJI voss_l_Page_139.jpg
2bcdf2d69e1cc451b78d7548cbd213d2
dd28a227ff6010a68325c4e3492d96bddeed63b5
24025 F20101111_AABMLX voss_l_Page_141.QC.jpg
8fede711b4389a6aa8e4814895c93a0b
53c7467c31837847b82e88efc8669a7b840d1617
86386 F20101111_AABLIV voss_l_Page_142.jpg
25b77f69d20b9493b95b09a20f7deff0
4f04d8a7be5ccfc52dc8ce4dd452a57e6ea0beef
5421 F20101111_AABLJJ voss_l_Page_041.QC.jpg
fba28c274393363d96784f2ccbdf0943
b7a07690065f9f49493a67b3aac494b07542dc38
6665 F20101111_AABMLY voss_l_Page_141thm.jpg
6b3db8114cddf97d25a9299b2fe940b0
6aabf5ae7dc596ed1a580a3c92e8bfb2e3ea23da
3487 F20101111_AABLIW voss_l_Page_084thm.jpg
80acf60e876b1524fec2869f55de17f8
45a9db6d05f23344437926df1fb754d16f35c339
3675 F20101111_AABLJK voss_l_Page_065thm.jpg
9065205fdd9f0d4981351120021d321b
ef39e1987d1c636cb621173e245de1931c39dd19
7109 F20101111_AABMLZ voss_l_Page_142thm.jpg
672dfdd4086c54131600e8280d0a99e2
8f743c46c27870a10b896cdd09035174c846edc6
223053 F20101111_AABLIX voss_l_Page_016.jp2
4544eae72d12dc96ba646f31ccb2e91c
377cad5d5210520a887bb4983eab68ca40e68610
1731 F20101111_AABLKA voss_l_Page_029.txt
38aa96ee662fc7f47ebe0add23ab0387
4adc317a897c7808d6eca2c93ad27357c13539ea
44260 F20101111_AABLJL voss_l_Page_144.jpg
7908033180bbcd0fbf3dbcd263b83195
acc591b599bd2c889c100fb2f829605b33bd503b
2024 F20101111_AABLIY voss_l_Page_050.txt
55f60bdda3f4e5f5bdcdc5414042c577
af7e0e19415caeacffca378d6f0f223d893ac7a8
82671 F20101111_AABLKB voss_l_Page_052.jpg
3098b4aeedd4fc21eae934fa8c96beea
0a9408c80eb61ec3795a4d35504308f7c002aeab
57559 F20101111_AABLJM voss_l_Page_055.pro
47a760e8ea7347dd00e404939a87804e
f0a4718e4a6b5fa7c4c0fbc887a391f02905d6ae
19996 F20101111_AABLIZ voss_l_Page_136.QC.jpg
bd07cad08be8d20787c3bf4352f19c9f
5e68f9be6ec82ad7384126dcf37f8a79dcc77035
33322 F20101111_AABLKC voss_l_Page_033.jpg
3eee1de37d140277d215f4a4cd2b50d6
d3fc54362b343cdba6e20a8c92a2db4edf8b83af
1051937 F20101111_AABLJN voss_l_Page_126.jp2
035bf6c2e2bb3c8090f5230247c92fbf
2af63d88761a746b09b4e229a5b8e74ddf923c4c
3906 F20101111_AABLKD voss_l_Page_035thm.jpg
1c38417b5c2c90641d87da0f9af82156
b0d04dfc72f59339a0f079b14b40b2c134fc79c3
1956 F20101111_AABLJO voss_l_Page_022.txt
8fbe3fee42455dccfedd1239bc6bf6b7
fd49c4280a6fe1dd9ffa8fe81397875aff0f5295
4391 F20101111_AABLJP voss_l_Page_017thm.jpg
3ce1ed6db199eef7291ade1e7a8f6918
c7eecb26b21c5f50bc11911f9e13661c571254fd
7686 F20101111_AABLKE voss_l_Page_047thm.jpg
b7578f76198be2b4fda934a0f1d2eac6
9482b97bc42034e06b0ae3bfc463adda0e379659
6341 F20101111_AABLJQ voss_l_Page_088.pro
7c00c8ecd5c48cfc96b46c167d192b6c
e9ad20c5b9b73ea6c969650dba33c55e140f01d1
39984 F20101111_AABLKF voss_l_Page_076.jp2
3a0435652ba68650b2dbc89d1c554467
af946e0b6f4bd9949ec6b86802e0afd06f52baaf
5828 F20101111_AABLJR voss_l_Page_062.QC.jpg
5896291657ec6fd188b65f98ab27fcfe
991970fdb72db561c41221358607d8ac2216b517
1051818 F20101111_AABLKG voss_l_Page_073.jp2
4dd4613d555469647930522859f5f762
c82c21d3b0ee46caa5e8e3d8e05f4610010ed14c
31305 F20101111_AABLJS voss_l_Page_069.jpg
b49daa3d4f2f42fd1c1dc1a1106487c4
d7a9b93a6295cbbb85a057a68b6f02b6c340a2fe
F20101111_AABLKH voss_l_Page_133.tif
d0899c2f176ba78b7ccd7f7bda4e9d22
81f27d1abf533974cc245322fba44768221f5076
887661 F20101111_AABLJT voss_l_Page_004.jp2
eeae479a37eb4259098db9c9e7cecd47
1b9c135bb9e081d18291e90b7846952b65a647ac
967407 F20101111_AABLKI voss_l_Page_029.jp2
ef3a3ff96bdc166190d7f042905367ca
eb143ad474cc6ea7e93bc024aa45b98b4a8b1bf3
6542 F20101111_AABLJU voss_l_Page_031thm.jpg
b4e967d6e13cca5270d810cc91674432
915758dfad53ab7a2146aa44333c115c3c5c39de
9232 F20101111_AABLKJ voss_l_Page_034.QC.jpg
3fad4405e1e2c8aecb31f46f35839603
deb57f300c1443bb8cb51d3d90c963b1327fd179
F20101111_AABLJV voss_l_Page_087.tif
1c152b2498c16acda5fbc920b1ec66c4
2312e1db084d895ed420348f8ad82289f0a50967
86550 F20101111_AABLKK voss_l_Page_140.jpg
9b6ddce4064a45e74d87f7ab9a23446d
7b83afa1132184e0b05781b33fd4abc7be3b1c1b
22529 F20101111_AABLJW voss_l_Page_016.jpg
77812ef99613eb8dc339eb02c237cd33
3ab36b48289c6327a5fc0a5c5a758a7ce10a9333
86532 F20101111_AABLLA voss_l_Page_015.jpg
271a26ddea2ac55de7ef4523e2eb9c4b
9e3915151a6af201b0392d7453ba9b376ca2979a
7445 F20101111_AABLKL voss_l_Page_101.pro
33ba9359f022d4dfb524409f0b740c18
fe7ae206eff2c61dac7db65e86c7c4bcc76a802b
51212 F20101111_AABLJX voss_l_Page_050.pro
2d52e1ac943776384fd260afd731126f
23d82ed34f77197f0d7be50daa8a6a51a69aac2b
34582 F20101111_AABLLB voss_l_Page_017.jpg
a2a3924d57d9798e1961438fa342e82f
c2274ef5150d4852a8ebf517dd0de0540d910f02
6221 F20101111_AABLKM voss_l_Page_114.pro
bb5b4efe7ae18847b7a754593c1badb4
8a02f838fea8651274cbad33556c1b9ebe35def3
359774 F20101111_AABLJY voss_l_Page_072.jp2
9e2b8919fd42d379fba9aef94e49832c
775c3d53afc0690743e062b65d98be2d75839e7e
73367 F20101111_AABLLC voss_l_Page_018.jpg
26fc6fb6fe56065f03f537bf3e6a4d06
b12b62c6398a10d18e450ee0c9fc2248efcbfb89
2173 F20101111_AABLKN voss_l_Page_024.txt
41cd6da815ab7093c30b32f06e52c05b
29cc47c2256863621c5b86483da77677d3184cfc
7391 F20101111_AABLJZ voss_l_Page_124thm.jpg
38ec1c0802e2fdb5da6bd3548a2afc6e
2f83b3a271eff7a12a3cadad4f4de95826849ba5
85750 F20101111_AABLLD voss_l_Page_020.jpg
bdca7e88e202a1294dbd83133e9c2a9c
c682baa872a724a03b1930f9f067fdb2f744b8dd
213314 F20101111_AABLKO UFE0021980_00001.xml
9eb6d8bd29f266d986a69c7c77cffb6f
2b51091df7f437e95de10f5bef524ed644573bb6
76167 F20101111_AABLLE voss_l_Page_022.jpg
8c116b49583fe28dd3d3e8929a53b1d0
b490613bcb33fe3dcc90b0e3b36014da4b8d57c4
78110 F20101111_AABLLF voss_l_Page_023.jpg
e818065ef837db31450c981c59af1509
9c508a9eeabb9f1fecc54e018e9f5ec09c6f6090
23177 F20101111_AABLKR voss_l_Page_001.jpg
b13749775fd5d5b811b9952858d20432
5f00f0e3ffcc580b8c9f1f0afe78e929f37918d5
86884 F20101111_AABLLG voss_l_Page_024.jpg
c0060f779b61fa0b0573777d37a3b321
e90b84fb3d105f1b95dade41ae55847a4f48a036
65865 F20101111_AABLKS voss_l_Page_004.jpg
c7074d6dbec256efbe84d53aa3fc5daa
5967683f907242229941be27f5bda6ee4feccc89
82265 F20101111_AABLLH voss_l_Page_028.jpg
ff790bce607f5f5486d378ff7ae03fdb
284c00d7a21c1fb388b020643fe3bf25f74368cb
82876 F20101111_AABLKT voss_l_Page_005.jpg
b69fc8c24d6b52cd5ed6bc944ef5acfd
e3d4af5bcef84863be4631eadb6323739fed2958
69105 F20101111_AABLLI voss_l_Page_029.jpg
e0423523232b217ca1e8c987035a0307
0f06e563d6e4e6a5d3900a56b9292c3e96021582
97527 F20101111_AABLKU voss_l_Page_008.jpg
f6d3d6d32c05de4478207f3571fd80cd
e81a4dd73779f801ecf2f62a386902926c3c6f73
68647 F20101111_AABLLJ voss_l_Page_030.jpg
7fae4d8da2fd50b49c3d6ae616b7d4fc
7eebaba5c070fd2a94560952aa523f42f639e1be
104358 F20101111_AABLKV voss_l_Page_009.jpg
006466cfac2a25d2c563f34fe0fe3052
81f56d4653eadba4ad6ee57aaf5f611ca30f6d77
70992 F20101111_AABLLK voss_l_Page_031.jpg
2a6910837b5343d3a064052cf974e2d0
cb96970a85933927ccfa394bccb5289bafbf3fe9
71648 F20101111_AABLKW voss_l_Page_011.jpg
26921492c3a7dc1a16cc01ef1ef71790
8b6dbcbc89f09368bdf450d88df5c42fd49335a8
32055 F20101111_AABLLL voss_l_Page_032.jpg
48fa5a06ac0ff69f858ec10b5fac2f2f
66a9f87c6325113d693ff9516ee9c242fabbc75d
56915 F20101111_AABLKX voss_l_Page_012.jpg
a17f806a15da1d74509386bb86219767
3e5f5cf9d01735dc32ebbf22dd0fb126bff37f80
84934 F20101111_AABLMA voss_l_Page_051.jpg
06fb10f0d492a01faa1d51bf3f2157fb
7e7e3ff25dbfdd2cda0a55fc7fc3ad8ba1d947dc
28353 F20101111_AABLLM voss_l_Page_034.jpg
b40870528b3ba4e4abdef6f01dc77030
673e98437a82201961007b8c81ff7f521ffd3209
87021 F20101111_AABLKY voss_l_Page_013.jpg
c32dd89a6bc40d4250c1f55ac388c2f6
7b82874a782790293a8eefa32f393ae22635c021
85121 F20101111_AABLMB voss_l_Page_053.jpg
7cd3011e9c91eafda4e7414ca1f6061a
ab5153610d1000d460b00ebdd24421da0ed307a7
38701 F20101111_AABLLN voss_l_Page_035.jpg
ae76c6b3c71c47690bff26ca32ed7925
3e3adaea69e9ddda2a5b2e521d9b017384443523
84658 F20101111_AABLKZ voss_l_Page_014.jpg
040e339e499dd3a3bd852f296bd16a19
7b37fddac826a9ab8fb7e132809ceb3fbbd6c986
90076 F20101111_AABLMC voss_l_Page_054.jpg
03f469aa3f4b78f11383c004da875dc2
4de2a081f33d3abb651b6f0e95006418b84f2532
19565 F20101111_AABLLO voss_l_Page_037.jpg
80afb932fe2ed88d4c67901beaf544d2
feb1037ecea4c2fb28e5b07453a398e3fbbd2dcc
90341 F20101111_AABLMD voss_l_Page_055.jpg
c903df5e66786c7ffbf95ce6c5ac6af6
15ce952255a2f9ead02f2a5758b0c86168682533
30572 F20101111_AABLLP voss_l_Page_039.jpg
7a8fedf29e8f199a135bb5bd89bac7d9
169e27b53afa80360d53da3e230eed9bbf5adb2e
89772 F20101111_AABLME voss_l_Page_056.jpg
6e68ce6c11116a830589284a3c270485
cd41bafdcb4a22d4459e0fe9bc40fe0a7607cd68
18105 F20101111_AABLLQ voss_l_Page_040.jpg
8d89c2594be476a0c3dda7bfe45a9199
1d699ca37ac39dd16f50859e22ac3fc6e8ed59f4
83515 F20101111_AABLMF voss_l_Page_057.jpg
eeaa3603be0d428c5ea5de4940dd45db
9b47706fdcfe75cc0a2c3b9d2b4106b6d1f7d2d3
24022 F20101111_AABLLR voss_l_Page_042.jpg
ec855cb3477693cd521ea3c347853bfd
796ad2454a13c25b9d32b3ed0e9ae8e3f49d479f
18594 F20101111_AABLLS voss_l_Page_043.jpg
8d3e5ad9176a674aa33297236156b588
6548d262b1902d774a7549773c9bb0ea581d5e8a
83932 F20101111_AABLMG voss_l_Page_058.jpg
3d62356c72085d14242fbc9fcbfdf135
a42c1cc725592dcb082c2ba9157c7c4711f01d56
17432 F20101111_AABLLT voss_l_Page_044.jpg
648b96deb7267633b5f30297ddec60ef
186e71c0e026eb30c0d6f2654ce2df00c139b848
77001 F20101111_AABLMH voss_l_Page_059.jpg
f47f7795a6e97249a97e380119c8c780
ee298b63a443f2227aaae6a442e3f0d9644bdbed
22215 F20101111_AABLLU voss_l_Page_045.jpg
4d5fa9508e5a5104ef25d0aa547b7035
7574381367c4451d3958d62ff0323747c499d41c
21684 F20101111_AABLMI voss_l_Page_060.jpg
46b8e16c685c167d9d34902fa38a2a82
df5a193ad6635f77cf21b53a8aa5223689d112d8
83205 F20101111_AABLLV voss_l_Page_046.jpg
92ade535e18427ba6672263456bd8b29
31690b444cc09cdf5e997d52750839dc1864eb46
25645 F20101111_AABLMJ voss_l_Page_061.jpg
4fe66571441e7407d696812e58bf4e71
93e2124a061c8c057214ca957f619080cc33cf1a
88094 F20101111_AABLLW voss_l_Page_047.jpg
5da75d807cdd3ce8826938870cc830f2
05f29cc1b1cce03f38a4de62b37a98ff418346cb
18235 F20101111_AABLMK voss_l_Page_062.jpg
64b8a32eee915cfc1cb6777d39e22c02
0381eb2228c10d08a9e89765f59c67d6f4de9157
85307 F20101111_AABLLX voss_l_Page_048.jpg
9eee3d767ace471fca9b793e6e650c23
e3239c28bd8ae5ac97387efc769157c587ea1a25
42542 F20101111_AABLNA voss_l_Page_083.jpg
7ce177399cc1f808a495bc3c97ec5417
168d0782d4b5c13f51ee3c4df490abf93e1b9132
32341 F20101111_AABLML voss_l_Page_063.jpg
41e2807ad18213d3e26e46042a280a22
1be70487eb12b343ba1d61bf4717ec4672266743
30383 F20101111_AABLNB voss_l_Page_084.jpg
5aea3e0151a61bba7ea48540f883d215
b821198b07c43b28e1de6d659e9846c02b7a5aeb
40578 F20101111_AABLMM voss_l_Page_065.jpg
4947dfbb844674b7f05ec20e11d3fba5
9be40f5c1fc4a64e8498477a66011a50bb57a2e5
87249 F20101111_AABLLY voss_l_Page_049.jpg
ce0be6421cad5b76702c10d8a6a00251
6a1305a2187a88d118b1d603b7498ad12ec3f998
31837 F20101111_AABLNC voss_l_Page_085.jpg
a5fa8f9833cbac6efb35fde822d885cf
0029c2b7e07d6d54d68433cf2ca277ee1d5494d6
45194 F20101111_AABLMN voss_l_Page_066.jpg
a53a410d432f51b1bc4fb4eaa6cc081b
7e6bbffccb796e2770690f13de76fdb64bfa410d
81218 F20101111_AABLLZ voss_l_Page_050.jpg
204364c2cebb80efad1ca36a72dc028e
5966af26e18827f5fa0e94bf3d397272339d5a7d
33945 F20101111_AABLND voss_l_Page_086.jpg
a4815d42e7b8609591c0a8a2fa52dd72
b1ae30ffbe98f7eb17e1f20750fea5ca0d8ecad1
34683 F20101111_AABLMO voss_l_Page_067.jpg
3afb2e08707b43ac351fc0a06168d24e
cc2265bb7d0b2c4a9628c6dd51775d0a43547b63
34247 F20101111_AABLNE voss_l_Page_087.jpg
591a69d64b0b9c0f9e45008ba56b0c1d
5cf7b1b0e3c79efbd7fd3f8bc24a5b2850e700e8
39587 F20101111_AABLMP voss_l_Page_068.jpg
4f300c3e2d4f7adf8c03890f8fff8f97
3bedd26fb7b909ef2d890bad9012ee6c52331fdb
41244 F20101111_AABLNF voss_l_Page_088.jpg
b8a564576bdeb428d209c6c44341a2bf
4edfcdc73521f70e371d3febd6417743c371d415
36463 F20101111_AABLMQ voss_l_Page_071.jpg
7b35a51616d9df3a8081e589a15766bb
25afefc2199daea70dd1de82e6b9d60c9cb4cee9
70609 F20101111_AABLNG voss_l_Page_089.jpg
8be83209ab61d36d76e9abdd6f923715
bb31cf3d6048300418b06d84af9c353ef096edbe
32857 F20101111_AABLMR voss_l_Page_072.jpg
0487c639085bc244d42a157c5b5526b8
b0ffa414d29611938e7f133a7751331ceef104d1
36310 F20101111_AABLMS voss_l_Page_073.jpg
1ad12ba23f7000d78d4831c0530cc501
2ec8dcbd2c3ddb2193073a38d3d3efb0705ad7d7
77870 F20101111_AABLNH voss_l_Page_090.jpg
8903afb81a266abfbc5f1b1cbbe85702
a30b2067b25a56f48752e7516ea79b5a0835c04f
29448 F20101111_AABLMT voss_l_Page_074.jpg
e50dc44c972af47244bc110093b533db
7a8533d1a082dc774d10adeed70c3f6ad1a71131
28022 F20101111_AABLNI voss_l_Page_093.jpg
06971d598775e129239a64d1bd3b20d3
9c855612c6e7f89ed65b07c6d05d0ac9f07a3096
33374 F20101111_AABLMU voss_l_Page_075.jpg
c6f66fd013913534f9729a5a42726316
76b3257556f59145ce69ca51fe38287aa02679b8
16900 F20101111_AABLNJ voss_l_Page_094.jpg
1c84a48109c02822dae9c1d1fcd79edc
8ad5a9bc66ba18e750b56493d03a5be4e3935bab
34956 F20101111_AABLMV voss_l_Page_076.jpg
c3535c31f98f73ee598f20f43e87f1c0
3802fa94e3f7b77c6b252e5a1fdf136a33ed82c8
30475 F20101111_AABLNK voss_l_Page_095.jpg
c8034fe73077953578a1df36ebe572c8
fa5c50205be272a0babf01de8ac3b9b91e4236d3
53686 F20101111_AABLMW voss_l_Page_077.jpg
326fa14114f3f11af413e4d4b10c1adf
92ee7edaa5e29b034c321c3c6747a64a4830e450
60886 F20101111_AABLNL voss_l_Page_096.jpg
8384e922a536eb0f3e519220e0cceaea
bebdc60fdbb8ce2a882c78bd4babbc75a4d0b7f6
31645 F20101111_AABLMX voss_l_Page_078.jpg
565f738d9a3271659de4e53f6e1d971c
2dbf5f47c317151b8d014f84ee347284a589a3a1
27508 F20101111_AABLOA voss_l_Page_114.jpg
4bc9f42ba39b625136190fef1683c111
05e4542b983066783a7c839b59b85c33db78569d
46256 F20101111_AABLNM voss_l_Page_097.jpg
460915578d40b8ecf0d93e8338c149ff
8dad80e9b28a1543a1a961519fe79eb77776c325
F20101111_AABLMY voss_l_Page_080.jpg
5cbddb54aa6b6db6ca9647d815ec3c32
a4fbc12fd10f3250bd44a712de3b70b43b9b5c9f
32815 F20101111_AABLOB voss_l_Page_116.jpg
8333883f46bfabf89d6ce6ca92aa9238
c8406c8ec5587a656a29352f3bbc70c7de8733d6
46076 F20101111_AABLNN voss_l_Page_098.jpg
d0defb0b0430c85832e74ecfa15fe86e
57407a0e3b982cbf9747d136a8409e1dcb883fa2
35043 F20101111_AABLMZ voss_l_Page_081.jpg
e77eb5666db0da3c29b55758a1058a81
908a740032c38b2ed74434e7455ee5b8f06e26fe
24785 F20101111_AABLOC voss_l_Page_117.jpg
600f254d8e898f531940592bc56188d3
3a77c8dbeb28f42d8f9e8493343e0001e2995c83
46501 F20101111_AABLNO voss_l_Page_099.jpg
cade0d5c87029e064f47f123c1c93a0a
6d0db6b371c50262d287dbc4d663db46780c20c7
26838 F20101111_AABLOD voss_l_Page_118.jpg
f18cd335d9723831dce50c3cceeb9142
89c6246ffa97ddc8446943041cb51a84a4c88e1c
31495 F20101111_AABLNP voss_l_Page_100.jpg
ed7a891e7fec3bd85e03ba10af985f62
eda6a37d799ed7867e1d44dbbf1b04415ba9e95a
30545 F20101111_AABLOE voss_l_Page_119.jpg
0628f5454b8e6dff569acca428b2df87
76d09c83ad43b19468c01f096fd7f5477419efe7
31775 F20101111_AABLNQ voss_l_Page_101.jpg
57e83f45fdfccd56a8d72b7d82e35179
c83cc5f75c7dbb8cef16cbed04d3a63f8f244ae8
26092 F20101111_AABLOF voss_l_Page_120.jpg
ca12420fd6e6e2520e68646ba6ac1a7e
950d2907cc6fae963615659aec303908aad1317b
33018 F20101111_AABLNR voss_l_Page_102.jpg
36c0d5c2db565a71eceb48b13b212ff8
abe11db4e618f4144de9a3ac1b78a5aff9a8050b
23289 F20101111_AABLOG voss_l_Page_121.jpg
4c540c27a1025677d7709e8dff63cd7d
ba7d37d9a830e85d5fbe44fe53a851137170074e
85451 F20101111_AABLNS voss_l_Page_103.jpg
b19a5c5c0aa2013f7ef79485ad60790d
0157e8c815db84d841ca48a8046ab5746c5e4e70
82840 F20101111_AABLOH voss_l_Page_124.jpg
ec2b98507dc46e203c204aeaf6a0bf40
6812291ddf9cde9c9ac626350359c7f323179e37
86216 F20101111_AABLNT voss_l_Page_104.jpg
1539dbf790fdd7d9164a7ef1a6942ab5
13fc6dca14e4493c01c79e5b12c9b0541ed800b5
72331 F20101111_AABLNU voss_l_Page_105.jpg
775623885c6657be8bb18fcaddd03eb9
0354094615d020408c43abb445fb7faad5264bd8
86801 F20101111_AABLOI voss_l_Page_125.jpg
1f885fb174cc7fde25cb036f1af384ea
57f67b38990cecfa4277a3da120b05b5f2cf3b40
82317 F20101111_AABLNV voss_l_Page_106.jpg
80d1f702df827d89bf5f95165b86d772
114ef9fbd768afc2aa685beed33ba261e3e21123
86802 F20101111_AABLOJ voss_l_Page_126.jpg
efe04670936fb5b888a31ad04850d64f
fb0f08411930598f7fde88a08766a2face18b2ce
86518 F20101111_AABLNW voss_l_Page_107.jpg
7224d5d39d3210ab18bb827c5b066c5f
40eefc7399eaba7b351bce716fb08f405a891de4
70841 F20101111_AABLOK voss_l_Page_127.jpg
51194305506e7f50a3cd8968d60643d0
3715f4a87220c58551456065af7f209663321c98
83848 F20101111_AABLNX voss_l_Page_110.jpg
cfd3d041b2fe464d266cb7e782b08051
141babc5b9bdb124de86798117fcee7813eeede9
975583 F20101111_AABLPA voss_l_Page_011.jp2
6408fa62bc73ced2cd893cbfd3b60494
7666330734a4da3cba46fe44294d0ce9fce921ed
25754 F20101111_AABLOL voss_l_Page_128.jpg
701cf0c0208c4792d3270b97b322f42b
94f94a1a8f8f100795352aea4e546efc3a5d79f9
22459 F20101111_AABLNY voss_l_Page_111.jpg
67dfaa76c47592acaa028f020be485b4
5c476dfcd53854310eed2bd5bb55076007985506
1051917 F20101111_AABLPB voss_l_Page_013.jp2
a13c7fef430f4b782e76395bcedbc690
b2b73c8acb41bc114760972e94a3dfae8b7a2339
37506 F20101111_AABLOM voss_l_Page_129.jpg
2620367f5820ace8ae18fa80f63a32a6
ceb65580867a0b67ba785a77183c0b3f7c074aaf
42767 F20101111_AABLNZ voss_l_Page_113.jpg
13d35be7ff8e06e5afc38bf5b2417936
88c178656dd6268273c709b2acba2ac9732f3d00
351141 F20101111_AABLPC voss_l_Page_017.jp2
1730f6e9b9ea01396acd8e1682ae1e44
e1e7ac6606422f5e8b3a9174604e0da0cab0e940
34497 F20101111_AABLON voss_l_Page_130.jpg
c58357d40647b0daa882268b78fdef93
0a5876236b164da4835d64f1e3dcd2b15ec125b8
995007 F20101111_AABLPD voss_l_Page_018.jp2
ae26da1bb0d77135e3760b6274a2827e
f7b04dfc3ccb0d73c1cd08be7659713238e82b88
27425 F20101111_AABLOO voss_l_Page_131.jpg
70202bee3517b1922b4445970ae93046
427d17af561ac1e9f695b2fbf86cc04e2e9fc0af
1051969 F20101111_AABLPE voss_l_Page_019.jp2
a93a39dc05686086abc8e9ca378e78ca
0201adbd319b97e7e5f414ee12f2562156f05ed3
84962 F20101111_AABLOP voss_l_Page_134.jpg
22563db549c7c647f98f07fa658f93a4
4114a4187013ed70d21baa7ae6222cc69419c5d4
F20101111_AABLPF voss_l_Page_020.jp2
882f708adb073466dc5c69f5a64d162f
bc6887a21e45c825af7f3405e4836f3daf1d439a
83156 F20101111_AABLOQ voss_l_Page_135.jpg
3b9d7784504fd77a24a050cf46ec12d6
c2c947b38e8908f5c5b86586421d07c40b2f5f60
F20101111_AABLPG voss_l_Page_022.jp2
dadd41a1f87a3fafabb13adacf410722
66331924c461bef44be6d1b4c42f2fce7fb2b6cd
61557 F20101111_AABLOR voss_l_Page_136.jpg
96a33fe052b450d7e26b53c6c9552c38
a1a4d84b052f62a246e3002720de483212a4ffe9
1051983 F20101111_AABLPH voss_l_Page_023.jp2
b898388344d3490a39c51b24d464fcac
e028cefe896636f273bc1498ee35760af8c65266
93588 F20101111_AABLOS voss_l_Page_138.jpg
01f4a8dbc3010b22c43e8e5d3a203114
66423321b5be2330158b3f3e441a6f89b53b4e94
F20101111_AABLPI voss_l_Page_024.jp2
cb1637e13bfc2e82a0eec3a9ab37bb05
00cb732d4efa8c5762add067117fc158d1d90cf9
26229 F20101111_AABLOT voss_l_Page_002.jp2
8a03577988129757a012b2f6b6b0bb92
07b284c687189c729f6fa832915e90e982c25346
21961 F20101111_AABLOU voss_l_Page_003.jp2
4cf47492931d7f1c5f47fe25553fbafa
0710d024544bcc38674244e6aa2075f8e8360a90
1051934 F20101111_AABLPJ voss_l_Page_025.jp2
e46a4bc0fceb637d86347f53d3aa7691
306dd323fa6952fa2902cd21bc583c5764dc0107
F20101111_AABLOV voss_l_Page_005.jp2
55bbcd77d78cc91806d6f8c58b7057fa
b22fe14914ccb3a0aa6f4fdad1ecdad59920515b
713950 F20101111_AABLPK voss_l_Page_026.jp2
fdba3a4db098da21deed183cb2b55f88
cab6a2b3a138ee03a25c6bd372d53771e0af11b2
1051981 F20101111_AABLOW voss_l_Page_006.jp2
1d156ce8677e7cdaaa7f18f67da485e4
5f23e611b9598e370aa18f000c235ae70ff3a9f5
F20101111_AABLQA voss_l_Page_047.jp2
2376156dbe63a41e417d8d121a55fbd7
70eb645f8629c24b8c19078316da585574eb7228
1026370 F20101111_AABLPL voss_l_Page_027.jp2
d7d406dd57207dfc5a5f3ccca47d403b
4df737d01ac3fd444916d5a696765ad9293bbf84
1051980 F20101111_AABLOX voss_l_Page_008.jp2
1ec1bd41a3b06bc1b8c020d37703292f
98ad9359368e20a52ddb6849a9927ef8752c1265
1051984 F20101111_AABLQB voss_l_Page_048.jp2
fa6db10b4b252cb43eef8dc4eb45e319
cddff213892b1dbdd83d4923bb89c2637631f266
1051970 F20101111_AABLPM voss_l_Page_028.jp2
080ef5fbaf6b6ec4780e1c10dc999f73
fd7914b5855e1f0a4fbefea14e350da7ea39b56e
1051941 F20101111_AABLOY voss_l_Page_009.jp2
9bcc8b20f44466ee7a1a3981cef7c065
56176071fde04463ba7c56c551568606db678d71
F20101111_AABLQC voss_l_Page_050.jp2
ed9b9b7265ecd4643a92607f972a0f02
72a9f989da11e2309d63613e15ce8a3ed5a917fa
939500 F20101111_AABLPN voss_l_Page_030.jp2
39ad624a64c16de7db9b2bc69eaa061d
cac6fc8896a84f3617ade2cf3110d00545931c83
1051955 F20101111_AABLOZ voss_l_Page_010.jp2
4cc1a8135d05f554ee6ae0f8d6e8960a
ede1b48fd475e4556cbf796aa21207e39acaf2a1
F20101111_AABLQD voss_l_Page_051.jp2
fc2639a84c2acc48ab099af08c084101
e951fd080f464d97c46f1753f004deff89c7fa5b
388845 F20101111_AABLPO voss_l_Page_032.jp2
910d7bfbd38f3b8827251efabe1204eb
fb9442fd9dae29e296bfdf253a79fb0186c28385
F20101111_AABLQE voss_l_Page_053.jp2
90c4a0c7fbe25d3b3f22bdb940861eae
46478e2337205145b23071deea3f9b0402811914
313979 F20101111_AABLPP voss_l_Page_034.jp2
bf2992a180472a5593da0af51fc085b5
ea51c0bef6652be6539f61f9e8bd88acd04493d2
F20101111_AABLQF voss_l_Page_055.jp2
8f87bc90260e968492bfe60a875ea0a4
558ce372bc3704830d32f38f6de3ae7f1e9c5b54
455328 F20101111_AABLPQ voss_l_Page_035.jp2
5b4981db344e904d2144899f0478f1ef
f956c466e391afd7bfd17d339147580af471b9fe
F20101111_AABLQG voss_l_Page_056.jp2
d9034eef7ab24dcc30bbcd383b5badbc
a70d5f20d9e70ec296e8b7f5232261d934145c09
224508 F20101111_AABLPR voss_l_Page_036.jp2
3bcbb7d38c6f4e07ea1febcb181bd7b7
cee64eb32a8c9107c40a1deafde176636603c195
1051951 F20101111_AABLQH voss_l_Page_057.jp2
252b1556070a7e965fc75b03cdd6e195
e6b8a564f98fc980fb0de5ce5d8e08ee343f04db
165054 F20101111_AABLPS voss_l_Page_037.jp2
4b6029668be36b817f9a8c641184a971
aec390728a449a7436cf038b639afaa4549506de
1051978 F20101111_AABLQI voss_l_Page_058.jp2
b4ee051a675fdd3ca82726d6c3b09c31
20feffc788266906350c3e90cc4c57fbe96afe4e
415273 F20101111_AABLPT voss_l_Page_039.jp2
1c2d76e4d45772e0ff796ccba43df120
ea02b80a44b225d8f3fd7c2842572593de5cf85d
1051932 F20101111_AABLQJ voss_l_Page_059.jp2
cd0d1081e24b43490a1dd6582d69243b
6866bac73caa04cdd49690a93eafd3c9d1367264
283898 F20101111_AABLPU voss_l_Page_040.jp2
9ca5d0de1e2c950bdb1489022bb3eb99
47e17a2d6a271dcdf72c0fb4bc0c99ddd923e767
282129 F20101111_AABLPV voss_l_Page_041.jp2
63264caed26600fcb2745785a43e284d
89a16b89a44800a3f8758e0770db05bef393f1aa
210426 F20101111_AABLQK voss_l_Page_060.jp2
81705e2c21226af2c58580018b5d98a9
3dcaf7c4ed368484c4e855249c6241fe3b72faf3
162509 F20101111_AABLPW voss_l_Page_043.jp2
574b269b8ffac433b153c7d0de6e4208
11fff306f3dbed2d563cb3af0094b8121005e453
267514 F20101111_AABLQL voss_l_Page_061.jp2
835ea0c5d200e75c6d605ca26957b9a3
dcb7cb82805d9c9e8e51291f923449ab97172517
120697 F20101111_AABLPX voss_l_Page_044.jp2
baf9cbcef66c70597baf3bd9e93deda3
e5fdfd3c87d9f45c8059c9f0c4c6e60082fca1f4
1051919 F20101111_AABLRA voss_l_Page_088.jp2
4cf25b5a4184061f0d1e22068522c88d
31a032b031d9b8b0ecc0762ea605bda65f05a747
1051839 F20101111_AABLQM voss_l_Page_066.jp2
2b6fe7ebd16aa176e068ea162e342212
df4f1e729bb6a883d8173e753350acf5e56020bb
258387 F20101111_AABLPY voss_l_Page_045.jp2
70be6f288b11c81346669238c75ef1d8
94ab94a986d1293860c3993f9985dab60e141948
971380 F20101111_AABLRB voss_l_Page_089.jp2
0e8bcedc3721269ec75b81076dc86781
6494ebc2dcc86e335c10a067da24a9efaf32b2d3
455913 F20101111_AABLQN voss_l_Page_067.jp2
d7e49c8c546ffbf9074d9d4a1817d364
1ed9268518850584f648afe7378e0c057e851592
1051950 F20101111_AABLPZ voss_l_Page_046.jp2
d92496ce70b7efe1e0f82cb8fe5df980
0c07973b66b94532be0583b3924a14ab4e0ac4e2
1051976 F20101111_AABLRC voss_l_Page_090.jp2
aa32c9378a6ce2a6b1fcbe12584c113c
fcf83001ab45e9b9ba53617dcf2652f9ac034f3b
387947 F20101111_AABLQO voss_l_Page_068.jp2
3edbde62c72ad615fdbbe0ab9a74a7df
c4cf6e31e76708fd374e95a5f4f80bebe7e86ba3
1051906 F20101111_AABLRD voss_l_Page_091.jp2
77ec29585217156561740d94c9c33c65
46176d1316b17cfa0065f38116a8c8be94744097
291834 F20101111_AABLQP voss_l_Page_069.jp2
819335e8a09c7f8758e499d57275589d
73924d5bef5fb4180f656bde3da779f95a506735
1051921 F20101111_AABLRE voss_l_Page_092.jp2
07b2b300a312c516e504232afd59231f
57d361523db37af1b6bf29df7b94e00725cb5afb
362986 F20101111_AABLQQ voss_l_Page_070.jp2
c52746551ee56df04374664e8413ff89
cea186c55f6865f8d78047f0bb9f8ae82dbf07a3
314465 F20101111_AABLRF voss_l_Page_093.jp2
2988ec5ffaf5afb3bbfc7f3bffdfcf92
55262718934cfbad4907e19214f20c0af0cd8622
443561 F20101111_AABLQR voss_l_Page_074.jp2
725ce87a93e4715f959010c5e63131f7
86f66fbceab4581519a5b93aa8f96891528d07cc
138046 F20101111_AABLRG voss_l_Page_094.jp2
3ba7636d9943ac717177ba7bd8328473
063b758b005e4503de8816c6a35824828a713af2
674016 F20101111_AABLQS voss_l_Page_075.jp2
b47de181c4b16db9d95faf40e8efe32d
553fe8703edcbc41245c70e8c7b08832f88a5cc5
F20101111_AABLRH voss_l_Page_096.jp2
df896874523c646a581026aa76f076d2
bb196f9664bdf2ba78e6a03db335b74e4194280c
963099 F20101111_AABLQT voss_l_Page_077.jp2
7082925f36e6dc54c016c1125a588f27
d28aabdcabadce16033f26a9f8ebbd4e64f95b85
883965 F20101111_AABLRI voss_l_Page_098.jp2
89f8be42c0dc926cc012d6eb7d44377c
e803af3539cfd40d3f85018e084ff1167d6a2846
774868 F20101111_AABLQU voss_l_Page_078.jp2
0c373363580c8844ebd7e2f7d3b1c7e9
7cacb7c17e2021000bc2c99beabf487210c413e2
880505 F20101111_AABLRJ voss_l_Page_099.jp2
b36583b848314f7694b5404f1ecd9612
b3a03779bfb786fc4224366789f2eaad5055152d
308740 F20101111_AABLRK voss_l_Page_100.jp2
75441e46f8a743db02906b295f8062bd
50d2c23b87686d30861f21cc38bbe5d1bc6daaa2
835837 F20101111_AABLQV voss_l_Page_079.jp2
971e524a38789102fcf032e8fc26ffa0
a7c976c11e543a9c0847b3d40aefbbd55c076ca5
541310 F20101111_AABLQW voss_l_Page_080.jp2
baa2faed245a52f6a9c9639b9aa68944
991c68055662119301fc3dea7cdc200550fa6656
34389 F20101111_AABLSA voss_l_Page_119.jp2
3bdebaecdcc5cc587f835a838cbc3f33
7e7139062084a583db8212a6acc4b45d39d5c400
363095 F20101111_AABLRL voss_l_Page_101.jp2
a76ba5310a763b33d167591499d12fe7
faf3aff55e4b79951067b9570600ef9293d28013
352375 F20101111_AABLQX voss_l_Page_081.jp2
5505f435f190bca6eb70eefb7063deb0
baca151189964f7f17779b9c4133b8d0b7582258
29222 F20101111_AABLSB voss_l_Page_120.jp2
35bdc355e2ab2480dba4e96dd1d03187
f86702281313e7adcf5d7ab69d5297d1b3584712
F20101111_AABLRM voss_l_Page_103.jp2
d481cc770de4a3970576a24c99828e44
4c3353a503c6c8050e20e077d3e2ced63c67a8ad
35687 F20101111_AABLQY voss_l_Page_085.jp2
98d9f3fe841803bdb426efba8c8ca22d
fa2359bfa448648e8beb73ea1003736c2e7ed438
845677 F20101111_AABLSC voss_l_Page_122.jp2
bcb286096228dc41b6ce29798ef0b9ab
c67197d0382f204d7faf59427bd36a26fb5b8766
F20101111_AABLRN voss_l_Page_104.jp2
5f95aa669f319e97accfdff1cd4b33ac
acd6a7c521f141e8c764f46af5e917a51b66714c
381841 F20101111_AABLQZ voss_l_Page_087.jp2
31b13ccf923b6996e32bed07b115ebd7
ed49181055ec2fc76983dd22686994da0ce38010
F20101111_AABLSD voss_l_Page_124.jp2
97348b9ecac02830e93140dea78691a4
1fc48be8603e24382cb9cea2bc7b577264763cfa
987369 F20101111_AABLRO voss_l_Page_105.jp2
8fc7f91b48e2b6dd6e982dfdc53b3d9c
e4e1e80a92b6f312ad1af8342b9d6ea3a52c2126
F20101111_AABLSE voss_l_Page_125.jp2
f1238a681c48c1364d4aa515fadf181c
576c7e45dc101bf67fc1eeadf427d030e142382e
F20101111_AABLRP voss_l_Page_106.jp2
c6dcd7c01f0b1813059a49cf1ac67e4e
5a39c2c3a5951848de6ca7eb4f3656609a1161fd
451177 F20101111_AABLSF voss_l_Page_129.jp2
08f4d5bd79572d7cd0d45b10587782d4
a8613505d47ececb2472d563596070f909eabc04
F20101111_AABLRQ voss_l_Page_107.jp2
52afac634ac4bc72341c9d645c277e4c
f43ea92ce2f20910ec825f6becd2bf04ed86d419
468184 F20101111_AABLSG voss_l_Page_130.jp2
1551b86051377d7ecaf7bb669a96f919
4671ff7cd4bb03a048c0c7e8931e7f8e677c5daf
F20101111_AABLRR voss_l_Page_108.jp2
a759dd097268acb45d84007fab6e2c29
02c8454664bc9030ce9aac4db2fe0ec066d172a5
1051945 F20101111_AABLSH voss_l_Page_132.jp2
2121d191dbdd024e83a3248560db0219
7b3dc3ba360ec2d3394a5a1d30e5630f191fc160
1051965 F20101111_AABLRS voss_l_Page_109.jp2
e0ce5956255c83ccc2f6791e3de1540b
b42f5029de5bd163c199d9d3ce9f002b882c792b
1051938 F20101111_AABLSI voss_l_Page_134.jp2
139911067a8113811262ae4be2df7f86
7c14a00d1de4096223f467b383147ab0a596f1cb
F20101111_AABLRT voss_l_Page_110.jp2
fe51888e6da659c809c1cd3ed144907b
72cc7414058f97ef1eb2807644c1470c563ab2c5
839214 F20101111_AABLSJ voss_l_Page_136.jp2
ed50ad8939478fe480d971efb7cd9630
83428813dddad7f3f6dcb967ad6938417ef95eda
449150 F20101111_AABLRU voss_l_Page_113.jp2
01add29f9db3b2eb07ab7f1c47eee6ac
c0778fa75929e61b885f512d66756da0b8c89b6f
F20101111_AABLSK voss_l_Page_137.jp2
a6e75c09349f93dfabcf40bf1eff48b6
71c2505dc5103c45a133aaf99e2f6edd5791c51e
29279 F20101111_AABLRV voss_l_Page_114.jp2
0c0637c624bcb98e68a300b947d93195
c941025311cb2c257b539af7568b8aacd5fae4d0
1051954 F20101111_AABLSL voss_l_Page_138.jp2
c159ba90b98f17ff44c3dc3d435383f8
67063096086e90926d02767e7a87824f37f75490
40728 F20101111_AABLRW voss_l_Page_115.jp2
023e06e401f46e4ea0264f0cfb05d2f5
89120fabb1c6ab14899143fee150df1c9d3a23fb
37414 F20101111_AABLRX voss_l_Page_116.jp2
736e188b9ab8f163710371aff2f74835
fa24c828a42bb9323db5b924ed721f8a20ce4e4c
F20101111_AABLTA voss_l_Page_018.tif
caafec093a24cf692d683353a8ea4d6c
775a47590a95580c92250add097a73a5e424bd7d
F20101111_AABLSM voss_l_Page_139.jp2
0a88d3d6533039afd89685b51d3c8e53
6d86ffef28c37b79f02a8e67e99179f4dfb91c50
26661 F20101111_AABLRY voss_l_Page_117.jp2
d495ace3b6a553618edf793372a99abb
61906feb4753fdacadc350e6ace2741aa847378d
F20101111_AABLTB voss_l_Page_019.tif
30d88e7c081d3b38bc5bad9a317852f2
30ab57d04f9310641c3c1b8b806707c5cb30e8cc
1051982 F20101111_AABLSN voss_l_Page_140.jp2
20c84e872472e7c71f25c19b010c142a
bba4e4cce4fb1fe38d69bdacdaebbd388f7fc21d
31218 F20101111_AABLRZ voss_l_Page_118.jp2
e9461277b6fae32cb8a82dccf13280de
e37ad260d10cff9ac473bb699d1e1fe853e6aa3d
F20101111_AABLTC voss_l_Page_021.tif
d815123f8825c95f25692cadb23453c0
502af819a950c19e2d2a4cb8d32963be9679fd93
F20101111_AABLSO voss_l_Page_142.jp2
2a4ac9060cd034ad6e2728a5415f4818
a13b8c0d9eda8c8df0b7d4b59d527a9afb7877b1
F20101111_AABLTD voss_l_Page_023.tif
fd2d3b2f06f20e4dce990642de0cc0db
49abca255e3052623ec4c5822ed52b3aeb2054b2
580070 F20101111_AABLSP voss_l_Page_143.jp2
eabaab57f0841753016bbdfc4bbc1326
b27a0032a9b0cfea2df0a287d69d8fb9994e5874
F20101111_AABLTE voss_l_Page_025.tif
67f7056367896edbd53eda2e96725bfd
98cb54aba56ab6dcf58ea148724216b406fb1e82
563886 F20101111_AABLSQ voss_l_Page_144.jp2
3c606c5a784d7cf705ab9d471446cd1c
a6387ff6065f402afc02cd5b0e3db8021db97973
F20101111_AABLTF voss_l_Page_026.tif
29448b4389820327f9c2eaaf254592d1
790af3c30aa76d6f581c204e336d1382cb56dd71
F20101111_AABLSR voss_l_Page_001.tif
0aa993b89dee87749c364af0254ae909
7e6663345d997a70b49a54e43ce5bb54018c99d1
F20101111_AABLTG voss_l_Page_027.tif
a4334014a43e0d57461392503d453cb6
c66da791357e3341df520c1ba73f6ef0f1edc09a
F20101111_AABLSS voss_l_Page_004.tif
60a1303789434a65efc4beb41fa9e3e0
622778a8de146ce4c45bb7a5da8ca77edb2746f7
F20101111_AABLTH voss_l_Page_028.tif
44b14df139442aab42f7ff6fed2b5404
e409b60774ac6919d1a708e99f966b59f538cea9
F20101111_AABLST voss_l_Page_006.tif
6e559f6d677665bc8c4c3bad9fc9f1c9
a44b956d52284d82066c0c9ef83bcfadfdd1e5c9
F20101111_AABLTI voss_l_Page_030.tif
6ee45e83bb396371289cc2e41d5cb134
e888bede2c2c5bff98223dd5954dfea7969aa773
F20101111_AABLSU voss_l_Page_007.tif
cfba615833e2b806e25d23a532b0d60a
08f42f87e8033f9438f0589981b2e71bc91cc5d0
F20101111_AABLTJ voss_l_Page_031.tif
1903eed875a26ebebb442f0255ce8c60
d9135e4ce1eceacabce766f9a2ba1265732bf3a8
F20101111_AABLSV voss_l_Page_008.tif
774b14ce0e0ed8eae538d3f916100d9f
670a3324dc91d5060c9772eb023734df43b61cbd
F20101111_AABLTK voss_l_Page_032.tif
1e43766803c2418817cbee6e0f21dc01
28deda4ba62c331553dc2d541aea119b6e927f0b
F20101111_AABLSW voss_l_Page_011.tif
bdc68dab41f4e44708feae4e7eb5ac3c
27b4b10db442d1a04032f152d265fa5cb220b70e
F20101111_AABLTL voss_l_Page_033.tif
027675d294dd05c0a346ac4f1c47897c
f32fee61d041624c158493bcf27c76c4f4edce43
F20101111_AABLSX voss_l_Page_012.tif
0ef0f86fb2af880de724c381ae4ba302
4d68afadec6b2fc7eab05e53c35f0eba61d3f855
F20101111_AABLUA voss_l_Page_062.tif
e8daddeb1d980db2c51f1ecb8f6b6e55
54126b295f291719986e5457ed2c1d6795bdc807
F20101111_AABLTM voss_l_Page_034.tif
5ca89d0ca4bc49676755d7c4b0c7c09a
d31c38d743c1814036a44078aa22a3b64e7c5406
F20101111_AABLSY voss_l_Page_013.tif
57aaa1ebc49446211711b32f22c9a0a0
c5e3b9f1bb88019c43abcaa56d9124b23879441c
F20101111_AABLUB voss_l_Page_064.tif
ab1566616d0a9900c2ddb0340c1e4d7a
e294ffed38466889e0510e2789a19ce2a02ac6ea
F20101111_AABLSZ voss_l_Page_017.tif
e074dd542e310de75e36e7748742157f
6007fba340e435a657a99b4deb42a705d1b81c78
F20101111_AABLUC voss_l_Page_066.tif
5ab08c24fff16fa7f227fae35b5adbc7
2f494f93157f4d435af93898154c0a898878f09f
F20101111_AABLTN voss_l_Page_035.tif
e186a5a56e5d90b47f4bf585fa2cdf40
bd31953c00efad0dddbf830534bfb796771a6a6f
F20101111_AABLUD voss_l_Page_067.tif
042ea80497f2dc3d90593fe90d90c59f
cee48d35979e4daea98d44679223a1496323256f
F20101111_AABLTO voss_l_Page_037.tif
3b8161b179d07d225a91ac43f061eff9
0440d48a948492b01715edf6de9d13f1288d3a69







THERMALLY STABLE OHMIC AND SCHOTTKY CONTACTS TO GaN


By

LARS FREDRIK VOSS


















A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA

2008

































2008 Lars Fredrik Voss



































To my grandparents.









ACKNOWLEDGMENTS

First and foremost I would like to thank my advisor Prof. Stephen J. Pearton for his

guidance and for all of the support and opportunities he provided. I would also like to thank my

other supervisory committee members, Fan Ren, Cammy Abernathy, David Norton, and Rajiv

Singh, for their help and their time.

Thanks go to Dr. Patrick M. Lenahan for providing me with my first experience working in

the semiconductor field while I was an undergraduate at Penn State as well as all of his advice

and encouragement.

I thank the members of the Pearton, Ren, and Abernathy research groups with whom I

have had the opportunity to work, including Luc Stafford, Kelly Ip, Jon Wright, Wantae Lim,

Hungta Wang, Sam Kang, Travis Anderson, Soohwan Jang, Brent Gila, Jerry Thaler, Jennifer

Hite, Mark Hlad and many, many more. I would also like to thank Ivan Kravchenko for his

support in the UF Nanofabrication Facility.

Thanks also go to all the people at Sandia National Laboratories who gave me the

opportunity to work there and with them for two enjoyable summers. I want to say thank you to

my mentors, Randy J. Shul, Albert G. Baca, and JeffE. Stevens, my managers, Charles Sullivan

and Dale Hetherington, as well as all of the people I had the pleasure to work with while at

Sandia including Carlos Sanchez, David Torres, Melissa Cavaliere, Karen Cross, Mark

Overberg, Michael Cich, and many others.

Thanks also go to all of my family and friends as well.









TABLE OF CONTENTS

page

A CK N O W LED G M EN T S ................................................................. ........... ............. .....

L IS T O F T A B L E S ................................................................................. 7

LIST OF FIGURES .................................. .. ..... ..... ................. .8

A B S T R A C T ................................ ............................................................ 1 1

CHAPTER

1 INTRODUCTION .................................................. .......... 13

2 BACKGROUND .................................. ... .. .... ...... ................. 18

2 .1 G allium N itride P roperties........................................................................ .................. 18
2.1.1 Fundam ental Properties ..............................................................................18
2.1.2 Electronic Properties ...................................................................... 19
2 .1.3 C ry stal Structure .............................. ........................................ ........... ... 19
2.2 Properties of the Contact Materials to be Studied ........................................................19
2 .2 .1 B o rid e s ................................. .......................................................... ............... 19
2 .2 .2 N itrid e s .......................................................................... 2 0
2 .2 .3 Iridiu m .............................................................................2 0
2.3 E electrical C contacts .................. ................. .............. ............... ...... .... 21
2 .3 .1 O hm ic C contacts ................ ........................................ .............. .. .... ...... .. 22
2.3.1.1 Ohm ic contacts to p-G aN .................................................................. 23
2.3.1.1 Ohm ic contacts to n-G aN .................................................................. 24
2.3.2 Schottky Contacts ................. ............... ..................... ............ 25
2 .4 E xperim ents ...............................................................................27
2.5 Characterization Techniques ................................................. ............................... 29
2.5.1 Current-V oltage ................................................... .............. .. ...... 29
2 .5.2 C apacitance-V oltage...................................................................... ...................30
2.5.3 X -ray Photoelectron Spectroscopy .................................... .................................. 31
2.5.4 Auger Electron Spectroscopy ................................................... ..................31

3 THERMALLY STABLE OHMIC CONTACTS TO p-GaN.................................................46

3.1 O hm ic C ontacts ......................................................... ................. 46
3.1.1 Fabrication of O hm ic Contacts......................................... .......................... 46
3.1.2 N itride-B asked C ontacts............................................................................ ...... 47
3.1.2.1 Experim ent and discussion.................................................. ...................47
3.1.2.2 Summary ....................... ..... ................. 50
3.1.3 Tungsten Boride and Chromium Boride-Based Contacts and Long Term
T herm al A going of B rides ........................................ ............................................50
3.1.3.1 Experim ent and discussion....................................... ......... ............... 50









3 .1.3 .2 Su m m ary ............................................................................. 52
3.1.4 Contact Resistance for Other Boride-based Contacts ..........................................52
3.1.4.1 Titanium boride-based contacts .................... ......... .................52
3.1.4.2 Zirconium Boride-based contacts.............................................................54
3.1.4.3 Gallium Nitride//Tungsten Boride-based contacts.................. ............55
3.1.5 Iridium-Based Contacts .............................. ........57
3.1.5.1 Experim ent and discussion....................................... ......... ...............57
3 .1.5 .2 Su m m ary ...............................................................59
3 .2 C o n c lu sio n s ............................................................................................................... 5 9

4 O H M IC C O N TA C TS TO n-G aN ...................................................................... ..................89

4 .1 E x p e rim e n t ........................................................................................................................8 9
4.2 Results and D discussion .................................... ..... .......... .............. .. 90
4 .3 C o n c lu sio n s ................................................................................................................. 9 2

5 BORIDE-BASED SCHOTTKY CONTACTS TO p-GaN.................... ........................... 103

5 .1 In tro d u ctio n ............................................................................................................... 1 0 3
5.2 E xperim mental D details ............................................................................ .................... 104
5.3 R results and D discussion ..................................... .................. .. .......... 106
5.4 C onclu sions ...................................... .................................................... 110

6 BORIDE AND IR BASED CONTACTS FOR LIGHT EMITTING DIODES......................123

6 .1 In tro d u ctio n ............................................................................................................... 12 3
6.2 E xperim mental .......................................... .... .......... ................. ............ 124
6.3 R results and D iscu ssion ................................................................................ ..... ......125
6 .4 C o n c lu sio n s ............................................................................................................... 12 7

7 C O N CLU SIO N ......................................... .... .............. ................... ............ 132

L IST O F R E F E R E N C E S .................................................................................... ...................137

B IO G R A PH IC A L SK E T C H ......................................................................... .. ...................... 144

















6









LIST OF TABLES


Table page

2-1 B ulk G aN properties .......................... .......................... .. .. .. ......... ........ 33

2-2 Properties of comm on semiconductors ..................................................... ... .......... 34

2-3 P properties of the b rides ..................................................................... .... .....................35

2-4 P properties of the nitrides ............................................................................. .............. 36

2 -5 P ro p erties o f Ir ..................................................................................... ........................3 7

3-1 Concentration of elements detected on the as-received surface (in atom%) .....................60

3-2 Concentration of elements detected on the as-received surfaces (in atom%) ....................61

3-3 Concentration of elements detected on the as-received surfaces (in atom%) ....................62

3-4 Sum m ary of specific contact resistances ........................................ ....................... 63

4-1 Percent change in specific contact resistance during thermal aging..............................94

5-1 Comparison of different barrier height calculations ........................................................ 112

6-1 Influence of long-term aging at 2000C and 350C on the turn-on voltage and reverse
current of InGaN/GaN MQW-LEDs. ................................... ............... 127









LIST OF FIGURES


Figure page

1-1 M market forecast for GaN-based devices................................ ........................ ........ 17

2-1 Intrinsic carrier concentration of GaN, GaAs, and Si........................................................38

2-3 Flat band diagram for a p-type Ohmic contact ...................................... ............... 40

2-4 Flat band diagram for a p-type Schottky contact..........................................................41

2-5 LED cross section (a) before and (b) after processing........................................... 42

2-6 L inear transm mission line pattern .............................................................. .....................43

2-7 R resistance v s. pad spacing plot............................................................... .....................44

2-8 Schottky contact schem atic............................................ ................... ............... 45

3-1 Specific contact resistance and sheet resistance under the contact ofNi/Au/ X/ Ti/Au
contacts as a function of anneal temperature. .......................................... ............... 64

3-3 Scanning electron microscopy images of Ni/Au/TaN/Ti/Au contacts (a) as deposited
(b) annealed at 600 o C (c) annealed at 7000C and aged at 200 C until the contacts
became non-Ohmic and (d) annealed at 1000 C............. ................... ..... ...........66

3-6 Specific contact resistance versus measurement temperature. ........................................69

3-7 Specific contact resistance and sheet resistance under the contact as a function of
long term therm al aging at 350 C. ............................................ ............................. 70

3-8 Depth profiles of W2B-based contacts (a) as deposited (b) annealed at 600 C (c)
annealed at 700 C and aged at 3500C and (d) annealed at 1000 C..............................71

3-9 Specific contact resistivity of Ni/Au/TiB2/Ti/Au Ohmic contacts and p-GaN sheet
resistance under the contact as a function of annealing temperature.............................72

3-10 Secondary electron images of Ni/Au/TiB2/Ti/Au contact pads on p-GaN as-deposited
(top) or after annealing at either 800(center) or 9000C (bottom). ..................................73

3-11 Surface scans of Ni/Au/TiB2/Ti/Au Ohmic contacts on p-GaN as a function of anneal
temperature. The as-deposited sample is at top, that annealed at 8000C at center and
that at 9000C at bottom .......................................... .............................74

3-12 Depth profiles of Ni/Au/TiB2/Ti/Au Ohmic contacts on p-GaN as a function of
anneal temperature. The as-deposited sample is at top, that annealed at 800 pC at
center, and that at 900 oC at bottom ............................................................................ 75









3-13 Specific contact resistance of Ni/Au/ZrB2/Ti/Au and ZrB2/Ti/Au Ohmic contacts and
p-GaN sheet resistance under the contact as a function of annealing temperature............76

3-14 Surface scans and depth profiles of Ni/Au/TiB2/Ti/Au Ohmic contacts on p-GaN as a
function of anneal tem perature. ............................................... .............................. 77

3-15 Scanning electron microscopy images of Ni/Au/TiB2/Ti/Au contact pads on p-GaN
as-deposited (top) or after annealing at either 750 (center) or 8000C (bottom) ...............78

3-16 Surface scans and depth profiles of ZrB2/Ti/Au Ohmic contacts on p-GaN as a
function of anneal tem perature. ............................................... .............................. 79

3-17 Elemental maps obtained from scanning AES of ZrB2/Ti/Au Ohmic contacts pads on
p -G aN ................... ............................................................ ................ 8 0

3-18 Specific contact resistivity of W2B/Ti/Au Ohmic contacts and measured p-GaN sheet
resistance under the contact as a function of annealing temperature.............................81

3-19 AES surface scans of W2B/Ti/Au Ohmic contacts on p-GaN as a function of anneal
tem p eratu re. ............................................................ ................. 82

3-20 Depth profiles of W2B/Ti/Au Ohmic contacts on p-GaN as a function of anneal
tem p eratu re. ............................................................ ................. 83

3-21 Current-voltage curves for Ni/Au/Ir/Au contacts. .................................. .................84

3-22 Current-voltage curves for Ni/Ir/Au contacts ........................................ ............... 85

3-23 Depth profiles for Ni/Au/Ir/Au contacts (a) annealed at 300 C (b) annealed at 500
C and (c) annealed at 700 C ................................................ ............................... 86

3-24 Depth profiles for Ni/Ir/Au contacts (a) annealed at 300 C (b) annealed at 500 C
and (c) annealed at 700 oC ......................................... ..................... .. .....87

3-25 Scanning electron microscopy images of Ni/Au/Ir/Au contacts........................................88

4-1 Specific contact resistance as a function of anneal temperature........................................95

4-2 Scanning electron microscopy images of annealed contacts.............................................96

4-3 Depth profiles of Ti/Al/TaN/Ti/Au contacts (a) as deposited (b) annealed at 6000C.
(c) annealed at 8000C and (d) annealed at 8000C and aged at 350C. .............................97

4-4 Depth profiles of Ti/Al/TiN/Ti/Au contacts (a) as deposited (b) annealed at 6000C.
(c) annealed at 8000C and (d) annealed at 8000C and aged at 350C. .............................98

4-5 Depth profiles of Ti/Al/ZrN/Ti/Au contacts (a) as deposited (b) annealed at 6000C.
(c) annealed at 8000C and (d) annealed at 8000C and aged at 350C. .............................99









4-7 Specific contact resistance as a function of anneal time..............................101

4-8 Specific contact resistance as a function of long term thermal aging ...........................102

5-1 XPS spectra without (top) and with (bottom) a boride overlayer. The left-hand
spectrum in the top figure corresponds to the Ga 3d core level whereas the right-hand
panel presents the spectrum of the valence band region ........................ .................. 13

5-3 Forward current-voltage characteristic of W2B-based (top) and W2B5-based (bottom)
Schottky diodes as a function of annealing........................................... ...............115

5-4 Influence of the annealing temperature on the characteristic energy related to the
tunneling probability. Dashed and dotted lines correspond to the values of Eo for
NA4 1019 and 5x1019cm -3 respectively. .................................................................... ...116

5-5 Influence of the annealing temperature on the apparent Schottky barrier height
derived from IV m easurem ents............................................................................ ... .... 117

5-6 Dependence of the apparent Schottky barrier height on the parameter defined as the
difference between the valence band maximum and the position of the Fermi level.
L o w a n d h ig h ....................................................................................................1 1 8

5-7 As-measured and after oxide correction dependence of C-2 versus V of Au/Pt/W2B/p-
GaN Schottky diodes. The measurement frequency was set to 1 kHz. .........................119

5-8 Reverse current-voltage characteristic ofW2B-based Schottky diodes as a function of
m easurem ent tem perature. ....................................................................... .................. 120

5-9 Influence of the annealing temperature on the breakdown voltage. ............................ 121

5-10 Depth profiles of W2B/Pt/Au contacts and W2B5/Pt/Au rectifying contacts (a,b)
before and (c,d) after annealing at 600 C.............. ................... ........................ 122

6-1 Optical micrograph of an as-fabricated MQW-LED. The p-contact at the center of
the diode is 80 m in diam eter ......... ............................ ..... ................... ............... 128

6-2 L-I characteristics of MQW-LEDs with Ni/Au, Ni/Au/TiB2/Ti/Au, and Ni/Au/Ir/Au
p-Ohmic contacts. The inset shows emission spectra from as-fabricated LEDs at
various injection currents ....... ................................... ....................... ............... 129

6-3 Influence of long-term aging at 2500C and 350C on the I-V characteristics of LEDs
with (a) Ni/Au, (b) Ni/Au/TiB2/Ti/Au, and (c) Ni/Au/Ir/Au p-Ohmic contacts ...........130

6-4 Image of aged LEDs with (a) Ni/Au and (b) Ni/Au/TiB2/Ti/Au p-Ohmic contacts. In
(a), the picture was taken for a forward bias of 10 V (I= 80 ptA), while the forward
voltage in (b) w as 4.5 V ( = 300 A ). ........................................................................ 131









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

THERMALLY STABLE OHMIC AND SCHOTTKY CONTACTS TO GaN

By

Lars Fredrik Voss

May 2008

Chair: Stephen J. Pearton
Major: Materials Science and Engineering

This dissertation is focused on the development of Ohmic and Schottky contacts to both n-

and p-type Gallium Nitride for use in microelectronic and optical devices. The goal is to develop

low resistance contacts with greater thermal budgets and superior thermal aging characteristics to

those commonly in use today as well as to understand the mechanisms by which these contacts

may fail. In addition, p-type Ohmic contacts have been used to fabricate light emitting diodes

(LEDs) which display far superior aging properties than those made with conventional Ni/Au

contacts.

Ohmic contacts to p-GaN were fabricated using a variety of refractory materials. The

materials examined were of three basic types: boride, nitride, and the refractory metal Ir. The

boride family includes W2B, W2B5, CrB2, ZrB2, and TiB2. The nitrides examined were TaN,

TiN, and ZrN. Contacts based on these materials were fabricated using either a

GaN//Ni/Au/X/Ti/Au, GaN//X/Ti/Au, or GaN//Ni/X/Au scheme, where X is the refractory

material. Contact resistances as low as -1 x 10-4 Q/cm2 were consistently achieved after

annealing at temperatures from 500-10000C for 60 s in N2 using these materials for p-GaN with a

carrier concentration of-1 x 107 cm-3. In addition, high temperature thermal aging was









performed on selected schemes as well as on devices in order to attempt to estimate long term

performance.

Nest, nitride-based Ohmic contacts to n-GaN are examined. The nitride was used to

replace the conventional Pt, Ni, or Mo diffusion barrier in Ti/Al based contacts, for a contact

scheme of GaN//Ti/Al/X/Ti/Au. These contacts achieve a similar specific contact resistance of

~1 x 10-4 Q/cm-2 for samples with a carrier concentration of 1 x 1017 cm-3 as that achieved with a

Ti/Au/Pt/Au contact. The contacts are also examined as a function of aging and are found to

display less intermixing of layers than those fabricated with a Ni diffusion barrier.

Schottky contacts to p-GaN were also fabricated using the family ofboride based materials

with the scheme GaN//X/Pt/Au. These were found to exhibit tunneling transport through the

metal-semiconductor junction and barrier heights of >3.5 eV were determined from the fit of the

current-voltage (IV) curves. X-ray Photoelectron Spectroscopy was used to determine a true

barrier height for these borides on p-GaN of 2.7 eV, close to that expected from the Schottky-

Mott model. Capacitance-voltage (CV) measurements confirm the IV barrier height, but reveal a

thin interfacial layer likely arising from oxidation or defects at the surface.









CHAPTER 1
INTRODUCTION

Semiconductor technology and its rapid improvements in device speed, yield, and density

have enabled much of the technology and advancements the world has experienced over the past

half century. Beginning with the first transistor developed in 1947 at AT&T Bell Labs and then

in 1957 with the first integrated circuit by Jack Kilby, the pace of improvement in

semiconductors and integrated circuitry is unmatched by any other technological development in

human history, and few have been more influential. Today, what once seemed ridiculously

optimistic is possible and billions of transistors are present in a single integrated circuit [1].

Silicon based technology has come to dominate an overwhelming share of the semiconductor

market, and for good reason. The main advantage of Si-based technology is the overwhelming

knowledge base that exists as well as the fact that is possesses a wide range of acceptable

properties. It possesses a higher melting point than Ge, from which the first devices were made,

as well as a stable oxide which displays low leakage current. Further, nearly all of its properties

fall into the acceptable range necessary for use in nearly all applications, including a reasonable

band gap, carrier mobility, and breakdown field.

For certain areas of application, however, Si is unacceptable. For applications requiring

higher speed electronics, GaAs dominates the market. This is due to a much higher electron

mobility, greater than 6 times that of Si. In addition, Si is also inappropriate for optoelectronic

applications due to its direct band gap. GaAs-based devices are also useful in this area.

However, for high power, high temperature electronics as well as emission of wavelengths in the

blue to ultra-violet range, GaAs's properties are insufficient. It is therefore necessary to develop

fabrication techniques for alternatives to Si and GaAs [2,3]. For these applications GaN is the

material of choice. Its wide, direct band gap of 3.4 eV allows for emission in the UV range,









while the use of InGaN and AlGaN allow for tunable emission over the range 1.92 to 6.2 eV. [4-

37]

Development of GaN-based electronics has been necessitated due to the limitations of SiC,

another semiconductor used for high power and high temperature electronics. SiC MOSFETs

are limited by primarily thermal oxides, as the gate contact degrades and becomes leaky at higher

operating temperatures. In addition, the electron mobility is low, only -400 cm2/V-s, resulting in

low power-added efficiencies of less than 30% for operation between 1 and 5 GHz. Because of

its nature, GaN is a superior choice due to its wider bandgap as well as greater chemical stability,

leading to improved performance at higher operating temperatures.

GaN based electronics are already in production, most notably high electron mobility

transistors (HEMTs). Other devices of interest include heterojunction bipolar transistors (HBTs)

and metal oxide field effect transistors (MOSFETs). In order to fully realize the potential of

GaN-based electronics, improvements in their processing must be achieved in addition to the

significant improvements in material quality that are necessary. This includes improvements in

high temperature processing and operation, one area of which is the development of both Ohmic

and Schottky contacts which are able to withstand high temperatures without degradation of

device characteristics or contact structure [4]. This becomes especially important as the quality

of the materials improves and failure is due to the contacts.

While electronic applications for GaN are important, optoelectronic applications are

currently dominant. As Figure 1-1 indicates, most of the current and projected market are for

optical applications. The first bright blue LED using GaN to be put into production was

demonstrated by Shuji Nakamura in 1993 at Nichia Corporation. Since that time, a great deal of

progress has been made in the attempt to realize the full potential of GaN for light emitting









applications. The promise of solid state lighting is one that grows more important with each

passing year, as the world's insatiable hunger for energy grows while reserves of conventional

fossil fuels struggle to keep up. Because production cannot indefinitely keep up with demand,

focus should and is on increasing the efficiency of consumption. Nearly 10% of world energy

consumption is used for lighting. Most of this is currently done with extremely inefficient

incandescent bulbs. While compact fluorenscent lighting has grown more popular and offers a

substantial improvement in energy efficiency, it cannot compete with that possible with the wide

spread use of LEDs. Further, CFBs present a hazard due to the non-trivial amount of mercury

present in each bulb. Because of this, any large scale use of the bulbs must go hand in hand with

proper disposal of them in order to prevent contamination of the environment. Solid state

lighting does not present this problem.

While many challenges still exist to realizing commercially viable solid state lighting

sources, including significant improvement in the efficiency of the green emitting alloys as well

as possible improvements in the phosphors used to emit the yellow and white light desired, it is

still important to develop improved contact structures in order to allow for longer life times and

operation in environments with higher temperatures. For p-type GaN especially, existing contact

schemes could be improved to offer more stable performance. Failure of GaN-based devices due

to contact failure should be minimized or eliminated.

The goal of this work is to develop improved contact schemes for both electronic and

optoelectronic applications that will offer acceptable specific contact resistances and barrier

heights which will be stable over a wide range of processing temperatures as well as during long

term high temperature aging. Chapter 2 deals with the basic properties of GaN as well as an

overview of the processing and characterization techniques that were utilized during this work.









The rest of this dissertation is devoted to the experiments and analysis. First, Ohmic contacts to

p-GaN using a variety of high temperature materials are discussed. The next section deals with

Ohmic contacts to n-GaN. A section on Schottky contacts to p-GaN follows. The final chapter

discusses the use of the Ohmic contacts to p-GaN in light emitting diodes.












MARKET FORECAST FOR GAN-BASED




Revenue


5-
0 Electr m s

O0ptical


3 -









2001 2003 2005 2007 20D9


Source: Strategies Unlimited


Figure 1-1 Market forecast for GaN-based devices









CHAPTER 2
BACKGROUND

2.1 Gallium Nitride Properties

2.1.1 Fundamental Properties

The materials properties of GaN make it desirable for a variety of applications. For optical

electronics, GaN holds great promise. Because it is a direct band gap semiconductor, it is an

efficient light emitter. GaN fills the need for a blue and ultraviolet wavelength emitter that is left

by other materials due to its band gap of 3.475 eV. Its exciton binding energy is 28 meV. This

is not as high as another blue emitters, ZnO, but GaN technology is more developed. The first

GaN LED was introduced in 1993; however, further development of the GaN system is

necessary for better LEDs and LDs.

In addition to optical applications, GaN is robust and therefore a candidate for high

temperature and high power applications, with potentially superior electronic properties when

compared with SiC [38]. The reason for this is due to the relatively wide band gap when

compared to Si and GaAs. Figure 2-1 displays the intrinsic carrier concentration as a function of

temperature of GaN, Si, and GaAs. Because of the wide band gap, the intrinsic carrier

concentration is much lower; 1015 cm-3 is reached at 3000C for Si, 5000C for GaAs, and at

10000C for GaN. Because of this, GaN can be operated at much higher temperatures before

device performance degrades and breaks down.

Table 2-1 contains some basic physical properties of GaN [39,40]. A key problem in

device operation is the presence of the lattice mismatch of GaN with the commonly used

sapphire (13%) and SiC (3%) substrates, which results in a large number of lattice defects which

hinder electrical conduction.









2.1.2 Electronic Properties

The properties GaN possesses are advantageous for a variety of reasons. They, as well as

those of other common semiconductors, are displayed in Table 2-2 [40,41,42]. It has a high

breakdown field, greater than 50 times that of Si or GaAs, which allows for its use in high power

electronic applications. The high electron mobility for its 2D electron gas and saturation velocity

allows its use in high speed electronics. In addition, heterostructures such as AlGaN/GaN allow

for the manufacture of interesting and high speed devices such as HEMTs. Recent work

suggests that the 1.5 W/cm may be a lower limit for the thermal conductivity of GaN. GaN also

shows excellent resistance to irradiation.

2.1.3 Crystal Structure

The thermodynamically stable phase of GaN is the hexagonal wurtzite structure, referred

to as the a-phase. In addition, a metastable P-phase with a zinc blende structure exists. These

two phases only differ in the stacking sequence of Ga and N; their coexistence in epitaxial layers

is possible due to stacking faults [43]. The lattice parameters are a = 3.189 A and c = 5.185 A.

The wurtzite structure of Ga-face and N-face GaN is shown in Figure 2-2.

2.2 Properties of the Contact Materials to be Studied

2.2.1 Borides

The family ofborides to be studied includes the refractory compounds W2B, W2B5, CrB2, TiB2,

and ZrB2. Previous work has shown the promise of using refractory materials such as W as

contacts to p-GaN. These borides also possess many of the requisite characteristics for use as

electrical contacts for semiconductor devices. Table 2-3 summarizes known properties of the

borides to be studied. Each of the borides possesses a high melting temperature, good thermal

and electrical conductivities, high work functions, and large heats of formation. Each of these

properties is important when considering a material for use in electrical contacts. The high









melting temperature is necessary in order to ensure that the material is able to withstand the

elevated annealing temperatures which are generally necessary for Ohmic contact formation. A

high melting temperature is an indication of material stability as well. A large electrical

conductivity is necessary so as to achieve a low resistance within the contact itself as well as to

avoid too much self heating, which can lead to inconsistent performance and diffusion between

contact layers during operation. A large thermal conductivity is necessary for this reason as

well. For a p-type material, a large work function is desirable in order to achieve the proper band

bending for thermionic carrier transport or at least to minimize the barrier for tunneling transport.

Finally, a large heat of formation is desirable in order to minimize reactions at the

semiconductor-contact interface during operation. These materials have previously been

examined for use in contacts to n-GaN and have displayed good characteristics as a diffusion

barrier in a Ti/Al-based contact compared to more conventional metals such as Pt or Ni.

2.2.2 Nitrides

The second class of materials to be studied is the nitrides. The materials chosen were TaN,

TiN, and ZrN. Table 2-4 summarizes the properties of these nitrides. They possess many of the

same characteristics as the borides and are therefore also worthwhile to study. These include a

high melting temperature, good thermal and electrical conductivities, a work function, and large

heats of formation. These nitrides have been studied for use in thin film resistors, wear-resistant

coatings on tools, thermal printer heads, gate electrodes, and diffusion barriers in Cu

interconnection, which suggest their suitability as diffusion barriers for GaN-based electronics.

They will be examined in this case for Ohmic contacts to both n- and p-type GaN.

2.2.3 Iridium

The final material to be studied for use in contacts to p-GaN will be Iridium. Properties of

Ir are summarized in Table 2-5. Note that Ir has the required high melting temperature, large









work function, and reasonable thermal and electrical conductivities. Ir has previously been

studied for use as a contact to p-GaN. Here, it will be examined for use in Ni-based contacts to

p-GaN as an intermediate layer. It has already been demonstrated that Ir acts a superior diffusion

barrier in Ti/Al-based contacts to n-GaN compared to Ni.

2.3 Electrical Contacts

High quality electrical contacts to semiconductors are critical for the manufacture of all

types of devices. While the requirements for specific devices and applications may differ, it is

always necessary to connect these devices to the outside world. Generally, a contact that

displays the lowest specific contact resistance with minimal drift during extended operation is

the most desirable, although for certain applications other properties may be more critical. For

instance, in light emitting devices it is often desirable to have either transparent contacts or

contacts with high reflectivities in the range of emission wavelengths in order to maximize light

output. For high temperature applications it is critical that contacts are stable during operation at

high temperature; thus materials which display little intermixing and minimal reactions are

necessary in addition to the requirement of good electrical and thermal conductivities.

Electrical contacts to semiconductors are generally understood as the junction between the

semiconductor and a metal or other contact material, as well as any layers of metal or other

material above this. The electrical properties of the contact are controlled by the materials at the

junction and it is often necessary to anneal the contacts at elevated temperatures in order to

achieve the required properties through elimination of defects, removal of compensators such as

Hydrogen, or formation of low resistance intermetallic phases. In addition, the layers above the

junction itself can also influence the performance of the contact. It may be necessary to use an

overlayer such as Au in order to promote current spreading so that the conduction is spread

evenly over the area of the contact as well as to prevent oxidation of the contact at room









temperature. Other materials may be used as diffusion barriers in order to minimize intermixing

between various contact layers which may lead to the formation of undesirable secondary phases,

segregation of layers when it is desirable that they be intermixed, or even to prevent out-

diffusion of the semiconductor to the surface of the contact. Two types of contacts exist, and it is

important to understand what materials properties may be important in order to fabricate each.

2.3.1 Ohmic Contacts

Ohmic contacts are those which obey Ohm's Law:

V = IR(2.1)

Note that this is a linear relationship and thus current is allowed to flow into and out of the

semiconductor without distortion of the signal. In general, the most critical component of an

Ohmic contact is a low specific contact resistance. This will minimize the power consumption of

the contact, which will also serve to minimize the internal heating of the contact and thus allow

for prolonged operation with predictable performance. An ideal Ohmic contact would have no

noticeable effect on device performance, as it would consume none of the power in the system

nor would it cause any rise in temperature during operation.

In theory, fabrication of Ohmic contacts should be straightforward. For semiconductors with

moderate doping, the dominant mechanism of current flow is generally thermionic emission,

which is governed by Equation 2-2:


Pc qA exp 0 (2.2)
qA- T kT

where k is Boltzmann's constant, q the electronic charge, A** is the Richardson's constant, T the

temperature, and (b the barrier height. In this model, electrons are excited over a barrier. It is

clear, then, that the barrier height should be small in order to produce a low contact resistance.

For p-type conduction, a metal with a work function larger than that of the semiconductor should









be sufficient. The desired flat band diagram for this situation is shown in Figure 2-3 [40]. It is

clear from this diagram that holes may pass easily between the metal and the contact at the

junction.

However, for some semiconductors it can be difficult to find materials systems that produce an

Ohmic contact through thermionic emission, especially for p-type semiconductors. A wide band

gap and large electron affinity, such as with GaN, effectively limits the choices to materials

which have a very high work function. A second approach for formation of Ohmic contacts to

semiconductors utilizes another form of carrier transport: tunneling. This type of transport is

characterized by the carriers tunneling through the potential barrier at the semiconductor-contact

junction. Equation 2-3 describes the specific contact resistance for this type of transport:


P = exp hND (2.3)


The relevant terms to consider in this equation are the barrier height and the doping density, ND

or NA for p-type semiconductors. It is obvious that in order to achieve a low contact resistance

the barrier height must be minimized and the doping density must be maximized. Barrier height

is determined by material choice. Doping density can be determined during growth of the

material, with ion implantation, or use of contacts which can increase doping through

interactions with the semiconductor.

2.3.1.1 Ohmic contacts to p-GaN

High quality, low resistance Ohmic contacts to p-GaN are difficult to achieve. This is due

to several factors: the relative dearth of appropriate metal systems due to the large work function

of GaN, low hole densities due to the difficulty in achieving a high doping concentration during

growth, the presence of compensating hydrogen interstitials, and the presence of compensating









nitrogen vacancies. Despite this, several materials systems have been found to produce Ohmic

behavior. These are generally based on a high work function metal such as Ni, Pd, or Pt with an

overlayer of Au and have been found to achieve specific contact resistances in the range of 1 x

10-4 Q-cm2 for hole concentrations of 1017 cm-3 [44-48]. Ni/Au contacts are the most common.

The Ni serves to remove hydrogen from the near surface region of the p-GaN at elevated

annealing temperatures, as the Ni hydride is thermodynamically favorable. In addition, at these

elevated temperatures Au begins to diffuse in towards the GaN and it has been found that the Au

can form intermediate phases with the Ga in the form of AuGa and AuGa2 [49]. This can serve

to further increase the near surface doping through the formation of Ga vacancies in the GaN

which act as acceptors. However, at these anneal temperatures of> 500C, the Au and Ni may

completely exchange places at the GaN surface [44]. This leads to a severe roughening of the

contact surface and degraded performance and limits the temperature at which anneals may take

place as well as the lifetime of the contacts when operated at high temperature or, equivalently,

high powers. Because one of the most attractive features of GaN is its ability to operate reliably

under these conditions, it is important to find a contact system that will perform better.

2.3.1.1 Ohmic contacts to n-GaN

Ohmic contacts to n-GaN are more studied, better understood, and easier to achieve

compared to those to p-GaN. The most common materials system used is GaN//Ti/Al/X/Au,

where X is a material chosen to act as a diffusion barrier to separate the Al and Au. This is

necessary, as Al and Au will react to form A1Au4 [50]. This phase is viscous at low temperatures

and lateral flow can create problems, especially as the gate/source separation becomes small.

Because of this, there is great interest in the use of high temperature metals. The Ti is present at

the junction in order to form TiN through reaction with the GaN, which creates nitrogen









vacancies and thus increases the donor density. The Au overlayer is present in order to prevent

oxidation as well as to promote current spreading in the contact.

2.3.2 Schottky Contacts

The other type of electrical contact required for electronic devices are known as Schottky

contacts or diodes. Schottky contacts allow little or no current to flow through them until a

critical voltage is reached, above which large amounts of current flow. This is called the forward

biased region. In the reverse bias region, voltage is applied in the opposite direction and no

current flows initially. However, if the voltage is pushed high enough then breakdown occurs, a

huge amount of current flows through the contact and the contact is generally destroyed.

Ideal Schottky contacts are formed in the absence of any surface states. When a metal and

semiconductor are brought together, the Fermi levels must line up. In order for this to occur, a

charge exchange occurs between the two materials. When this occurs, a space charge or

depletion region is formed and a barrier to current flow is created. For p-type semiconductors,

the flat band diagram is shown in Figure 2-4.

The barrier height, 4bp, for a p-type semiconductor is given by the equation

q4bp = Eg -q (m -X) (2.4)

where Eg is the band gap, 4m is the metal work function, and x is the electron affinity of the

semiconductor. Invariably, however, surface states are present. This makes the prediction of

barrier height using the equation above inaccurate. In order to determine barrier height in this

case, it is necessary to fabricate a specimen to test.

The predominant carrier in Schottky contacts is the majority carrier unlike a pn junction

diode. Two mechanisms for current flow exist: thermionic emission (TE) and thermionic field

emission (TFE) or tunneling. While some authors have reported values for barrier heights on p-









GaN using the TE model, most of the work has demonstrated that TFE is the dominant transport

mechanism. Reported barrier heights on p-GaN range from -0.5 to 2.9 eV even for the same

metallization scheme because the mechanism of current flow is not definitively established [51-

65].

In the case of TE, electrons or holes are excited over the potential barrier and the forward

current is given by the following expression:

2 e, eV
JF = AT2exp exp e 1(2.5)
SkT) nkkT)

where A*=103.8 Acm-2K-2 is the effective Richardson's constant for p-GaN, Tis the absolute

temperature, e is the electronic charge, OB is the SBH, ks is the Boltzmann's constant, and Vis

the applied voltage.

In the case of TFE, quantum tunneling of charge carriers through the barrier occurs and the

forward current can be described by


qJ e
JF = oexp qV (2.6)
Eo)

where the saturation current density Jo is given by

A'T \Eo q (] -V- 05 (_ q _
J = k cosh(Eoo/kBT) exp kT kT 7)


In Eq. (2.7), -(E-Ev)/q is the difference between the valence band maximum and the

position of the Fermi level and E0 = E00 coth (Eo00/kT) is the characteristic energy related to the

tunneling probability.









2.4 Experiments

The goal of this work is to find suitable contact systems for use at high temperatures on n-

and p-GaN. Towards this end, the previously discussed materials will be investigated in order to

determine which of them allow good quality Ohmic and Schottky contacts. In addition, a study

will be made of devices which use p-GaN, mainly LEDs. The contacts will be aged at high

temperatures and compared with more conventional schemes. A variety of techniques will be

used to characterize the contacts and devices in order to determine their performance and to

better understand what is happening.

The main experiments that will be done are the processing of the Ohmic and Schottky

contacts for electrical characterization. The following schemes will be produced and analyzed:

1. GaN//Ni/Au/X/Ti/Au for p-type Ohmic contacts, where X indicates the refractory material to
be examined

2. GaN//X/Ti/Au for p-type Ohmic contacts

3. GaN//Ni/Ir/Au for p-type Ohmic contacts

4. GaN//Ti/Au/X/Ti/Au for n-type Ohmic contacts

5. GaN//X/Pt/Au for p-type Schottky contacts

In case (1), the Ni/Au should act as an Ohmic contact while the boride layer provides a

diffusion barrier to prevent breakdown of the metallization at elevated temperatures. The Ti

layer is present in order to promote adhesion of the Au to the rest of the contact. The Au

overlayer is to prevent oxidation and act as a current spreading layer. In case (2), it is hoped that

at least some of the borides may provide an Ohmic contact on p-GaN. Case (3) may provide

some insight into the importance of Au at the GaN surface for these types of contacts. Case (4)

is based on standard Ohmic contacts to n-GaN, with a nitride used as the diffusion barrier. Case









(5) will be used to fabricate Schottky contacts, as the Pt should not be as reactive as Ti should it

reaches the GaN surface and is therefore used as the adhesion layer.

Samples will be prepared on p-GaN and n-GaN wafers grown in an MOCVD reactor. P-

GaN will be 3 |tm thick Mg-doped and grown on c-plane A1203. Hole concentration after

activation obtained from Hall measurements is -1 x 17 cm-3. The same doping will be used for

both Ohmic and Schottky contacts. This is because it is extremely difficult to obtain a higher

doping level for use in Ohmic contacts. At lower doping concentrations, it is not possible to

obtain an Ohmic contact for use in Schottky contacts. For Ohmic contacts to n-GaN, epilayers 3

|tm thick Si-doped and grown on c-plane A1203 with an electron concentration of-1 x 1017 cm-3

will be used.

In order to fabricate the samples, photolithography will be performed using a Karl Suss

MJB3 aligner using positive photoresists. Samples for Ohmic contacts will then be etched in a

Unaxis ICP tool in order to isolate devices and prevent current spreading. Prior to patterning for

the metal layers, p-GaN samples will be dipped in a 1M KOH solution for one minute in order to

remove surface oxide. Following patterning, surface treatments may be necessary to improve

performance. A 30 s RIE 02 plasma exposure to remove carbon followed by a 10% HC1 dip for

one to ten minutes immediately prior to insertion into the deposition chamber will be employed

to ensure a clean surface. Initial Ni/Au layers for contacts to p-GaN will be deposited using an

e-beam evaporator in order to minimize surface damage. For deposition of all other layers, a

Kurt Lesker sputterer will be employed. This is necessary, as the refractory materials cannot be

evaporated. In addition, it is necessary to deposit all overlayers in the sputterer in order to

minimize oxidation. Following lift off of the metal layers, rapid thermal annealing will be









performed at a range of temperatures from 300-10000C in order to determine how the properties

of the contacts are affected at a variety of annealing temperatures.

Following the development of these contacts, devices will be fabricated. While p-GaN is

unsuitable for use in HEMTs due to its low hole mobility, it is necessary for use in LEDs, LDs,

and HBTs. LEDs will be fabricated on wafers with the structure shown in Figure 2-5. These

will be compared to devices using more traditional metal schemes such as Ni/Au. They will then

be thermally aged in order to determine the performance relative to Ni/Au when maintained at an

elevated temperature.

2.5 Characterization Techniques

A variety of techniques will be utilized in order to determine device characteristics as well

as to observe the physical and chemical structure of the contacts. The techniques to be employed

are

1. Current-voltage measurements

2. Capacitance-voltage measurements

3. X-ray Photoelectron Spectroscopy

4. Auger Electron Spectroscopy

5. Scanning Electron Microscopy

2.5.1 Current-Voltage

Current-voltage (IV) measurements will be performed using an Agilent 4156

Semiconductor Parameter Analyzer. Two probe measurements will be performed when possible.

However, if problems which make it impossible to extract contact resistance arise due to high

sheet resistance and the resistance of the probes, a four probe measurement technique will be

employed. Using this method, current will be injected with two outer probes while the voltage

drop is measured between two inner probes. This allows for measurement of only the voltage









drop between pads. IV measurements will be the main technique employed to determine the

suitability of both Ohmic and Schottky contacts. The IV curves allow for the extraction of

contact resistance for the former and barrier height and reverse breakdown voltage for the latter.

IV measurements can be performed using either a circular or linear transmission line method.

The CTLM is useful for materials which are difficult to etch and therefore may demonstrate

problems with current spreading. In this study, linear TLM patterns will be used, as shown in

Figure 2-6. Data from these curves allow for the creation of a plot such as that shown in Figure

2-7. Characterization of the pads is then performed using the relation


R, =2Rc + Rs


Pc =RW2) (2.8a,b)



Figure 2-8 shows a schematic for Schottky contacts. Schottky pads are formed using an

inner Schottky contact with an outer Ohmic contact.

2.5.2 Capacitance-Voltage

Capacitance-voltage (CV) measurements will be employed on Schottky contacts in order

to further confirm results of IV measurements as well as to provide a deeper understanding of the

barrier. If an interstitial layer exists between the contact and the p-GaN, utilizing a CV

measurement can reveal this. This is a distinct possibility that could arise either from an

interstitial oxide layer forming or from defects at the surface possibly caused by sputtering

damage. Measurements were performed using an Agilent 4284A precision LCR meter in the

parallel mode. The modulation frequency was set to 1 kHz. Barrier height is determined using









B = V,, + EA + (k,T/q)lng (2.9)


where V,nt is the extrapolated intercept voltage of the reverse bias in the 1/C2 versus Vplot,

EAO0. 12 eV is the activation energy of Mg dopants in p-GaN, and g-2 is the degeneracy factor

for acceptors. Note that in our case, the series resistance (-2 kM) and the junction conductance

(-10-8-10-9) were low enough so that the measured capacitance corresponds to the proper

junction capacitance.

2.5.3 X-ray Photoelectron Spectroscopy

X-ray Photoelectron Spectroscopy (XPS) measurements will be further used to confirm the

barrier height of Schottky contacts. XPS functions by bombarding a specimen with soft x-ray

radiation to examine core energy levels. When an x-ray collides with an atom, an electron is

ejected from its shell. The electron is then collected and its energy measured. Each element has

a characteristic binding energy for its core electrons, and thus each element has a characteristic

spectrum. Comparison of the spectrums of a bare specimen, such as GaN, and a specimen with a

thin film coating, such as GaN//W2B, can yield the barrier height of the surface. The barrier

height can be determined from the binding energy of the Ga 3d core level EB and the energy

difference between that core level and the valence band maximum Evc according to



B = EB Ev (63,64)(2.10)


2.5.4 Auger Electron Spectroscopy

Auger Electron Spectroscopy (AES) will be used to determine the elemental composition

versus depth of the contacts. This will allow for an understanding of how diffusion plays a role

in the behavior of the contacts. AES functions by bombarding the surface of a specimen with a









beam of focused electrons. These electrons penetrate into the sample and collide with the atoms

of the specimen. A core shell electron is then ejected, creating a vacancy. Following this, an

electron from an outer shell relaxes and fills this lower energy state, releasing energy. This

excess energy then results in the ejection of an electron from an outer shell. This Auger electron

possesses an energy characteristic of the element from which it is ejected, allowing for the

composition of the specimen to be determined. Using an ion gun in combination with the

electron beam allows for information about the composition versus depth to be collected.









Table 2-1: Bulk GaN properties


Property
Lattice parameters at 300 K (nm)

Density (g-cm-3)
Stable phase at 300K
Melting point (C)
Thermal conductivity (W/cm-K)
Linear thermal expansion coefficient

Static dielectric constant
Refractive index
Energy bandgap (eV)
Exciton binding energy (meV)
Electron effective mass
Hole effective mass


Value
ao = 0.3189 nm
Co = 0.5185 nm
6.095 g-cm-3
Wurtzite
2500
1.3, 2.2 +/- 0.2 for thick, free-standing GaN
Along ao = 5.59x10-6 K-
Along co = 7.75x10-6 K-
8.9
2.67 at 3.38 eV
Direct, 3.45
26
0.20
0.59









Table 2-2: Properties of common semiconductors


Bandgap
(eV)
@300C
Electron
mobility
(cm2/V-s),
RT
Hole
mobility
(cm2/V-s),
RT
Saturation
velocity
(cm/s), 107
Breakdown
field (V/cm)
x 106
Thermal
conductivity
(W/cm)
Melting
temperature
(K)


Si
1.1
indirect

1400



600



1


0.3


1.5


1690


GaAs
1.4
direct

8500



400



2


0.4


0.5


1510


GaN
3.4
direct

1000 (bulk)
2000
(2DEG)

30



2.5


>5


1.5


>1700


A1N
6.2
direct

135



14



1.4


3000


6H-SiC
2.9
indirect

600



40



2


4


>2100












Table 2-3: Properties of the borides


Properties
Melting Point
(C)
Structure
Thermal
expansion
coefficients x
106 (/deg)
Phonon
component of
heat
conduction
(wt/m-deg)
Elastic
modulus E x
10-6 (kg/cm2)
Characteristic
Temperature
(K)
Density of
electronic
state g x 10-21
(eV1 cm1)
Work
function (eV)
Heat of
Formation
(kcal/mol)
Lattice
constant (A)
Thermal
conductivity
(W/m-K)
Electrical
resistivity
(uOhm-cm)


TiB2
2980 -3225

Hexagonal
4/6


20.6


ZrB2
3040 -3200

Hexagonal
5.9


W2B
-2670


18.9


1100


4.76


4.19(?)


3.94(?)


71.4


3.028


3.169


Unknown 32


W2B5
-2385


Hexagonal


CrB2
2200

Hexagonal
10.5


10.4


54.6


3.18(?)


2.982


2.969











Table 2-4: Properties of the nitrides
Properties TaN TiN ZrN
Melting point (K) 3633 3203 3253

Structure FCC FCC FCC
Thermal expansion 3.6 9.4 8.7
coefficients x 106 (/deg)
Work function (eV) 4.7 3.74 4.6
Lattice constant (A) 4.3 4.32 4.574
Thermal conductivity (W 57.5 19.2 20
m-1 K-1)
Electrical resistivity (tLQ- 15.52 25 13.6
cm)









Table 2-5: Properties of Ir
Properties Ir
Melting point (K) 2719
Structure FCC
Thermal expansion 6.4
coefficients x 106 (/deg)
Work function (eV) 5.27
Lattice constant (A) 3.8
Thermal conductivity (W 147
m-1 K-1)
Electrical resistivity (LtQ- 4.71
cm)















Temperature (C)


400 300 200


1.5


2.0


2.5


Figure 2-1 Intrinsic carrier concentration of GaN, GaAs, and Si


10201


100


25 0


1010




100


f?
u

o


om
u



ci


I I I


Si


GaAs -


GaN


10-20


3.0


3.5


1000 / T (K1)


4.0


' '


I
I


.










Ga-face


Ga N

N Ga










Substrate Substrate

Figure 2-2: Stable wurtzite GaN crystal


N-face











Metal p-Semiconductor


Figure 2-3: Flat band diagram for a p-type Ohmic contact











p-Semiconductor


Figure 2-4: Flat band diagram for a p-type Schottky contact


Metal


















- p-A1GaN


- n-A1GaN


p-contact
p-GaN
<- p-A1GaN
InGaN/GaN MQW
-A1GaN


n-GaN


GaN buffer


Sapphire


Figure 2-5: LED cross section (a) before and (b) after processing


p-GaN


InGaN/GaN MQW


n-GaN

GaN buffer


Sapphire


n-contact











Metal Pads


L2 m3
Semiconductor film


Figure 2-6: Linear transmission line pattern


















Slope= Rs/W


L, L2
Distance L


Figure 2-7: Resistance vs. pad spacing plot









Schottky contact


Ohmic contact
/


Semiconductor film


Figure 2-8: Schottky contact schematic









CHAPTER 3
THERMALLY STABLE OHMIC CONTACTS TO P-GAN

3.1 Ohmic Contacts

In this section, two approaches to fabricating Ohmic contacts using the boride family of

materials will be discussed. The first is as an overlayer on conventional Ni/Au contacts and the

second as the actual contact material. Because these borides have a large work function, it is

expected that both approaches will be successful in yielding Ohmic behavior, provided that

sputter damage and oxygen incorporation into the borides during deposition does not

compromise the contacts, especially the latter.

3.1.1 Fabrication of Ohmic Contacts

The samples used were light-emitting diode wafers with a 0.1 um Mg-doped p-GaN layer

on top of a 0.1 um p-AlGaN layer followed by a 0.3 um InGaN/GaN superlattice on top of a n-

type GaN layer grown on 2 um-thick undoped buffers on c-plane A1203 substrates. Hole

concentrations in the p-GaN layer were 1017 cm-3 obtained from Hall measurements. Mesas for

the linear transmission line method (LTLM) method were etched using a C12/Ar inductively

coupled plasma for pad isolation and to minimize current spreading. Samples were dipped in 1

M KOH for 1 minute and then rinsed in acetone and ethanol prior to photolithography. Prior to

insertion into the evaporation chamber, the samples were exposed to a 30 s 02 RIE plasma in

order to remove any residual carbon and then dipped in 10:1 H20:HC1 for 5 minutes, which

serves to remove any native oxide. A Ni/Au contact of 500 A/800 A was deposited using an e-

beam evaporator in order to minimize surface damage. After unloading, the samples were put

into a sputtering system for deposition of the X/Ti/Au layer of thickness 500 A/200 A/800 A,

where X is the refractory metal used. Au is necessary to decrease the sheet resistance while the

Ti is necessary to prevent peeling of the Au from the surface due to the large lattice mismatch.









In the case of Ir, this Ti layer is unnecessary. Contacts were also fabricated of the form Ni/Ir/Au,

in order to confirm the importance of the Au near the GaN surface, and of the form X/Ti/Au to

investigate the use of the refractory materials without the increased doping present due to the

Ni/Au layer. Sputtering was performed using an Ar plasma-assisted rf sputter at 5 mTorr and rf

(13.56 IMHz) power of 150 W. The sputter rates were 5 A/s for Au, 0.5 A/s for Ti, and were 0.8-

2 A/s for the refractory materials. Following lift-off, the contacts were annealed in a flowing N2

ambient for 60 s in a rapid thermal annealing furnace at temperatures up to 1000 oC.

3.1.2 Nitride-Based Contacts

3.1.2.1 Experiment and discussion

Ohmic contacts based on Ni/Au but with a nitride overlayer were fabricated in order to

improve the temperature stability. Figure 3-1 shows the specific contact resistivity of the nitride-

based contacts as a function of annealing temperature, as well as the sheet resistance of the p-

GaN under the contacts. At annealing temperatures of less than 500 C, contacts were rectifying.

The contacts transition to Ohmic behavior at higher temperatures. The specific contact

resistance was not a function of temperature in the range 25-175C, indicating that tunneling is

the dominant transport mechanism. The specific contact resistance is fairly stable on annealing

up to 1000 oC, with a minimum of 2 x 10-4 Q-cm2 at 6000C for the Ni/ Au/ ZrN/ Ti / Au

contacts. This indicates that there is minimal change in the doping of the near surface region of

the GaN at these elevated anneal temperatures. The formation of AuGa and AuGa2 occurs at

600C at the interface 49]. A sharp maximum in the sheet resistance is observed at 800 oC, with a

drop at higher temperatures. This is likely due to increased in-diffusion of metal to the GaN

without complete alloying, resulting in interstitial atom incorporation into the lattice. The drop

in sheet resistance at higher temperatures is an indication that these interstitials are increasingly

incorporated into the GaN. The magnitude of the increase in sheet resistance is related to the









size of the metallic atom in the nitride, with the largest increase seen with the smallest atom Ti

and vice versa for Ta, suggesting the atoms causing the spike are from the nitride. The reason

the smaller atoms cause a more pronounced increase is the greater ease of diffusion into the GaN

as well as a likely higher solid solubility.

Figure 3-2 shows the AES depth profiles for the TaN-based contacts under a variety of

conditions. Upon annealing at intermediate temperatures, most of the Ti has migrated to the

surface and oxidized while Ni and Au are both present at the GaN interface. In addition, the

nitride layer is present at the GaN surface, suggesting that the increase in sheet resistance is due

to incorporation of the metallic atoms from the nitride into the GaN. After annealing at 700 C

and aging at 200 C or annealing at 1000 C, there is a large amount of intermixing between the

metals and the GaN. Note that the upon annealing at 6000C the Ta and N peaks are still

correlated, indicating that TaN has not reacted a great deal with the GaN yet. However, in the

profile of the thermally aged contact there is no longer a correlation between the peaks of the Ta

and N. This is an indication that the nitride is breaking down during the thermal aging. In fact,

the profile of the aged contacts bears a much stronger resemblance to that of those annealed at

much higher temperatures. SEM micrographs are shown in Figure 3-3 for the TaN-based

contacts. The contacts show a smooth morphology upon deposition, with an increasingly rough

surface as anneal temperature increases due to the increased intermixing of the contacts as well

as oxidation of the surface. TiN and ZrN contacts display the same type of profile structure upon

annealing and after thermal aging.

As mentioned previously, the Ni/Au/GaN structure is preferred due to the greater

difference in electronegativities between Au and Ga than between Ni and Ga. By incorporating a

layer on top of the Au, the dynamics of this relationship may change. The electro-negativities of









Ta, Ti, and Zr are 1.5, 1.5, and 1.33, respectively. The values for Au, Ni, and Ga are 2.4, 1.8,

and 1.6, respectively. From this, it is apparent that the differences between Au and either Ti, Ta,

Zr, or Ga are roughly the same, indicating no preferred bonding to any of them. Various

equilibrium phases exist for Au-Ti, Au-Zr, and Au-Ta, but none for Au-N, supporting the

conjecture that Au is bonded to the metal. Thus, neither configuration is strongly preferred.

From the Au-Ni phase diagram, a miscibility gap exists with a maximum of 810.3 C at around

70 at% Au. At higher and lower Au concentrations, the temperature above which Au and Ni

intermix freely is lower; for example, at 10 at% Au, it is below 500 C and at 30 at% Au below

700 C. Since there is no driving force in terms of greater bond strength to cause Au and Ni to

switch places when using these diffusion barriers, the Ni and Au should each be present at the

surface of the GaN. It is expected that the Ni and Au would separate with slow enough cooling

but the layered structure of pure Ni/Au contacts may not form. The AES depth profiles confirm

this, displaying significant intermixing at the GaN interface. We also see that the boundary

between the GaN and the Ni-Au is not well-defined. This is consistent with the formation of the

Au-Ga phases. From the Au-Ga phase diagram, these phases have melting points of 461.3 C

and 491.3 C, respectively. These are lower than the temperatures required to obtain Ohmic

behavior, suggesting it is beneficial to have some Ni present at the GaN interface in order to

maintain contact integrity. [67]

Figure 3-4 shows specific contact resistance and sheet resistance under the contacts as a

function of thermal aging at 200 C. This simulates long term operation of these contacts in

devices. These contacts display poor stability upon aging at this temperature, lasting for nearly

10 days before the I-V characteristics are no longer completely linear. From the AES depth

profiles of the contacts that failed, there was significant in-diffusion of metals into the GaN. The









temperature during aging is not sufficient for the formation of the Au-Ga phases and as a result

the metal in-diffused to the GaN is in the form of interstitials, which hinder electrical conduction.

In addition to this, the presence of the nitride on the surface of the GaN will lead to a breakdown

the near surface GaN layer. Failure is attributed to the fact that the nitrides will intermix with the

GaN. Since no change in chemical composition is expected to take place, only for Ga to

exchange places with a metal such as Ti, there is no contribution to a change in the Gibb's free

energy from formation or decomposition of compounds. In fact, it is expected that the

intermixing may be favorable, as it will serve to increase the entropy of the system. From the

equation AG = AH TAS it is apparent that an increase in entropy leads to a lower free energy,

and is therefore favorable. This leads to decomposition of the GaN. This is in agreement with

the sheet resistance under the contacts increasing in the failed contacts. It is also consistent with

the finding in the AES profiles that the aged contacts display a breakdown in the nitride barrier

layer.

3.1.2.2 Summary

In conclusion, nitride-based contacts display low specific contact resistance and are stable

upon anneals up to 1000 C, with reasonable stability upon aging at elevated temperatures. Due

to the nature of the contacts, they display extremely poor stability during aging and are therefore

unsuitable for long term device operation.

3.1.3 Tungsten Boride and Chromium Boride-Based Contacts and Long Term Thermal
Aging of Borides

3.1.3.1 Experiment and discussion

Figure 3-5 displays the specific contact resistance of the contacts as a function of anneal

temperature as well as the sheet resistance under the contact. The contacts transition to Ohmic

behavior at annealing temperatures of > 600C. This likely corresponds to the creation of Ga









vacancies due to the formation of AuGa and AuGa2 at the GaN surface and thus a localized

increase in the hole concentration. This implies that the current transport mechanism is

tunneling. Figure 3-6 supports this conclusion, as the contact resistance is not dependent on

measurement temperature in the measured range 25-140 C. The sheet resistance is also

relatively stable as a function of annealing temperature, suggesting that there is little intermixing

between the boride and the GaN, even during elevated anneal temperatures.

Figure 3-7 displays the specific contact resistance of the W2B and CrB2 contacts as a

function of long term thermal aging at 350C. Both W2B and CrB2 display excellent stability for

the duration of the aging, 23 days. While the nitrides provided good stability over the same

range of anneal temperatures, the contacts failed after approximately 9 days at a lower

temperature. This is attributed to the fact that the nitrides will intermix with the GaN. In the case

of the borides, however, changes in the compounds must occur for significant intermixing to take

place. This could take several forms, including the reactions XB + GaN => XN + GaB or XB +

GaN => XGa + BN. GaB phases are not present in phase diagrams [51]. Full thermodynamic

evaluation of these is not possible, as most of the values for the Gibb's free energy are not

available. However, it would be expected that these reactions could only proceed at very

elevated temperatures, as the compounds involved are each very stable while any Ga-metal

phases have relatively low melting temperatures and thus can be assumed to be less stable [51].

For example, the maximum melting temperatures of selected Ga-metal phases are 1670C for

GaTi, 1863C for GaCr, and 1855C for GaZr. These are much lower than the melting

temperatures of either GaN or the borides or nitrides. Thus, while a reaction of this nature may

proceed during very high temperature anneals, during the thermal aging it will only proceed very

slowly, if at all.









Figure 3-8 displays the AES depth profiles for the W2B-based contacts as deposited,

annealed at 600 oC, annealed at 700 C and aged at 350 oC, and annealed at 1000 C. Upon

annealing at 600 oC, the Ni and Au show strong mixing at the GaN interface. However, the

boride barrier prevents most intermixing between the Ni/Au and the Ti/Au overlayer. The boride

has minimal reaction with the GaN surface and the surface had excellent morphology at 6000C.

Note that the boride peak remains strongly correlated, indicating the W and B atoms remain

together. This is the case even at 10000C anneals and after long term thermal aging at 350 C,

indicating that the boride is not chemically active in these contacts. At higher temperatures, the

AES profiles were not entirely reliable due to non-uniformity depending on where the profile

was taken. However, the IV characteristics were still Ohmic and the surface still had reasonable

morphology. At temperatures of 10000C and higher, the contact layers became totally

intermixed, although the peaks of W and B still correspond to one another. Similar results were

obtained with CrB2-based contacts.

3.1.3.2 Summary

These boride-based contacts display excellent thermal stability both upon high temperature

anneals and during long term thermal aging at very high temperature. This is due to the fact that

there is little intermixing between the boride and the GaN itself, due to the large change in Gibbs

Free Energy that would be necessary for any chemical reaction to occur. This is supported by

the depth profiles obtained from AES, which show that the boride has not broken down during

experiments.

3.1.4 Contact Resistance for Other Boride-based Contacts

3.1.4.1 Titanium boride-based contacts

Contacts with a TiB2 diffusion barrier for enhancing the thermal stability of Ni/Au Ohmic

contacts on p-GaN were fabricated. Figure 3-9 shows the specific contact resistivity of the









Ni/Au/TiB2/Ti/Au/p-GaN structure as a function of annealing temperature, along with the sheet

resistance of the p-GaN under the contacts, extracted from the TLM measurements. The contacts

are rectifying below an anneal temperature of 5000C but transition to Ohmic behavior for higher

temperatures. The specific contact resistivity improves steadily with temperature above 700C

and reaches a minimum of 2 x10-4 Q.cm-2 after annealing at 800 OC for 60 secs. Note that the

contact resistance stays at a similar value even for temperatures up to 9500C, even though the

higher annealing temperatures produced increases in the sheet resistance of the GaN and less

reproducible contact properties due to roughening of the contact morphology. The mechanism

for the increase in sheet resistance may include loss of N (producing nitrogen vacancy-related

donors) or a reduction in the soluble Mg concentration. This specific contact resistance is

comparable to that achieved on the same samples with Ni/Au metallization annealed at 5000C.

The incorporation of the TiB2 diffusion barrier clearly provides a much wider range of thermal

stability compared to the standard Ni/Au contacts.

Figure 3-10 shows SEM pictures of the TLM contact pads as a function of annealing

temperature. After 8000C annealing the morphology is degraded, but is still much smoother than

conventional Ni/Au contacts under these conditions. The darker appearance of the contacts after

high temperature anneals is mainly a result of the outdiffusion of Ti, which rapidly oxidizes.

Figure 3-11 shows the AES surface scans from the as-deposited sample (top) and those

annealed at 8000C (center) and 9000C (bottom). Ti is evident on the surface after annealing at

8000C and the concentration increases with annealing temperature. The increased oxygen

concentration evident from the summary of the data in Table 3-1 most likely comes from

oxidation of this out-diffused Ti. The carbon signal in all cases comes from adventitious carbon

on the surface.









The AES depth profiles from the as-deposited and 800 and 9000C annealed samples are

shown in Figure 3-12. The as-deposited sample shows sharp interfaces between the metals and

between the Ni and the GaN. After annealing at 8000C, the Ni shows significant movement

through all of the overlying layers and by 950C, most of the contact metallurgy is intermixed

and the Ti is essentially all removed to the surface. There is an accompanying decrease in

abruptness of the TiB2 interfaces with the metals on either side. The TiB2 appears to be a barrier

for Ti diffusion, so this probably excludes formation of TiN phases at the interface of the GaN to

account for the improved contact resistance.

3.1.4.2 Zirconium Boride-based contacts

ZrB2-based contacts were also fabricated. Figure 3-13 shows the specific contact

resistance of the ZrB2/Ti/Au and Ni/Au/ZrB2/ Ti/Au/p-GaN structures as a function of annealing

temperature, along with the sheet resistance of the p-GaN under the contacts, extracted from the

TLM measurements. The contacts are rectifying below an anneal temperature of 750C but

transition to Ohmic behavior for higher temperatures. The Ni/Au/ZrB2/Ti/Au contact resistance

improves with higher anneal temperature, at the expense of poorer morphology. The ZrB2/Ti/Au

contacts do not show low resistance values until 8000C and the value degrades at higher

temperatures. The associated sheet resistance generally decreases with temperature. The specific

contact resistance values obtained with both types of ZrB2-based schemes is comparable to that

achieved on the same samples with Ni/Au metallization annealed at 5000C but the former have

much higher thermal stability, as expected from the high melting temperature of the ZrB2.

Figure 3-14 shows the AES surface scans and depth profiles from the as-deposited

Ni/Au/ZrB2/ Ti/Au/p-GaN sample (left) and from the sample annealed at 7500C (right). Ti, Ni

and Zr are evident on the surface after annealing at 7500C and the concentration increased with

annealing temperature. The as-deposited sample shows sharp interfaces between the metals and









between the Ni and the GaN. Table 3-2 summarizes the near-surface composition AES data. The

increased oxygen concentration evident from this data most likely comes from oxidation of the

out-diffused Ti. The carbon signal in all cases comes from adventitious carbon on the surface.

After annealing at 7500C, the Ni shows significant movement through all of the overlying layers

and even the Zr diffuses out of the boride later, leaving the boron in its original location. Both Zr

and Au diffuse to the interface with GaN. The increased contact intermixing at higher annealing

temperatures did roughen the contact morphology as shown in the SEM pictures of the TLM

contact pads as a function of annealing temperature in Figure 3-15. After 8000C annealing the

morphology is degraded, but is still much smoother than conventional Ni/Au contacts under

these conditions. The darker appearance of the contacts after high temperature anneals is mainly

a result of the outdiffusion of Ti, which then oxidizes.

The same basic trends were observed with the ZrB2/Ti/Au structures. Figure 3-16 shows

the AES surface scans and surface scans of ZrB2/Ti/Au Ohmic contacts on p-GaN as a function

of anneal temperature. The as-deposited sample shows abrupt interfaces but after annealing at

9000C, the Zr shows a very broad distribution (more so than the B). The contacts after annealing

showed the presence of reacted islands, as shown in the elemental maps of Figure 3-17. The

islands contain both Au and Ga and show that the GaN epi layer has begun to dissociate at this

temperature, at least under the contact metallurgy. Note that the areas with Au and Ga overlap,

indicating that they are indeed forming the Au-Ga phases reported in the literature.

3.1.4.3 Gallium Nitride//Tungsten Boride-based contacts

Figure 3-17 shows the specific contact resistivity of the W2B/Ti/Au/p-GaN structure as a

function of annealing temperature, along with the sheet resistance of the p-GaN under the

contacts, extracted from the TLM measurements. The contacts are rectifying below an anneal

temperature of 5000C but transition to Ohmic behavior at this temperature. The specific contact









resistivity improves steadily with temperature and a minimum of 1.7 x10-3 Q.cm-2 was obtained

after annealing at 800 OC for 60 sees. Higher annealing temperatures produced sharp increases in

the sheet resistance of the GaN and irreproducible contact properties. The mechanism for the

increase in sheet resistance may include loss of N (producing nitrogen vacancy-related donors)

or a reduction in the soluble Mg concentration. However the contact morphology as determined

by SEM was similar over the entire annealing range used here. This specific contact resistance is

comparable to that achieved on the same samples with Ni/Au metallization annealed at 5000C.

Figure 3-18 shows the AES surface scans from the as-deposited sample and those

annealed at 500, 700 and 8000C.Ti is evident on the surface after annealing at 5000C and the

concentration increases with annealing temperature. The increased oxygen concentration evident

from the summary of the data in Table 3-3 most likely comes from oxidation of this out-diffused

Ti. The carbon signal in all cases comes from adventitious carbon on the surface.

The AES depth profiles from the as-deposited and annealed samples are shown in Figure

3-19. The as-deposited sample shows sharp interfaces between the metals and between the W2B

and the GaN. After annealing at 5000C, the Ti shows significant movement through the Au layer

and by 800C is essentially all removed to the surface. There is an accompanying decrease in

abruptness of the W2B/GaN interface, which suggests the onset of reactions that are responsible

for the improved contact resistance. Since the W2B appears to be a barrier for Ti diffusion, this

probably excludes formation of TiN phases. Cole et al.[29] reported the formation of the P-W2N

and W-N interfacial phases after annealing sputtered W on GaN [68]. The former appeared after

anneals at 6000C and the latter after 10000C. That work demonstrated the importance of these

interfacial phases in the resulting contact properties. The specific contact resistance obtained for

the W2B-based contacts is lower than reported previously for pure W on p-GaN [69], likely due









to the lack of these phases and the resulting compensation of holes by nitrogen vacancies. The

absence of Ti diffusion through the W2B also suggests that an optimum metal scheme might

involve a thin sandwich of the boride layers around the Ti, with the Au overlayer to reduce the

sheet resistance. The reflectivity of pure W is around 60% at 460 nm, while that of Au is around

95% [70], so clearly the thickness of the boride layers should be small relative to the Au

thickness. The long-term stability and reliability of the boride-based contacts must still be

established on LEDs and laser diodes. In the former case, the reflectance of the multi-layered

contacts at the emission wavelength of the LEDs must also be established.

3.1.5 Iridium-Based Contacts

3.1.5.1 Experiment and discussion

Figures 3-21 and 3-22 show the IV plots for both types of contacts at different anneal

temperatures and 20 pm spacing. Only the Ni/Au/Ir/Au contacts annealed at 500 C were found

to have Ohmic characteristics. It was observed that these contacts displayed a specific contact

resistance of -2.3 x 10-4 Q-cm2, comparable to results previously achieved on similar wafers for

other schemes. Table 3-4 shows a summary of previous results achieved using a similar

approach with other thermally stable materials as well as the anneal temperatures above which

the contacts failed either due to unreasonable morphology or breakdown of Ohmic

characteristics. From the IV curves for the Ni/Au/Ir/Au contacts, we see a decrease in the

amount of current upon annealing at 300 C, which likely corresponds to indiffusion into the

GaN of Ni but at a temperature too low for the Ni hydride to form and increase the doping. At

500 C the contacts transition to Ohmic behavior due to increased near surface doping. At higher

temperatures the contacts revert to rectifying behavior. For the Ni/Ir/Au contacts, the as

deposited display the largest amount of current. At 300 C the amount of current decreases









again, only to increase at 500 C due to removal of H by Ni. At 700 C it begins to decrease

again.

Figures 3-23 and 3-24 show the Auger electron spectroscopy (AES) depth profiles for the

Ni/Au/Ir/Au and Ni/Ir/Au contacts, respectively, at several different anneal temperatures. At

300C, the contacts maintain well defined layers, with a distinct lack of intermixing. At this

temperature, the contacts do not display Ohmic characteristics because the Ni has not

depassivated the H-Mg complexes. At 500 C, the Ni/Au/Ir/Au contacts display significant

interaction between the GaN and the Ni/Au underlayer. This likely corresponds to the formation

of the Au-Ga phases, which serve to increase the doping. The Ni/Ir/Au contacts show that the Ni

has diffused towards the Au, but the Au has not diffused in towards the GaN. With little or no

Au at the surface of the GaN, the hole concentration is not expected to increase as much as with

the Ni/Au-based contact. This is the reason why Ohmic behavior is not observed in the Ni/Ir/Au

contacts. In addition, we can observe that the Au seems to be essentially insoluble in the Ir at the

moderate temperature anneals. At 700 C, we can see that the layer structure has broken down

and the metals are completely intermixed.

Figure 3-25 displays scanning electron microscopy (SEM) images of the Ni/Au/Ir/Au

contacts as deposited and annealed at several temperatures. The surface at 500 C displays good

morphology while that of the 700 C anneal shows signs of the breakdown of the layer structure.

The failure of any contacts to display Ohmic behavior at high anneal temperatures is likely

due to the nature of the Ir. Ir is expected to be stable on the GaN surface because it is not

expected to form extensive secondary phases with the Ga or the N. While IrN has been recently

been observed, it has only been synthesized at very high pressures. For the high temperature

anneals, the Ir diffuses into the GaN. However, because no nitride can be formed under these









conditions, the Ir must occupy interstitial sites and cannot be incorporated into the crystal

structure of the GaN itself. Because the Ir atoms are very large and are likely sitting in

interstitial sites, they can cause significant lattice distortions and it is expected that the

crystallinity of the GaN would be significantly decreased. These distortions, as well as the Ir

atoms themselves, would act as scattering sites for charge carriers. It is in this manner that the

contacts fail when annealed at high temperatures. Thus, even when Au arrives at the surface in

the Ni/Ir/Au contacts at 700 C, the Ir has already caused enough disruption to the GaN that it

does not matter.

This same lack of chemical interaction of the Ir on the GaN may allow for improved long

term operation of LEDs using the Ni/Au/Ir/Au contacts. Because the Ir is not expected to

interact chemically with the GaN, at the lower temperatures which were examined during aging,

and which would be present during long term operation, the GaN itself is not expected to

decompose over a long period of time.

3.1.5.2 Summary

Ir diffusion barriers for contacts to p-GaN have been fabricated. The importance of Au

being in contact with the GaN for the formation of low resistance Ohmic contacts has been

confirmed, as the Ni/Ir/Au contacts do not display Ohmic characteristics.

3.2 Conclusions

Ohmic contacts have been fabricated to p-GaN using three classes of materials. It has been

determined that boride based contacts are superior to either Ir or nitride based ones due to both

their ability to withstand a wide range of annealing temperatures and to withstand long term

thermal aging. From these results, it can be noted that it is important to choose materials for

contacts which will not display a great deal of either chemical reaction or indiffusion.










Table 3-1: Concentration of elements detected on the as-received surface (in atom%)


Filename Sample ID C(1) 0(1) Ti(1) Au(3)
Sensitivztyfactors [0.076] [0.212] [0.188] [0.049]
051829101 As deposited 36 1 nd 64
051829201 8000C annealed 38 31 19 12
051829301 9000C anneal 27 33 28 12
051829401 950C Anneal A 30 26 33 nd
051829401 950C AnnealB 28 26 35 nd









Table 3-2: Concentration of elements detected on the as-received surfaces (in atom%)


Sample ID C(1) 0(1) Ti(1) Ni(1) Zr(2) Au(3)
Sensitivity factors
0.076 0.212 0.188 0.227 0.043 0.049
As deposited 44 4 nd nd nd 52
Ni/Au/ZrB2/Ti/Au
annealed 750C 45 19 2 1 14 19
annealed 800C 42 22 9 1 16 10
As deposited 37 4 nd nd nd 59
ZrB2/Ti/Au
annealed 800C 43 23 nd nd 12 22
annealed 1000C 48 19 nd nd 14 19










Table 3-3: Concentration of elements detected on the as-received surfaces (in atom%)


Sample C(1) 0(1) Ti(2) Au(3)
[0.076] [0.212] [0.296] [0.049]
As deposited 41 nd nd 59
5000C annealed 43 24 5 28
7000C annealed 44 34 7 15
8000C annealed 38 34 15 13









Table 3-4: Summary of specific contact resistances


Contact Scheme
Ni/Au/W2B/Ti/Au
W2B/Ti/Au
Ni/Au/CrB2/Ti/Au
Ni/Au/TiB2/Ti/Au
Ni/Au/ZrB2/Ti/Au
ZrB2/Ti/Au
Ni/Au/TaN/Ti/Au
Ni/Au/TiN/Ti/Au
Ni/Au/ZrN/Ti/Au
Ni/Au/Ir/Au
TaN/Ti/Au
ZrN/Ti/Au


Psc,minium (~-cm2)
3.5 x 10-5
1.69x 10-3
7x 10-5
1.93 x 10-4
1 x 10-4
1.8x 10-3
2.5 x 10-4
2.45 x 10-4
2x 10-4
2.3 x 10-4
3.7x 10-4
3.1x 10-4


Anneal temperature (C)
1000
800
1000
850
900
800
800
600
600
500
800
600


Failure temperature (C)
>1000
>900
>1000
>900
>900
>900
>1000
>1000
>1000
>500
>1000
>700














E x10" x -3 -- *100

\"; -80 a



S-*- TiN
-"--TaN -60 =

1x104 -* N *
ZrN u
-v- TaN "40 4)
-0- TiN
<- ZrN \
0 /20


S1105 ,--- 0
500 600 700 800 900 1000
CO Anneal Temperature (OC)


Figure 3-1: Specific contact resistance and sheet resistance under the contact ofNi/Au/ X/ Ti/Au
contacts as a function of anneal temperature.














1UU 100
Sg90 Au Ti Au Ni (a) 90 (b)
C 80- C 80-
o 0
70- 70-
S60- 60-
C Ta Ga N
8 50 50- Ga
C 40-N N C
UN Ti TaN
S30- 30-
E 20- E 20 Au
O 10 10- N
0 0
0 500 1000 1500 2000 2500 3000 35C 0 1000 2000 3000 40C
Depth (A) Depth (A)

100 100
90- (C) 90- (d)
C 80- C 80
o 0
S70- 70-
S60- 60-
S50- 50-
CC
S 040 N 40-
S30- Ga 30Ga
30 A .2a Au
E 20- E 20
O Ti Ta 0 Ta
10 10 -.
0 1000 2000 3000 40C 0 1000 2000 3000 40C
Depth (A) Depth (A)




Figure 3-2: Depth profiles ofNi/Au/TaN/Ti/Au contacts (a) as deposited (b) annealed at 600 C
(c) annealed at 700C and aged at 200 C until the contacts became non-Ohmic and
(d) annealed at 1000 C.





































Figure 3-3: Scanning electron microscopy images of Ni/Au/TaN/Ti/Au contacts (a) as deposited
(b) annealed at 600 C (c) annealed at 700C and aged at 200 C until the contacts
became non-Ohmic and (d) annealed at 1000 oC.













E ,X10_3 160
S1x10o-3
140

S/ 120

0-4 100 S
1x10- ,
S80 80
TaN 60 "
TiN -

C -v--TaN S
TiN
a ZrN CO

00 2 4 6 8 10

Om Day


Figure 3-4: Specific contact resistance of Ni/Au/X/Ti/Au contacts as a function of aging at 200
oC.








E

4)

Cu
0


0)

0.
C




e+
0


1x10-2


1x10-


1x10-4


1x10-5


1x106


600 700 800 900 1000


30


* CrB .
* W2B


CO Anneal Temperature (oC)




Figure 3-5: Specific contact resistance and sheet resistance under the contact versus anneal
temperature.


25 7
C
20 8

15

10 (

5 e
O
0










E 1x10-2 -
S! 7000C
1 3- -*- 1000C
x 0
o 1x103


x -4
re







U, Measurement Temperature (_C)
Figure 3-6 Specific contact resistance versus measurement temperature.
o

E 1x10
o 0 25 50 75 100 125 150
co Measurement Temperature (C)
Figure 3-6: Specific contact resistance versus measurement temperature.










E 1x10-3- 50


0 -40
o8
m 1x104 CrB )
30 a


S2- 20
1x10 5

0o 10Q

6
1x10-6 0 O
S0o 5 10 15 20 25
0.
CO Days



Figure 3-7: Specific contact resistance and sheet resistance under the contact as a function of
long term thermal aging at 3500C.

















uu i


= 80-

, 60-
w -

o 40-
U,)

E 20-
0
S -


100

S80

060
E 60
u
o 40
E 20
E 20
o


Au Ti Au (a)
B
Ni
Ga



W / I


1000 2000 3000 4000
Depth (A)


(b)
Au



N

T Ni

2000 4000
Depth (A)


100

" 80
0

60
U
o 40

0
E 20







100

S80-
0

S60-

o 40-
.u
E 20'
o
i


0 500 1000 1500 2000
Depth (A)



(b)


B

Ga
/Ni W N

Ti
S 1000 2000
Depth (A)


Figure 3-8: Depth profiles ofW2B-based contacts (a) as deposited (b) annealed at 600 C (c)
annealed at 700 C and aged at 3500C and (d) annealed at 1000 C


Au (b)
B



f\
W Ga
N
















I I I I I I'
0- -- Contact Resistance
E Sheet Resistance
o80
S1E-3

O 60 D
(D

40 (D
__ 1E-4 -


0 -20



0 I I I I
o -----------------

700 750 800 850 900 950

Temperature



Figure 3-9: Specific contact resistivity ofNi/Au/TiB2/Ti/Au Ohmic contacts and p-GaN sheet
resistance under the contact as a function of annealing temperature.























































Figure 3-10: Secondary electron images of Ni/Au/TiB2/Ti/Au contact pads on p-GaN as-
deposited (top) or after annealing at either 800(center) or 9000C (bottom).


'104





10/03/05 10.OkV 250.0 X 100.0m
5 SOOC annealed
T-











10/03/05 10.OkV 250.0 XI 100.Omm
5 900C annealed


































0


4000

3000

2000

1000
1 om

0

c/s
-1000

-2000

-3000

-4000

-5000

-6000



8000

6000

4000

2000

0

c/s -200


-400

-600


'F> Qj
Bi AiM


500 1000 1500 2000
Kiretic Energy (eV)













B AuAu





TII 27.8
Tl C1 27.2

B2 5.7


0



500 1000 1500 2000
Kinetic Energy (eV)


Figure 3-11: Surface scans of Ni/Au/TiB2/Ti/Au Ohmic contacts on p-GaN as a function of

anneal temperature. The as-deposited sample is at top, that annealed at 8000C at

center and that at 9000C at bottom .


50C 10o 1500
Kinetic Energy (eV)


-12000


x10


-1 5

















090 Au3 1s8

80


Au3 1s8


rm Concentration (%)


6q
Gal Is7
50

40

30 \

20

10


0 500 1000 1500 2000 2500 3000
Sputter Depth (A)


oncertrdton (%)

12 Is1

Ai3Js4 / s


Ga ls2


M ls2


500 1000
Sputter Depth (A)


Doncertraton (%)


50 -Gl 1s7

Au3 1s8


30 T2 ls5 N1 s3




01 IQ

0 500 1000 1500 2000
Sputter Depth (A)


Figure 3-12: Depth profiles of Ni/Au/TiB2/Ti/Au Ohmic contacts on p-GaN as a function of

anneal temperature. The as-deposited sample is at top, that annealed at 800 pC at

center, and that at 900 oC at bottom.


















0





,)


O
C


0)






0


7In
750


CD
11

10

9 D

8

7 C
0
6

5

4 3

3


900


Temperature (oC)


Figure 3-13: Specific contact resistance ofNi/Au/ZrB2/Ti/Au and ZrB2/Ti/Au Ohmic contacts
and p-GaN sheet resistance under the contact as a function of annealing temperature.


D
Contact Resistance \
- Ni/Au/ZrB /TiA/
SZrB2/Ti/Au

0^


Sheet Resistance
-0- Ni/Au/ZrB2/Ti/Au
- ZrB2/Ti/Au
0 0















2000 6000

S4000 Sample #2
S0Annealed at 750C
o 2000

C/S 2000 amle Au C/S 0 Au
c/s 2000
Sample #1 u Ti l i AlNi
As deposited Au Zr Au
-4000 -2000 Au
0 % Atomic %
Atomic % Au Cl 44.9
Au3 52.1 Af01 18.9
Auc : -4000
u Cl 44.2 -4000 Au3 15.7
-6000 l 2 Zr2 14.0
01 1.5 A12 2.9
-6000 Til 2.2
0 Nil 1.4
-8000 c
500 1000 1500 2000 -8000 500 1000 1500 2000
Kinetic Energy (eV) Kinetic Energy (eV)


100 100
90 90
S80 Ni 80
.2 70 .2 70
7Ga An
S60 Zr Ga Au
,50r ,60
50o \ 50/
40 0 40
*30 zr
S30 30 Zr
O O


0 / 20

0 500 1000 1500 2000 2500 3000 0 1000 2000 3000 4000 5000
Sputter Depth (A) Sputter Depth (A)





Figure 3-14: Surface scans and depth profiles ofNi/Au/TiB2/Ti/Au Ohmic contacts on p-GaN as
a function of anneal temperature.


4UUU


OUUU





















































Figure 3-15: Scanning electron microscopy images of Ni/Au/TiB2/Ti/Au contact pads on p-GaN
as-deposited (top) or after annealing at either 750 (center) or 8000C (bottom).













6000


500 1000 1500 2000
Kinetic Energy (eV)


100
90 Au Ga
80
70
60
50
40
30
20
10 0

0 500 1000 1500
Sputter Depth (A)


-10000



100
90
gso
-80
o70
S60
550
040
.o 30
E


10
0


Same #6
Annealed at


MgAI


Atomic %
Cl 52.9
01 16.3
Zr2 12.0
Au3 8.9
Mg2 4.2
AI2 4.0
N1 1.7


500 1000 1500 2000
Kinetic Energy (eV)


500 1000 1500 2000
Sputter Depth (A)


Figure 3-16: Surface scans and depth profiles of ZrB2/Ti/Au Ohmic contacts on p-GaN as a
function of anneal temperature.
















Gal


130400


Au3


0.0 0.0


47.0 191.0
Gal+Au3 SEM












0.0 151.0






Figure 3-17: Elemental maps obtained from scanning AES of ZrB2/Ti/Au Ohmic contacts pads
on p-GaN.


25900















< 3.6
E 140

3.2- C)

=C
0 120

( 2.8
c --Contact Resistance *
100
S-*- Sheet Resistance
.* 2.4 -

O8 -80
S2.0 .

C 60
0 1.6

500 550 600 650 700 750 800

Temperature (oC)



Figure 3-18: Specific contact resistivity ofW2B/Ti/Au Ohmic contacts and measured p-GaN
sheet resistance under the contact as a function of annealing temperature.












051535101 spe


Kinetic Energy (eV)

051535301 spe


500 1000 1500
Kinetic Energy (eV)


2000


05 k


Kinetic Energy(eV)


051535401 spe


500 1000 1500
Kinetic Energy (eV)


Figure 3-19: AES surface scans of W2B/Ti/Au Ohmic contacts on p-GaN as a function of anneal

temperature.


Sample #3
700C annealed
Surface survey







Atomic %
C 44 .0
01 33 .7
Au3 14.7
I T12 7.1
1 0.5




0


Sanmple #4
800C annealed
Surface survey

At oml %
C1 38.1
01 33.6
T12 15.1
o


2000


051535201 spe





















051535103 pro


o 70 70
Au3 18
S60 W2 1s6 60 W2 s6
Sa Gall s9

o 509 0 50
O N1 Is3

E 40 E 40
ClMiCl 30
30 < 30 s1

20 2001 1s4

10 10


0 500 1000 1500 2000 0 500 1000 1500 2000 2500
Sputter Depth (A) Sputter Depth (A)




051535303 pro 051535403 pro
100 100

90 90

o80 o

o 70
S70o 70 1s4

S60U W21s6 \ 60
6060
SGal Is8 Wo Gal Is8
o 50
o 50


4 B1 Is9 40 B





Cl Is B1


0 500 1000 1500 2000 2500 0
Sputter Depth( 0 500 1000 1500 2000 2500
Sputter Depth (A












Figure 3-20: Depth profiles of W2B/Ti/Au Ohmic contacts on p-GaN as a function of anneal

temperature.



















83


051535203













0.1
0.01
1E-3-
1E-4
1E-5
1E-6
As dep
0 1E-7- 300o
1E-8- 500c
1 E-9- 7000
1E-10 ,
-3 -2 -1 0 1 2 3
Voltage (V)




Figure 3-21: Current-voltage curves for Ni/Au/Ir/Au contacts.


















I-
I-
0l


0.01

1E-3

1E-4

1E-5

1E-6

1E-7


1E-8-

1E -9 ,
-3 -2 -1 0 1 2 3

Voltage (V)




Figure 3-22: Current-voltage curves for Ni/Ir/Au contacts
















Au Ir Au Ni (a)
? 80-
0
SGGa
S60-

o 40- N



o
0 1000 2000 3000
Depth (A)




100
A r (b)
Au
c 80-
0
Ga
C 60-
a Au
o 40- N

2 2Ni
E 20-
o
0
0 1000 2000 3000 4000 5000
Depth (A)




100-
F (c)
80-
o
C 60- Ir
a) /Ga
o 40- 0

E 20- Ni N
0

0 1000 2000 3000 4000
Depth (A)


Figure 3-23: Depth profiles for Ni/Au/Ir/Au contacts (a) annealed at 300 C (b) annealed at 500
C and (c) annealed at 700 C.













Au Ir Aa) Au (r)
S80- 80-
o 0 / \
s 60- 60- Ga
Ga Ni
o 400 o 40
o N o
E 20- E 20-
o o
01 0
0 1000 2000 0 1000 2000
Depth (A) Depth (A)



100
S(c)
80-
0
I-
*a 60-
= /G
o 40-
.2 Ni N
E 20
-, Au
0
0 1000 2000 3000
Depth (A)







Figure 3-24: Depth profiles for Ni/Ir/Au contacts (a) annealed at 300 C (b) annealed at 500 C
and (c) annealed at 700 oC.

















































Figure 3-25: Scanning electron microscopy images of Ni/Au/Ir/Au contacts


i -1
5/07/07 350.0'X 100.Omm
I Annealed at 300C













i -1
5/07/07 350.0 X 100.Omm
I Annealed at 700C


5/07/07 350.0 X 100.0mm
I Annealed at 500C









CHAPTER 4
OHMIC CONTACTS TO N-GAN

In this chapter, we report results on Ohmic contacts based on these materials to n-GaN.

Specific contact resistance has been studied as a function of annealing temperature, annealing

time, and thermal aging for TaN, TiN, and ZrN diffusion barriers. All nitrides show comparable

specific contact resistance to conventional Ti/Al/Pt/Au Ohmic contacts and display similar

stability when aged at 3500C in air.

4.1 Experiment

The samples consisted of 3 |tm thick Si-doped GaN grown by Metal Organic Chemical

Vapor Deposition on c-plane A1203 substrates. From Hall measurements, the electron

concentration was -3 x 107 cm3. Mesas 0.6 [m deep were formed by C12/Ar inductively

coupled plasma etching to provide for electrical isolation of the contact pads and to prevent

current spreading. Deposition of a Ti (200 A)/Al (800 A)/X (500 A)/Ti (200 A)/Au (800 A)

metallization was performed by sputtering, where X is TaN, TiN, or ZrN. Sputtering was

performed using a Ar plasma-assisted system at a pressure of 5 mTorr and rf (13.56 MHz)

powers of 150-350W. For comparison, a Ti (200 A)/Al (800 A)/Pt (400 A)/Au (800 A) was

deposited by e-beam evaporation. The contacts were patterned by lift off and annealed at 400-

10000C for 30-240s in a flowing N2 ambient in a rapid thermal annealing (RTA) furnace.

Specific contact resistances were obtained using the linear transmission line (L-TLM)

method. 100 |tm x 100 |tm pads with spacings of 80, 40, 20, 10, and 5 as well as 76, 36, 16, 6,

and 1 |tm were used. The total resistance is given by

RT= 2R + R,(L/W) (4.1)









where Rc is the contact resistance, Rs the sheet resistance, L the pad spacing, and W the pad

width. A plot of RT vs pad spacing provides both Rc, from the y-intercept, and Rs, from the

slope. The specific contact resistance is then calculated according to the equation

pc=Rc2 W/R,(4.2)

All current-voltage plots were measured using an Agilent 4145B Parameter Analyzer.

Auger electron spectroscopy (AES) depth profiling of the contacts was performed with a

Physical Electronics 660 scanning Auger microprobe. The electron beam conditions were 10

keV, 1 ptA beam current at 300 from the sample normal. For depth profiling, the ion beam

conditions were 3 keV Ar+, 2.0 ptA, (3 mm)2 raster. Before AES profiling, scanning electron

microscopy images (SEMs) were acquired from the samples. The SEMs were obtained at

magnifications of 125x and 1000x and were used to locate and document analysis area locations

and the surface morphology. Quantification of elements was achieved using the elemental

sensitivity factors.

4.2 Results and Discussion

Figure 4-1 shows the contact resistances as a function of anneal temperature for the nitride-

based Ohmic contacts as well as the conventional Ohmic contact. As deposited, the contacts

display slight curvature in the current-voltage curves. Upon annealing at 600-10000C, the IV

characteristics transform to Ohmic behavior. The specific contact resistances are stable up to

10000C anneals, except for the TiN, which shows increased specific contact resistivity at

10000C. The minimum specific contact resistance achieved was -6 x 10-5 Q.cm2 for the TiN-

based contacts at an anneal of 8000C. This is likely due to diffusion of additional Ti to the

surface to form the TiN phase. The ZrN-based contract appears to be the most stable with

increased annealing temperature, maintaining a fairly constant value of -1 x 104 for the range









600-9000C. These specific contact resistance values are expected for the moderately low doped

samples used in this study. This is because the current transport mechanism for Ti/Al-based

Ohmic contacts to n-GaN is tunneling and is governed by the relation


2 ',esnm O
RSCR exp h
(4.3)

Obviously, this relation shows a strong dependence on the doping of the n-GaN. For

doping levels an order of magnitude higher, 3 x 1018 cm-3, we would expect a decrease in the

contact resistance of approximately an order of magnitude.

Figure 4-2 shows the SEM images of the contacts as a function of anneal temperature. The

as-deposited samples for TiN and ZrN show a smooth morphology while bubbles are present on

the TaN as-deposited sample, possibly due to trapped sputter gases. At 600C and 800C anneals

the contacts maintain a reasonable morphology. Furthermore, there appears to be little

degradation of the surface even after aging at 350C. All contacts show good edge acuity, a

concern for devices with small contact separations.

Figures 4-3,4, and 5 show the AES depth profiles of the nitride-based contacts. As

deposited, all contacts contain a significant amount of oxygen incorporation in the Ti and nitride

layers. This is due to the relatively high base pressure of the sputtering system of 2 x 10-6 Torr.

Note that despite this, the contact resistances are not adversely affected when compared to the

Ti/Al/Pt/Au contacts, which were deposited in an electron beam evaporation operating at a base

pressure of 3 x 10-7 Torr. As deposited, the metallic interfaces are sharp and there is little

intermixing between the layers. Upon annealing at 6000C, Al has outdiffused to the surface and

some Au has indiffused to the interface. At higher temperatures, the amount of interdiffusion

increases. The amount of interdiffusion is much less than that seen with a standard metallic









diffusion barrier, such as Ni. Figure 4-6 shows a profile for a Ti/Al/Ni/Au contact annealed at

500C. From the profiles, the nitrides appear to be breaking down somewhat, with TiN the least

affected, as the location of the Zr and Ta peaks do not correspond precisely to those of the N.

Little change in the contact structure is seen with long term aging even after greater than 20 days.

Because there is only minimal intermixing of the layers, little of the viscous A1Au4 phase has

formed compared to conventional contact schemes.

Figure 4-7 shows the dependence upon anneal time at 8000C. Upon annealing for up to

180 seconds, an increase in contact resistance is seen in each of the contacts. The standard

Ti/Al/Pt/Au contact exhibit the largest increase in contact resistance, while the contacts

containing TaN and TiN display more moderate increases. The ZrN-based contact displays only

a slight increase in contact resistance even upon annealing for 180 seconds.

Figure 4-8 displays the specific contact resistance as a function of long term thermal aging.

Samples placed on a hot plate at 3500C display all display a gradual increase in the specific

contact resistance throughout the period. Table 4-1 shows the percentage increase in specific

contact resistance for each scheme. Compared to the Ti/Al/Pt/Au contacts, the ZrN and TaN

contacts display better stability, with the specific contact resistance increasing only about 100%.

The TiN contacts displayed a larger increase.

4.3 Conclusions

Ohmic contacts to n-GaN were fabricated using a Ti/Al/X/Ti/Au scheme, where X is either

TaN, TiN, or ZrN. These contacts display similar specific contact resistances to the conventional

Ti/Al/Pt/Au contacts deposited on the same wafer, -10-4 ohm-cm2 over a wide range of

annealing temperatures. In addition, the TaN and ZrN contacts display only about half the

increase in specific contact resistance that is seen in the Ti/Al/Pt/Au contacts. The AES depth

profiles confirm that little intermixing has occurred during the aging process. Because of this,









these contacts show promise for use in high temperature applications. While the contact

resistance does not seem to be better than that of conventional contacts after aging, there is far

less mixing between layers. This is a primary concern, due to the A1Au4 phase, for devices with

short spacings for high temperature applications. Further evaluation in devices such as HEMTs is

needed in order to determine to potential of these contacts. In addition, the effect of oxygen

incorporation in the nitride should be examined.









Table 4-1: Percent change in specific contact resistance during thermal aging
Contact Percent Change

Ti/Al/Pt/Au 236%
Ti/Al/TaN/Ti/Au 107%
Ti/Al/TiN/Ti/Au 330%
Ti/Al/ZrN/Ti/Au 091%










CM
E

0

0
c
Cu
C,


0
e-

0


1x10-2




1x10 3




1x10-4




1x10-5


600


800


1000
1000


vC Anneal Temperature (OC)





Figure 4-1: Specific contact resistance as a function of anneal temperature


-*- Ti/Al/Pt/Au
-*-- Ti/AI/ZrN/Ti/Au
Ti/Al/TiN/Ti/Au
-v- Ti/Al/TaN/Ti/Au

-0


1 ^\y
^1^ ==












TaN TiN ZrN


TaN-based as deposited TiN-based as deposited










TaN-based 600C anneal TiN-based 600C anneal





S.0


ZrN-based as deposited










ZrN-based 600C anneal


TaN-based aged TiN-based aged ZrN-based aged

Figure 4-2: Scanning electron microscopy images of annealed contacts





















4U- /
Ta Ti Ga
20-


0 1000 2000 3000 4000 5000 6000
Depth (A)
100-
(c)
80-

60-

40- Al
O G
20 u Ta

0
0 1000 2000 3000 4000 5000 6000
Depth (A)


100 1
(b)
80 U

60
N O Ga
40- \a
40

20 A Ti

0-
0 1000 2000 3000 4000 5000 6000 7000
Depth (A)


1000 2000 3000 4000 5000 6000
Depth (A)


Figure 4-3: Depth profiles of Ti/Al/TaN/Ti/Au contacts (a) as deposited (b) annealed at 6000C.
(c) annealed at 800C and (d) annealed at 800C and aged at 3500C.

























Depth (A)
100-
(c) F
80- N 0
N .
60-

Al C
40 V 0
\ 0 Ga
Au o
20- ZZ
20 Ti Ti E
0-0
0 1000 2000 3000 4000 5000 6000 7000
Depth (A)


Depth (A)


Ti 4


0 1000 2000 3000 4000 5000 6000 7000
Depth (A)


Figure 4-4: Depth profiles of Ti/Al/TiN/Ti/Au contacts (a) as deposited (b) annealed at 6000C.
(c) annealed at 800C and (d) annealed at 800C and aged at 3500C.

















30
Ti Ti

30-
S N
N

Ga
0 2

0T -- ---I 3 -- -
0 1000 2000 3000 4000 5000 6000


s
i- 8
.0
0e


4
0

E 2
0
'I


1000 2000 3000 4000 5000
Depth (A)


(b)

0-


0- Au NGa
N Au

0-


0- Al
0 Ti Z N


0
0 1000 2000 3000 4000 5000 60
Depth (A)


0 1000 2000 3000 4000 5000 6000
Depth (A)


Figure 4-5: Depth profiles of Ti/Al/ZrN/Ti/Au contacts (a) as deposited (b) annealed at 6000C.
(c) annealed at 800C and (d) annealed at 800C and aged at 3500C.


1-


o

, '
O



,
O




E


I t-tN


I I I I I















co
0


c0
0
0


o
=l


100


80

60


40


20


0


0 1000 2000 3000
Depth (A)




Figure 4-6: Depth profile of Ti/Al/Ni/Au contact annealed at 5000C.













0
E lx10--

C


80 4




0

-5
0
0.
n,)







1x10"-
Q.
(0


60 80 100 120 140 160 180

Anneal Time (s)


Figure 4-7: Specific contact resistance as a function of anneal time.


-- Ti/Al/Pt/Au
-*- Ti/Al/TaN/Ti/Au
T Ti/AI/TiN/Ti/Au
-v- Ti/AI/ZrN/Ti/Au











E 1x10-3



C

) 10-4 __,
---- Ti/AI/Pt/Au
S-- -- Ti/AI/TaN/Ti/Au
Ti/Al/TiN/Ti/Au
C --v- Ti/AI/ZrN/Ti/Au
0


1x105
o.
80 2 4 6 8 10 12 14 16 18 20 22 24
Cn Days






Figure 4-8. Specific contact resistance as a function of long term thermal aging.









CHAPTER 5
BORIDE-BASED SCHOTTKY CONTACTS TO P-GAN

5.1 Introduction

The unique combination of wide bandgap, high breakdown field, high saturation velocity,

and ability to form high-quality heterostructures with good transport properties make GaN an

ideal candidate for several device applications including light emitting diodes (LEDs) for

displays, laser diodes for data storage, and high electron mobility transistors (HEMTs) for high-

power and high frequency electronics. An important aspect for improved reliability of GaN-

based devices is the development of more thermally stable Ohmic and Schottky contacts to both

n-type and p-type GaN. This aspect has been one of the major issues in the operating lifetime of

III-nitride lasers and LEDs and it is yet to be overcome completely. Another example is

AlGaN/GaN HEMTs in advanced power amplifiers and converters for which the development of

Ohmic and Schottky contacts ensuring long-time operation under uncooled and high temperature

conditions without metal spiking or loss of edge acuity remains a challenge. One approach that

has proven promising for contact formation on n-GaN is the use of very high melting

temperature metals like W and WSix [24,68,71-74]. Recently, other promising metallurgies

based on transition metal borides like W2B, W2B5, ZrB2, CrB2, and TiB2 have also been

proposed [75-78]. The interest in borides is due to their high melting temperature (e.g. 3200C

for ZrB2) and high thermodynamic stability. For p-GaN, most rectifying contact studies have

been performed with simple metals such as Ni, Pd, Ti, Cr, and Pt, [17, 54-66] and studies on

boride-based metallurgies are still in their infancy.

In this work, we report on the annealing and measurement temperature dependence of W2B

and W2B5-based rectifying contacts to p-GaN. The barrier heights determined by X-Ray

Photoelectron Spectroscopy (XPS) are in the range 2.7-2.9 eV. These contacts are found to be









reasonably stable upon annealing to -600-7000C. While current-voltage characteristics as a

function of measurement temperature suggest thermionic field emission (TFE) over a Schottky

barrier as the most dominant mechanism of forward current flow, the bias and measurement

temperature dependence of the reverse-bias current indicates that leakage must originate from

surface leakage or generation in the depletion layer.

5.2 Experimental Details

The p-GaN samples were 1 [tm-thick Mg-doped GaN layers grown by metal organic

chemical vapor deposition (MOCVD) on 1 utm-thick undoped buffers on c-plane Al203

substrates. The hole concentration obtained from Hall measurements after acceptor activation

annealing was ~1017 cm-3. The surface was cleaned by sequential rinsing in acetone, ethanol and

10:1 H20:HC1 prior to insertion in the deposition chamber. Boride(500A)/Pt(200A)/Au(800A),

where the boride was either W2B or W2B5, was used as the Schottky metallization scheme. Au

was added to lower the contact sheet resistance while Pt is a diffusion barrier. All metals or

compounds were deposited by Ar plasma-assisted rf (13.56 MHz) sputtering at a pressure of 15

mTorr and rf powers of 250-400 W. The contacts were patterned by liftoff of lithographically-

defined photoresist and annealed at temperatures up to 7000C for 1 min in a flowing N2 ambient

in a RTA furnace. For Ohmic contacts, we used Pt/Au annealed at 5000C in 02 for 30s prior to

deposition of the Schottky metallization. Ring contact geometry for the diodes was employed,

with the Schottky contacts surrounded by the Ohmic contacts. The Schottky contact dots were

40 [tm in diameter, the surrounding was 60 [tm in inner diameter and 70 [tm in outer diameter.

XPS, current-voltage (I-V), and auger electron spectroscopy (AES) were used to

characterize boride-based rectifying contacts. XPS measurements were taken with a Physical

Electronics 5100LSci spectrometer with an aluminum x-ray source (energy 1486.6 eV). High-









resolution spectra were acquired to determine the binding energy (i.e. chemical state) and

concentration of specific elements observed in the survey spectra. Charge correction was

performed by using the known position of the C-(C,H) line in the C Is spectra at 284.8 eV. The

SBH was determined from the binding energy of the Ga 3d core level EB and the energy

difference between that core level and the valence band maximum Evc according to

B = EB -Evc .[62,79]

The I-Vs were recorded over the temperature range 25-200C using a probe station and an

Agilent 4145B parameter analyzer. For the relatively high doped p-GaN samples investigated in

this work, TFE over a Schottky barrier is expected to be the dominant mechanism of forward

current flow [56,57,62,63,80]. We fit the forward I-V characteristics to the relation for TFE over

a barrier [81]


JF =0 exp E)(5.1)
l-Eo

where JF is the current density, e is the electronic charge, and Vis the applied voltage. In

Eq. (1), Eo = E00 coth(E00/kBT) is the characteristic energy related to the tunneling probability,

kB and Tbeing the Boltzmann's constant and absolute temperature, respectively. For p-GaN,

Eoo (eV) 7.75 x 10 12O N (c 3) [56,62] NA being the density of acceptors. The saturation

current density Jo is given by

A*T ;T^Eooq( V -f ) 05 e e (OB -f
J= kB cosh(Eoo/kBT) exp kBT E (5.2)

where A*=103.8 Acm-2K-2 is the effective Richardson's constant for p-GaN, [82] OB is the

barrier height, and -(EF-Ev)/e is the difference between the valence band maximum and the

position of the Fermi level. This parameter depends on the doping level, but, as will be shown









later, it only slightly influences the SBH of W2B and W2B5-based rectifying contacts to p-GaN.

Nakayama et al. [83] have suggested that a hole concentration of2x1017 cm-3 in p-GaN samples

leads to -=0.13 V. This hole concentration is very close to that achieved in the present study so

that we also assume = 0.13 V.

Auger electron spectroscopy (AES) was used to analyze the depth profiles. A Physical

Electronics 660 Auger Microprobe Electron Beam at 10 keV, 0.3 utA, 300 from sample normal

was used for the data collection while for profiling the ion beam conditions were 3 keV Ar+,

2.0 utA, (4 mm)2 raster with sputter-etch rates of 80, 40, and 104 A/min for the boride, Pt and Au

layer respectively. A survey spectrum (a plot of the first derivative of the number of electrons

detected as a function of energy) was used to determine the composition of the outer few

nanometers of each sample. Quantifications were accomplished by using elemental sensitivity

factors.

5.3 Results and Discussion

Figure 5-1 presents an example of the Ga 3d core level and the valence band spectrum

collected on a p-GaN surface without a boride overlayer. From Figure 5-1, we find

Evc=17.54 eV, which is in reasonable agreement with the value of 17.8 eV reported previously.

[62,84] After W2B deposition, Figure 5-1 shows that the binding energy of the Ga 3d core level

is EB=20.25 eV, thereby yielding OB=2.71 eV. Similarly, we find EB=20.41 eV for W2B5, which

gives B=2.87 eV. These results are similar to the 2.68-2.78 eV values reported by Yu et al. [56]

for Ni/p-GaN. Based on the Schottky-Mott model [52], the SBHs for n-GaN and p-GaN should

add up to the GaN band gap 3.4 eV. For W2B and W2B5-based rectifying contacts to n-GaN,

Khanna et al. [76,77] have reported OB-0.55 eV, thereby suggesting ~B-2.85 eV for p-GaN. Our

XPS results are clearly in excellent agreement with the predictions of this model.









The I-V characteristics as a function of the measurement temperature for the W2B-based

diodes are shown in Figure 5-2. All log Ivs Vcurves measured between 25 and 2000C exhibit a

linear region at low forward bias (up to -2.0 V) and a less steep region at higher voltages.

Furthermore, all log Ivs Vcurves are parallel. As noted in refs. [56,57,62,63] for p-GaN samples

with relatively high doping levels (>1017 cm-3), this latter result is inconsistent with the

thermionic emission model but typical for carrier transport with a dominant tunneling

component. Note that Shirojima et al. [61] have reported accurate determination of the SBH

using the TE model. However, the acceptor concentration in their p-GaN samples was relatively

low (~1016 cm-3). In our case, the Mg concentration is expected to be in the 1019 cm-3 range, thus

confirming that the forward current cannot be analyzed using the thermionic emission model.

Figure 5-3 displays the I-V characteristics obtained from the diodes annealed at different

temperatures. The extracted characteristic energy related to the tunneling probability as a

function of annealing temperature is shown in Figure 5-4. For W2B, the value of Eo is seen to

remain stable in the 120 meV range and then increases beyond -5000C. This value of Eo

corresponds to an acceptor density of about 2x1020 cm-3, which is higher than that expected from

the Mg concentration alone (Eo~25meV for NA-1019 cm-3). This can probably be attributed to the

presence of acceptor-like deep level defects induced by both the high Mg doping [63,85,86] and

the sputter-deposition process [29,87]. In the case of W2B5, a slight decrease of Eo is observed at

intermediate annealing temperature which can probably be attributed to the annealing of sputter-

damage [29,87].

The influence of annealing on the SBH derived from the I-V measurements of W2B/Pt/Au

and W2B5/Pt/Au contacts is presented in Figure 5-5. Over the whole range of temperatures

investigated, the SBH only slightly decreases with increasing annealing temperature. However,









these values are unphysical, being larger than the GaN band gap, and are inconsistent with both

our XPS measurements and the predictions of the Schottky-Mott model. One may argue that this

discrepancy results from the value of used in our calculations. However, it is shown in Figure

5-6 that over the wide range of values investigated, this parameter only slightly influences the

SBH. The discrepancy between XPS and I-Vmeasurements is more likely related to the presence

of an interfacial defect layer or due to the presence of an interfacial oxide layer. Indeed, it was

shown by Shiojima et al. [88] using high-temperature isothermal capacitance transient

spectroscopy that the defects induced by the high Mg doping in p-GaN are essentially located at

the surface vicinity and that this region acts as a series capacitance [88,89]. Therefore, when such

defect or oxide layer is present, the carriers have to tunnel through an additional barrier, thus

resulting in an apparent SBH given by 0, = 0 + A0 where is the true SBH between the

contact metal and p-GaN and where Ao is the increment of the SBH due to the presence of the

thin defect or oxide layer. Our separate capacitance-voltage measurements (C-V) have indeed

shown that by accounting for such defect layer, the SBH obtained from the C-V characteristics

becomes similar to that determined by XPS [80]. We have calculated a corrected junction

capacitance Coorr for W2B/p-GaN from the equivalent circuit using the simple relation

(1/Ccor) = (1/Cm) -(1/C,), where Cm and Cox are the as-measured and oxide capacitances

respectively. From our XPS measurements, the true SBH for W2B/p-GaN is O 2.7 eV. Using

our C-V measurements, Figure 5-7 shows that the value of Co required to obtain a SBH similar

to that determined by XPS is Cox-0.70 nF. Assuming a contact area of -2x109 m2 and a relative

permittivity of -10 for the oxide layer [90], we estimate an oxide layer thickness of -0.25 nm,

i.e. about one monolayer. For 20 [tm-diameter diodes, the value of Co required to fit the XPS









data was 4 times lower than that for 40 [tm-diameter diodes, as expected from the surface area

ratio. In addition, the density of acceptors obtained from the corrected C-Vs was 1.8x102 cm-3,

which is very similar to the value determined above from the I-Vs. Table 5-1 summarizes the

calculated barrier heights from XPS, IV with both thermionic and thermionic field emission, and

CV measurements.

Figure 5-8 presents the influence of the measurement temperature on the reverse-bias

current for the W2B-based diodes. All log I vs V curves measured between 25 and 200C exhibit

a linear region at high reverse bias and a steeper variation at lower reverse voltages. In addition,

as the measurement temperature increases, the reverse leakage current also increases. Given the

large barrier height typical for rectifying contacts to p-GaN, one can assume that thermionic

emission over the Schottky barrier makes only a negligible contribution to the reverse-bias

current flow. This assumption is further supported by the fact that all log Ivs Vcurves in Figure

5-8 are parallel, which is clearly inconsistent with the thermionic emission model. On the other

hand, in the tunneling approach, the reverse current density JL is given by the Fowler-Nordheim

tunneling expression J, = CE2 exp(Eb/D), where C=C(OB) and D=D(OB) are parameters

independent of the measurement temperature [91-93]. For tunneling, the reverse-bias current is

thus independent of the measurement temperature, which again is inconsistent with the

experimental data displayed in Figure 5-6. Leakage must therefore originate from other

mechanisms such as generation in the depletion layer or surface leakage. The large band gap of

GaN makes the intrinsic carrier concentration in a depletion region very small, suggesting that

contributions to the reverse-bias leakage from generation in the depletion layer are negligible.

However, given the large number of deep-level defects located at the surface vicinity [88] as a

result of both the high Mg doping [29,85,86] and the sputter-deposition process [29,87], one may









expect these defects to act as generation centers when the carrier density is below its equilibrium

value as in the reverse-bias regime of a Schottky diode.

The breakdown voltage, VB, of the boride-based rectifying diodes as a function of

annealing temperature is shown in Figure 5-9 VB shows the opposite trend to the characteristic

energy related to the tunneling probability displayed in Figure 5-4, decreasing where Eo

increases. As Eo is directly related to the density of acceptor-like deep-level defects, this

observation is consistent with our expectations that leakage essentially results from a defect-

mediated mechanism. As observed for boride-based rectifying contacts to n-GaN [75,76], the

improvement of breakdown voltage at intermediate annealing temperatures is likely to result

from the annealing of sputter-damage in the near-surface region of GaN. At high annealing

temperature, VB is seen to decrease, indicating that the diodes become very leaky.

AES depth profiles of the as-deposited and annealed contacts are shown in Figure 5-10.

For both diodes, the as-deposited layers exhibit relatively sharp interfaces, consistent with the

featureless surface morphology observed with both optical and scanning electron microscopy.

After high-temperature annealing, W2B-based contacts show the onset of metallurgical reaction

in the contact scheme whereas W2B5 presents significant interdiffusion of layers.

5.4 Conclusions

In summary, our XPS measurements have shown that W2B- and W2B5-based rectifying

contacts to p-GaN produce barrier height in the 2.7-2.9 eV range, in excellent agreement with the

predictions of the Schottky-Mott model. By comparison, the SBH determined from the I-Vs

using the TFE model are unphysically large and much higher than those deduced from XPS due

to the presence of a defect layer acting as an additional barrier to carrier transport. The apparent

SBH only slightly decreases upon annealing to 600-7000C due to the onset of metallurgical









reaction with the GaN. This suggests that that the boride-based metallurgies may be promising

for rectifying contacts to p-GaN where thermal stability is a critical issue. The bias and

measurement temperature of the reverse leakage current indicates that leakage must originate

from surface leakage or generation in the depletion region through deep-level defects.









Table 5-1: Comparison of different barrier height calculations

XPS TE TFE
B = 2.7 eV B = 1.2 eV B = 3.8 e


n =4.4


V


Eo= 110 meV


C-V
= 4.1 eV

















Valence band
Region (X5)


I I

24 22 20 18 16 4 2 0
Binding Energy (eV)




W W2 B, p-GaN
Ga 3d / --- W2BI p-GaN






,I II


II
26 24 22 20 18 18 14
Binding Energy (eV)




Figure 5-1: XPS spectra without (top) and with (bottom) a boride overlayer. The left-hand
spectrum in the top figure corresponds to the Ga 3d core level whereas the right-hand
panel presents the spectrum of the valence band region.











10-2


1-3 .- ---
10 -



10-4



10 -



10-6 ,
0.0 0.5


1.0 1.5 2.0 2.5 3.0 3.5 4.0


Voltage (V)



Figure 5-2: Forward current-voltage characteristic ofW2B-based Schottky diodes as a function
of measurement temperature.











10-3


10-4


10-5


10-6


10.7
0


I W B/Pt/Au |I

12 /





S As-deposited
S -- 3000C anneal
/...... 6000C anneal

1 2 3 4 5 6
Voltage (V)


0 1 2 3 4 5 6
Voltage (V)




Figure 5-3: Forward current-voltage characteristic of W2B-based (top) and W2B5-based (bottom)
Schottky diodes as a function of annealing.


10-3


10-4


10-5


10-6


W2B/Pt/Au








;. /r As-deposited
S 3000C anneal
.- --A- o 6000C anneal





















E

LU


260
240
220
200
180
160
140
120
100
80
60
40
20
0


0 100 200 300 400 500 600

Anneal temperature (C)


700 800


Figure 5-4: Influence of the annealing temperature on the characteristic energy related to the
tunneling probability. Dashed and dotted lines correspond to the values ofEo for
NA-1019 and 5x1019cm-3 respectively.


- W2B/Pt/Au
--- W2B /Pt/Au


NA-5x1019 cm

N ~1019 cm-3
N-0 cm --


I I I I I I I I


' ' '


































0 100 200 300 400 500 600

Anneal temperature (oC)


700 800


Figure 5-5: Influence of the annealing temperature on the apparent Schottky barrier height
derived from IV measurements.


5.5 k


v W2B/Pt/Au
W2B /Pt/Au
*


4.5

4.0

3.5

3.0

2.5


. . .
















-- W2B/Pt/Au
4.5 WB/Pt/Au



4.0 -

-e-

3.5

Highly-doped Lightly-doped

3 .0 I 1I I 1 _
0.08 0.12 0.16 0.20 0.24

(eV)





Figure 5-6: Dependence of the apparent Schottky barrier height on the parameter defined as the
difference between the valence band maximum and the position of the Fermi level.
Low and high




























-3 -2 -1 0


1 2 3 4


Voltage (V)






Figure 5-7: As-measured and after oxide correction dependence of C2 versus V of Au/Pt/W2B/p-
GaN Schottky diodes. The measurement frequency was set to 1 kHz.


3.5

3.0

2.5

2.0


0~


I I I I


As-measured
0 After oxide correction


0
0




<^-:::-'


0.5


0.0 L
-4


' '


1.5

1.0














10-4


105 -


6
2 10-6 ....-2000C '
--1500C .
S...... 100 C
10- --- 50oc
-- 250C
10-8
-4.0 -3.5 -3.0 -2.5 -2.0 -1.5 -1.0 -0.5

Voltage (V)



Figure 5-8: Reverse current-voltage characteristic of W2B-based Schottky diodes as a function
of measurement temperature.



























0 100 200 300 400 500 600
Anneal temperature (oC)


700 800


Figure 5-9: Influence of the annealing temperature on the breakdown voltage.

















100 1
90 Au pt W2B/Pt/Au as-deposited -
80
70
60
W
50 Ga
40 B
30-
20
0
10
0
0 50 100 150 200

Sputter depth (nm)


uu
90 Au Pt W2B5/Pt/Au as-deposited -
80
70 -
60 Ga
50 W
40 N
30 B
20-
10
O -
0
0 50 100 150 20

Sputter depth (nm)


0 50 100 150 200

Sputter depth (nm)


0 50 100 150

Sputter depth (nm)


Figure 5-10: Depth profiles of W2B/Pt/Au contacts and W2B5/Pt/Au rectifying contacts (a,b)

before and (c,d) after annealing at 6000C.









CHAPTER 6
BORIDE AND IR BASED CONTACTS FOR LIGHT EMITTING DIODES

6.1 Introduction

InGaN/GaN multiple quantum well light-emitting diodes (MQW-LEDs) are commercially

available in a broad range of wavelengths for use in applications such as full color displays,

traffic signals, and exterior lighting. There is also interest in shorter wavelength LEDs with

AlGaN active regions to excite down conversion phosphors for white light [94]. Nevertheless, to

compete with fluorescent and other high-efficiency lighting sources, it is essential to drive GaN-

based LEDs at very high current densities to maximize light output. One drawback of the high

current densities is self-heating of the heterostructure which can produce Ohmic contact

degradation and generation of non-radiative recombination centers [95-98]. Cao et al. [99-101]

demonstrated improved heat dissipation and current spreading by reducing the density of

dislocations in the heterostructure through the use of free-standing GaN instead of the more

common sapphire substrates. Further reduction of self-heating effects and improved efficiency

and operating lifetime can be obtained through the use of low-resistance and thermally stable

Ohmic contacts to the p-GaN layer. Indeed, power dissipation by Joule heating across the p-

GaN/metal interface at high current injection levels was found to produce indiffusion of the

Ohmic contact elements along dislocations in the III-nitride-based epilayers, leading to an

electrical short of the pn junction. [102,103]

The conventional metallization schemes used for making Ohmic contacts to p-GaN for

InGaN/GaN LEDs are based on high work function metals such as Ni, Pd, Cr or Pt with an

overlayer of Au to reduce the sheet resistance. Specific contact resistances of 10-2-10-4 Q.cm2

are obtained by annealing at 450-650 C [106,107]. These contacts have stability problems at

high temperature. For example, the initial Au/Ni/GaN structure transforms to Ni/Au/GaN with a









rough Ni surface after annealing at 6000C [44,108]. A way to prevent excessive intermixing and

contact morphology degradation is to use a high-melting-point diffusion barrier in the contact

stack [109-111]. TiB2, with a melting temperature of 30000C, reasonable electrical resistivity

(28 [tQ.cm) and thermal conductivity (26 W.m-.K-1), and heat of formation comparable to those

for silicides or nitrides [112], shows promise as a diffusion barrier.

In this letter, we report on the long-term annealing characteristics at 200-350C of

InGaN/GaN MQW-LEDs with TiB2- and Ir-based p-Ohmic contacts. This high-temperature

stress stimulates accelerated aging of GaN-based LEDs and gives an idea of the expected

reliability of the Ohmic contacts. The use of TiB2 or Ir as a diffusion barrier in Ni/Au-based

contacts is found to produce superior long-term stability of turn-on voltage, leakage current and

output power.

6.2 Experimental

The MQW-LED structures were grown by metal organic chemical vapor deposition on c-

plane sapphire substrates. The layer structure consisted of a low-temperature GaN buffer, 3 |tm-

thick n-GaN, 0.1 |tm n-AlGaN clad, three period undoped InGaN/GaN MQW active, 0.1 [tm p-

AlGaN clad, and 0.3 |tm p-GaN layer. 1 [tm-deep mesas were fabricated using C12/Ar

inductively coupled plasma etching. Ohmic contacts to n-GaN were formed by lift-off of e-beam

deposited Ti (20 nm) / Al(80 nm) / Pt(40 nm) / Au(80 nm) subsequently annealed at 8000C for 1

min in a flowing N2 ambient in a rapid thermal annealing (RTA) furnace. The first p-

metallization scheme investigated consisted of Ni (50 nm) / Au (80 nm) / Ir (50 nm) / Au

(80 nm). The second contact scheme was Ni (50 nm) / Au (80 nm) / TiB2 (50 nm) / Ti (20 nm)/

Au (80 nm). Ti was added to improve the adherence of Au on TiB2. The Ni/Au layers were

deposited by e-beam evaporation while the Ir/Au and TiB2/Ti/Au overlayers were deposited by









Ar plasma-assisted rf magnetron sputtering. For comparison, devices with Ni/Au contacts were

also fabricated using the same wafer. All contacts were patterned by liftoff of lithographically-

defined photoresist and annealed at 6000C for 1 min in a flowing N2 ambient. Prior to their

insertion in the e-beam chamber, the surface of both n- and p-type GaN were cleaned in a

1HC: 10H20 solution for 1 minute. Figure 6-1 shows a typical optical micrograph of as-

fabricated MQW-LEDs with Ni/Au/TiB2/Ti/Au Ohmic contacts. Similar contact morphology

was achieved using the other metallization schemes. All devices were first aged for a period of

10 days at 2000C on a heater plate in air. The samples were then removed from the heater block,

allowed to cool to room temperature, and characterized before being returned for further 35 days

aging at 3500C. MQW-LEDs were analyzed by luminescence-current-voltage (L-I-V)

measurements using a probe station and an Agilent 4145B parameter analyzer. The light output

power was measured using a Si photodetector located at -2 cm from the sample surface.

Electroluminescence (EL) spectra were acquired by a fiber optic spectrometer.

6.3 Results and Discussion

Figure 6-2 presents the 300 K L-I characteristics from each as-fabricated device. Typical

EL spectra measured for different forward currents are also shown in the inset. For an injected

current of 500 ptA, the EL peak wavelength of the InGaN/GaN MQW-LEDs is 459 nm with a

full width at half maximum (FWHM) of 26 nm. As the injected current increases, the peak

exhibits a blue shift while the FWHM remains fairly constant. This behavior results from the

decrease by carrier screening of the quantum-confined Stark effect induced by the built-in

piezoelectric field in the strained InGaN/GaN QWs [113]. The L-I characteristics are similar for

all metallization schemes investigated, showing that the TiB2/Ti/Au or Ir/Au overlayers did not

reduce the light output with respect to the more conventional Ni/Au contacts. The non-linear









increase and relatively low output power can probably be attributed to enhanced current

crowding and self-heating effects [114,115]. These effects are expected to be particularly

important for the LED structures investigated in this work due to the high density of dislocations

in InGaN and GaN epilayers grown on sapphire as well as the low thermal conductivity of

sapphire [99-101]. As the dislocation density or LED junction temperature increases,

confinement in the MQW becomes less efficient, leading to premature saturation of the emission

power.

I-V characteristics from MQW-LEDs with Ni/Au, Ni/Au/TiB2/Ti/Au, and Ni/Au/Ir/Au

Ohmic contacts to the p-GaN layer are shown in Figure 6-3. For all diodes, the ideality factor

was >2, consistent with other data reported previously [95-101,116,117]. There is still no

agreement whether such high values result from tunnelling [101,117], recombination in the

depletion layer [117], or rectifying junctions present in heterostructure diodes [118]. The turn-on

voltage (voltage at which the current reaches 0.05 mA) and reverse leakage current (measured at

-5V) are given in Table 6-1. The uncertainty on each value was estimated from the deviation

between different devices on the same wafer. For as-fabricated devices, the turn-on voltage was

3.2-3.6V and the reverse leakage current was ~10-6 A. Such high leakage currents are commonly

observed in GaN-based LEDs grown on sapphire substrates and essentially result from the high

density of dislocations in such samples [101]. After 10 days aging at 2000C, an increase of the

turn-on voltage to -4.1 V and a decrease of the leakage current to -10-9 was observed for all

devices. After aging at 3500C for an additional 30 days, MQW-LEDs with TiB2- and Ir-based

Ohmic contacts show turn-on voltage and leakage current similar to those in the as-deposited

state whereas there was serious degradation in LEDs fabricated with Ni/Au contacts. This

reduction of injected current at a given forward voltage had a large impact on the EL output as









shown in Figure 6-4. While the brightness of MQW-LEDs with Ni/Au Ohmic contacts at 10 V

(I= 80 pA) was low, optical images of aged MQW-LEDs with boride-based Ohmic contacts

show much stronger emission, even at a lower applied voltage of 4.5 V (I= 300 pA).

Several mechanisms can be invoked to explain the drastic degradation of Ni/Au contacts

upon aging, including formation of an islanded contact morphology, formation of reactions with

the GaN resulting in a modification of the doping profile, and excessive intermixing of Ni and

Au leading to oxidation on the rough Ni surface and an increase of the series resistance

[44,109,119]. Although the dominant degradation mechanism remains unknown' the use of TiB2/Ti/Au

or Ir/Au overlayers o" the Ni/Au-baed Ohmic contacts clearly improves the long-term thermal stability of

InGaN/GaN MQW-LEDs.

6.4 Conclusions

In conclusion, TiB2 and Ir diffusion barriers in p-Ohmic contacts on GaN LEDs produced

less change in turn-on voltage, leakage current, and output power after long-term annealing at

2000C and 3500C for 45 days. These schemes look promising for high temperature applications

where improved stability over Ni/Au is mandatory.






Table 6-1: Influence of long-term aging at 2000C and 350C on the turn-on voltage and reverse
current of InGaN/GaN MQW-LEDs.


Contact to p-GaN Turn-on voltage (V) Reverse current ( -5V (A)
Day 0 Day 10 Day 45 Day 0 Day 10 Day 45
(@ 2000C ,@ 3500C (@ 2000C ,@ 3500C
Ni/Au 3.30.3 4.90.4 9.1+0.5 (21) x10-6 (62)x10-9 (53) x105
Ni/Au/TiB2/Ti/Au 3.50.2 4.1+0.5 4.00.2 (52)x 10-6 (21) x10-9 (2+1)x10-8
Ni/Au/Ir/Au 3.30.2 4.30.3 3.90.3 (3l)x 10-6 (21)x10-9 (32)x10-7



























Figure 6-1: Optical micrograph of an as-fabricated MQW-LED. The p-contact at the center of
the diode is 80 tm in diameter.












1.0
0.8
0.6
0.4
0.2
0.0
1.0-
0.8
0.6
0.4
0.2
0.0
1.0
0.8
0.6
0.4
0.2
0.0
-4


6 8 10


Figure 6-2 : L-I characteristics of MQW-LEDs with Ni/Au, Ni/Au/TiB2/Ti/Au, and Ni/Au/Ir/Au
p-Ohmic contacts. The inset shows emission spectra from as-fabricated LEDs at
various injection currents.


-Day 0
- Day 10 @ 2000C
-----Day 45 @ 3500C

------------------............---_


-2 0 2 4
Voltage (V)



























0.0 1/ I I -'Y**" **i I
0.0 0.5 1.0 1.5 2.0
Current (mA)




Figure 6-3: Influence of long-term aging at 250C and 350C on the I-V characteristics of LEDs
with (a) Ni/Au, (b) Ni/Au/TiB2/Ti/Au, and (c) Ni/Au/Ir/Au p-Ohmic contacts.

























Figure 6-4: Image of aged LEDs with (a) Ni/Au and (b) Ni/Au/TiB2/Ti/Au p-Ohmic contacts. In
(a), the picture was taken for a forward bias of 10 V (I = 80 pA), while the forward
voltage in (b) was 4.5 V (I= 300 LIA).









CHAPTER 7
CONCLUSION

Improved device processing is necessary in order to realize the full potential of GaN based

electronics and optoelectronics. While improved material quality is critical, especially for p-

GaN as well as related alloys such as InGaN to increase the efficiency of green emitting devices,

it is not the only challenge. Developing reliable, stable, low resistance Ohmic and Schottky

contacts to both n- and p-type GaN is still a challenge and hence remains of interest to the GaN

community at large. Without contacts that can withstand elevated temperatures and other harsh

environments, many of the properties that make GaN a unique and desirable semiconductor

mean nothing.

The goal of this work was to develop such contacts to both n- and p-GaN. In order to

improve upon existing contact schemes, it is necessary to achieve at least comparable contact

resistances or barrier heights. In addition, the contacts should be able to withstand more stressful

processing conditions, such as elevated annealing temperatures, while maintaining predictable

and stable characteristics. Perhaps the most critical factor is the ability of the contacts to

withstand long periods in a harsh environment, simulated here by placement on a hot plate at

temperatures of 200C and 350 C. A final consideration is the amount of intermixing of the

contact layers, especially in Ohmic contacts to n-GaN, as large amounts of intermixing can lead

to undesirable phases, such as A1Au4 and issues with lateral flow. In addition, if any undesirable

phases may form at device operation temperatures, or if the contacts themselves interact with the

GaN at these temperatures, devices will prove unreliable during prolonged operation. To tackle

these issues, three separate materials systems were examined: borides, nitrides, and Ir.









The first section of this dissertation involved the fabrication of Ohmic contacts to p-type

GaN. All three material systems mentioned were examined for this use. Contacts were

fabricated of the following structures:

1. Nickel / Gold / X / Titanium / Gold, where X is a chosen boride or nitride

2. X / Titanium / Gold, where X is a chosen boride of nitride

3. Nickel / Gold / Ir / Gold

4. Nickel / Ir/Gold

Each of these was then subjected to annealing at temperatures ranging from 300C to 10000C in a

flowing N2 ambient for 60 s. Schemes 1 and 3 consistently produced Ohmic contacts in the

range of 500-1000 C, 2 only inconsistently at specific temperatures, 3 at only 500C and 4 not at

all. Schemes 1 and 3 produced specific contact resistance values similar to those reported for

Nickel / Gold contacts in the literature. Contacts of scheme 1 were subjected to long term

thermal aging on a hot plate at both 200C and 350C. Nitride-based contacts failed early in

aging, however boride based contacts displayed stable specific contact resistances throughout.

Auger Electron Spectroscopy showed breakdown of the nitride structure during aging due to

severe intermixing with the GaN. The borides showed minimal intermixing with the GaN,

accounting for their stability. AES depth profiles of scheme 3 revealed severe intermixing of the

contacts with the GaN at anneal temperatures above 500C, accounting for their failure at

elevated anneal temperatures. Contacts fabricated with scheme 4 were not Ohmic due to the

absence of Au at the surface to promote increased hole concentration as well as severe

intermixing at high temperatures.

The second section of the work dealt with the examination of nitride based contacts to n-

GaN. Contacts using the borides and Ir have previously been examined. Contacts of the









structure Titanium / Aluminum / X / Titanium / Gold were fabricated using standard

semiconductor processing procedures, where X is either TaN, TiN, or ZrN. For comparison,

conventional contacts of the form Titanium / Aluminum / Platinum / Gold were fabricated as

well. The contacts were then subjected to a range of anneal temperatures between 500C and

1000C in a flowing N2 ambient for 60 s, as well as anneals at 8000C for up to 180 s, in order to

determine if the nitrides would allow for an increased thermal budget. Contacts were also aged

on a hot plate at 3500C and aged for a period of 24 days. Current-voltage measurements, Auger

Electron Spectroscopy, and Scanning Electron Microscopy were used to characterize the

contacts. Current-voltage measurements of the nitride contacts did not show any improvement

over the performance offered by the conventional Pt-barrier contacts in any of the experiments.

However, AES depth profiles revealed that the nitride based contacts displayed far less

intermixing than Ni-barrier contacts, even when the nitrides were annealed at higher

temperatures. Further, the profiles of thermally aged nitride contacts displayed no noticeable

difference from unaged contacts. Thus, while electrical performance of these contacts was

essentially unchanged from that of the conventional ones, they do offer the benefit of

dramatically reducing the amount of intermixing between layers even after being subjected to

harsh long term aging. This is significant, as one of the sources of failure for small gate width

GaN devices is lateral flow arising due to the mixing of the Al underlayer and Au overlayer,

leading to the formation of the viscous A1Au4 phase. This leads to short circuiting and thus

failure of devices.

Chapter five dealt with the fabrication of Schottky contacts to p-GaN. For this purpose,

W2B and W2B5 sputter targets were chosen. Nitrides were ignored due to the severe intermixing

expected between them and the GaN. It is also expected Ir would diffuse a great deal into the









GaN at elevated annealing temperatures, also making it unsuitable for use as a Schottky contact.

Contacts were fabricated with the scheme X / Platinum / Gold, where X is the boride. Platinum

was chosen as the adhesive layer instead of Titanium in order to eliminate formation of TiNx

phases. Current-voltage measurements were used to evaluate the barrier height of the contacts.

Because IV curves measured at different temperatures were parallel, it was determined that the

current transport mechanism present was thermionic field emission. Unphysical barrier heights

of approximately 4 eV were observed for anneals up to 7000C, with good stability. A slight

decrease is seen at increased annealing temperatures. This likely was the result of either an

increased near surface defect concentration due to sputtering or due to incorporation of an

interfacial oxygen layer between the GaN and the contact. X-ray Photoelectron Spectroscopy

measurements of thin boride layers on GaN reveal a true barrier height of 2.85 eV, in close

agreement with that predicted by the Schottky-Mott model. Capacitance-voltage measurements

confirm the large barrier height in agreement with the IV results, but if an interfacial layer of 1-2

monolayers is included the corrected barrier height would be in agreement with XPS

measurements.

Chapter six deals with the application of boride and Ir-based Ohmic contacts to p-GaN for

light emitting diodes. Contacts to LEDs were fabricated with the structures Nickel / Gold,

Nickel / Gold / Titanium Boride / Titanium / Gold, and Nickel / Gold / Iridium / Gold. They

were then annealed at 5000C, in order to prevent damage of the LED structure as well as metal

spiking into the active layers. Initial IV and photoluminescence measurements revealed all

contacts to have similar properties and performance. Contacts were then aged initially at 2000C,

after which all displayed a slight degradation of performance. After further aging at 3500C, the









Ni/Au contacts had degraded severely while the TiB2 and Ir-based contacts maintained good

performance. This increased device stability confirms the earlier work.

In conclusion, contacts to n- and p-GaN were fabricated. Nitride diffusion barrier contacts

to n-GaN show much less intermixing that that present with a Ni diffusion barrier, although no

improvement in the IV characteristics either as a function of annealing or during long term aging.

Ohmic contacts to p-GaN using the borides displayed both the ability to withstand a wide range

of anneal temperatures, indicating their increased thermal budget as compared to conventional

Ni/Au contacts, as well as far superior long term aging characteristics due to the stability of the

borides on GaN. This has implications for processing of devices using p-GaN which may be

used at elevated temperatures. While the specific Ohmic contacts may not be completely

suitable for immediate use in light emitting applications due to their poor reflectivity and

opacity, they show great promise. Development of transparent contacts using very thin layers of

these borides may be possible, as are improvements in the reflectivity by experimenting with

different metallization schemes and low anneal temperatures. In addition, if MOSFETs are to be

achieved using GaN, low resistance, stable Ohmic contacts will be necessary and these may fit

the requirements.









LIST OF REFERENCES


1. S. A. Campbell, The Science and Engineering of Microelectronics Fabrication 2nd ed.,
Oxford University Press, Oxford, UK, 2001.

2. J. C. Bean, in High-Speed Semiconductor Devices, edited by S.M. Sze, John Wiley & Sons,
New York, NY, 1990.

3. S. M. Sze, Physics of Semiconductor Devices, John Wiley & Sons, New York, NY, 1981.

4. D.C. Look, Mater. Sci. Eng. B80, 383 (2001).

5. 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).

6. 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).

7. 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).

8. V. Kumar, L. Zhou, D. Selvanathan, and I. Adesida, J. Appl. Phys. 92, 1712 (2002)

9. A. Motayed, R. Bathe, M.C. Wood, O.S. Diouf, R. D. Vispute, and S. Noor Mohammad, J.
Appl. Phys. 93, 1087 (2003)

10. J. Burm, K. Chu, W. A. Davis, W. J. Schaff, L.F. Eastman, and T. J. Eustis, Appl. Phys. Lett.
70, 464 (1997)

11. M. E. Lin, Z. Ma, F. Y. Huang, Z. F. Fan, L. H. Allen, and H. Morkoc, Appl. Phys. Lett. 64,
1003 (1994).

12. B. P. Luther, S. E. Mohney, T. N. Jackson, M. Asif Khan, Q. Chen, and J. W. Yang, Appl.
Phys. Lett. 70, 57 (1997).

13. Z. Fan, S. N. Mohammad, W. Kim, 0. Aktas, A. E. Botchkarev, and H. Morkoc, Appl. Phys.
Lett. 68, 1672 (1996)

14. S. Ruvimov, Z. Liliental-Weber, J. Washburn, D. Qiao, S. S. Lau, and Paul K. Chu, Appl.
Phys. Lett. 73, 2582 (1998).

15. A. N. Bright, P. J. Thomas, M. Weyland, D. M. Tricker, C. J. Humphreys, and R. Davies, J.
Appl. Phys. 89, 3143 (2001).

16. N. A. Papanicolaou, M. V. Rao, J. Mittereder, and W. T. Anderson, J. Vac. Sci. Technol. B
19, 261 (2001).

17. Q. Z. Liu and S. S. Lau, Solid-State Electronics, 42, 677-691 (1998).


137









18. R. Li, S. J. Cai, L. Wong, Y. Chen, K. L. Wang, R. P. Smith, S. C. Martin, K. S. Boutros, J.
M. Redwing, IEEE Electron Device Letters, 20, 323 (1999).

19. S. T. Sheppard, K. Doverspike, W. L. Pribble, S. T. Allen, J. W. Palmour, L. T. Kehias, T. J.
Jenkins, IEEE Electron Device Letters, 20, 161 -163 (1999).

20. D. Qiao, Z. F. Guan, J. Carlton, S. S. Lau, and G. J. Sullivan, Appl. Phys. Lett. 74, 2652
(1999)

21. V. Tilak, R. Dimitov, M. Murphy, B. Green, J. Smart, W. J. Schaft, J. R. Shealy and L. F.
Eastman, Mat. Res. Soc. Symp. Proc., 622, T7.4.1(2000).

22. S. -H. Lim, J. Washburn, Z. Liliental-Weber, and D. Qiao, Appl. Phys. Lett. 78, 3797 (2001).

23. M. AsifKhan, M. S. Shur and Q. Chen, Appl. Phys. Lett., 68, 3022(1996).

24. S. Murai, H. Masuda, Y. Koide, and M. Murakami, Appl. Phys. Lett. 80, 2934 (2002)

25. C. Lu, H. Chen, X. Lv, X. Xie, and S. N. Mohammad, J. Appl. Phys. 91, 9218 (2002)

26. J. S. Foresi and T. D. Moustakus, Appl. Phys. Lett., 62, 2859(1993).

27. M. W. Fay, G. Moldovan, P. D. Brown, I. Harrison, J. C. Birbeck, B. T. Hughes, M. J. Uren,
and T. Martin, J. Appl. Phys. 92, 94 (2002).

28. K. O. Schweitz, P. K. Wang, S. E. Mohney, and D. Gotthold, Appl. Phys. Lett. 80, 1954
(2002).

29. M. W. Cole, D. W. Eckart, W. Y. Han, R. L. Pfeffer, T. Monahan, F. Ren, C. Yuan, R. A.
Stall, S. J. Pearton, Y. Li and Y. Lu, J. Appl. Phys. 80, 278 (1996).

30. A. Zeitouny, M. Eizenberg, S. J. Pearton, and F. Ren, J. Appl. Phys. 88, 2048 (2000).

31. D. Selvanathan, F. M. Mohammed, A. Tesfayesus and I. Adesida, J.Vac.Sci.Technol. B22
2409(2004).

32. B. Luo, F. Ren, R. C. Fitch, J. K. Gillespie, T. Jenkins, J. Sewell, D. Via, A. Crespo, A.G.
Baca, R. D. Briggs, D. Gotthold, R. Birkhahn, B. Peres and S. J. Pearton, Appl. Phys. Lett.
82, 3910 (2003).

33. H. W. Jang and J. -L. Lee, J.Appl.Phys.93,5416(2003).

34. R. C. Fitch, J. K. Gillespie, N. Moser, T. Jenkins, J. Sewell, D. Via, A. Crespo, A.M.
Dabiran, P. P. Chow, A. Osinsky, J. R. LaRoche, F. Ren and S. J. Pearton, Appl.Phys.Lett.
84, 1495 (2004).

35. R. C. Fitch, J. K. Gillespie, N. Moser, T. Jenkins, J. Sewell, D. Via, A. Crespo, A.M.
Dabiran, P. P. Chow, A. Osinsky, J. R. LaRoche, F. Ren and S. J. Pearton,
J.Vac.Sci.Technol. B22, 619(2004).









36. D. Selvanathan L. Zhou, V. Kumar, I. Adesida and N. Finnegan, J.Electron.Mater.32
335(2003).

37. X. A. Cao, S. J. Pearton, G. Dang, A. P. Zhang, F. Ren, and J. M. Van Hove, Appl. Phys.
Lett. 75, 4130 (1999).

38. H. Morkoc, S. Strite, G. B. Gao, M. E. Lin, B. Svedlov, and M. Burns, J. Appl. Phys. 76 (2),
1363 (1994).

39. S. Nakamura and G. Fasol, The Blue Laser Diode -Springer, Heidelberg, 1997.

40. S. M. Sze, Physics of Semiconductor Devices, John Wiley & Sons, New York (1981).

41. http://www.onr.navy. mil/scitech/information/312_electronics/ncsr/properties.asp.

42. D. W. Palmer, www.semiconductors.co.uk, 2004a.

43. 0. Ambacher, J. Phys. D: Appl. Phys. 31, 2653 (1998).

44. H. Omiya, F. A. Ponce, H. Marui, S. Tanaka, and T. Mukai ,Appl. Phys. Lett. 85, 6143
(2004).

45. M. -R. Park and W. A. Anderson, Mat. Res. Soc. Symp. Proc. 680E, E6.5.1 (2001).

46. C. H Lin, D. L. Hibbard, A. Au, H. P. Lee, Z. J. Dong, F. J. Szalkowski, J. Chen, and C.
Chen, Mat. Res. Soc. Symp. 639, G4.8.1 (2001).

47. T. Gessmann, Y.-L. Li, E.L. Waldron, J.W. Graff, E.F. Schubert, J.K. Sheu, Appl. Phys. Lett.
80, 986 (2002).

48. J. K. Kim, J.-L. Lee, J. W. Lee, H. E. Shin, Y. J. Park, and T. Kim, Appl. Phys. Lett. 73,
2953 (1998).

49. V. R. Reddy, Mater. Chem.Phys. 93 286 (2005).

50. D. Selvanathan, L. Zhou, V. Kumar, I. Adesida and N. Finnegan, J.Electron. Mater. 32, 335
(2003).

51. G. -T. Chen, C. -C. Pan, C. -S.Fang, T. -C. Huang, J. -I.Chyi, M. -N. Chang, S. -B. Huang
and J. -T.Hsu, Appl. Phys. Lett. 85, 2797 (2004).

52. K. A. Rickert, A. B. Ellis, J. K. Kim, J. L. Lee, F. J. Himpsel, F. Dwikusuma, and T. F.
Kuech, J. Appl. Phys. 92, 6671 (2002).

53. Q. Z. Liu and S. S. Lau, Solid-State Electron. 42, 667 (1998).

54. M. W. Wang, J. O. McCaldin, J. F. Swenberg, T. C. McGill, and R. J. Hauenstein, Appl.
Phys. Lett. 66, 1974 (1995)









55. T. Mori, T. Kozawa, T. Ohwaki, Y. Taga, S. Naagai, S. Yamasaki, S. Asami, N. Shibata, and
M. Koike, Appl. Phys. Lett. 69, 3537 (1996).

56. L. S. Yu, D. Qiao, L. Jia, S. S. Lau, Y. Qi, and K. M. Lau, Appl. Phys. Lett. 79, 4536 (2001).

57. L. S. Yu, L. Jia, D. Qiao, S.S. Lau, J. Li, J.Y. Lin, and H.X. Jiang, IEEE Trans. Electron.
Devices 50, 292 (2003).

58. X. A. Cao, S. J. Pearton, G. Dang, A. P. Zhang, F. Ren, and J. M. Van Hove, Appl. Phys.
Lett. 75, 4130 (1999).

59. P. J. Hartlieb, A. Roskowski, R. F. Davis, W. Platow, and R. J. Nemanich, J. Appl. Phys. 91,
732 (2002).

60. D. L. Hibbard, R. W. Chuang, Y. S. Zhao, C. L. Jensen, H. P. Lee, Z. J. Dong, R. Shih, and
M. Bremser, J. Electron. Mater. 29, 291 (2000).

61. K. Shiojima, T. Sugahara, and S. Sakai, Appl. Phys. Lett. 74, 1936 (1999).

62. Y. J. Lin, Appl. Phys. Lett. 86, 122109 (2005).

63. K. Shiojima, T. Sugahara, and S. Sakai, Appl. Phys. Lett. 77, 4353 (2000).

64. T. Sawada, Y. Ito, N. Kimura, K. Imai, K. Suzuki, and S. Sakai, Appl. Surf. Sci. 190, 326
(2002).

65. C. K. Tan, A. Abdul Aziz, and F. K. Yam, Appl. Surf. Sci. 252, 5930 (2006).

66. Y. J. Lin and C. W. Lin, J. Electron. Mat. 33, 1036 (2004).

67. T. B. Massalski (Ed.), Binary alloy phase diagrams (ASM International, Materials Park,
Ohio, 1990).

68. X. A. Cao, S. J. Pearton, F. Ren, and J. R. Lothian, Appl. Phys. Lett 73, 942 (1998).

69. X. A. Cao, F. Ren, J. R. Lothian, S. J. Pearton, C. R. Abernathy, J. C. Zolper, M. W. Cole, A.
Zeitouny, M. Eizenberg, and R. J. Shul, MRS Internet Journal of Nitride Semiconductor
Research. 4 (1999).

70. Handbook of the Physiochemical Properties of the Elements, edited by G.V. Samsonov, IFI-
Plenum, New York, (1968).

71. M. W. Cole, F. Ren, and S. J. Pearton, J. Electrochem. Soc. 144, L275 (1997).

72. A. Zeitouny, M. Eizenberg, S. J. Pearton, and F. Ren, J. Appl. Phys. 88, 2048 (2000).

73. J. Kim, F. Ren, A. G. Baca, and S. J. Pearton, Appl. Phys. Lett. 82, 3263 (2003).









74. B. Luo, F. Ren, R. C. Fitch, J. K. Gillespie, T. Jenkins, J. Sewell, D. Via, A. Crespo, A. G.
Baca, R. D. Briggs, D. Gotthold, R. Birkhahn, B. Peres, and S. J. Pearton, Appl. Phys. Lett.
82, 3910 (2003).

75. R. Khanna, S. J. Pearton, F. Ren, and I. Kravchenko, J. Electrochem. Soc. 11, G804 (2005).

76. R. Khanna, S. J. Pearton, F. Ren, I. Kravchenko, C.J. Kao and G.C. Chi, Appl. Phys. Lett.
87, 052110 (2005).

77. R. Khanna, S. J. Pearton, F. Ren, and I. Kravchenko, Appl. Surf. Sci. 252, 5814 (2006).

78. T. N. Oder, P. Martin, J. Y. Lin, H. X. Jiang, J. R. Williams, and T. Isaacs-Smith, Appl.
Phys. Lett. 88, 183505 (2006).

79. R. T. Tung, Mater. Sci. Rep. R35, 1 (2001).

80. L. Stafford, L. F. Voss, S. J. Pearton, J. -J. Chen, and F. Ren, submitted.

81. Michael Shur, Physics of Semiconductor Devices (Prentice-Hall, Engelwood Cliffs, NJ,
1990).

82. C. Merz, M. Kunzer, U. Kaufmann, I. Akasaki, and H. Amano, Semicond. Sci. Technol. 11,
712(1996).

83. H. Nakayama, P. Hacke, M.R.H. Khan, T. Dtechprohm, K. Hiramatsu, and N. Sawaki, Jpn. J.
Appl. Phys. 35, L282 (1996).

84. J. R. Waldrop and R. W. Grant, Appl. Phys. Lett. 68, 2879 (1996).

85. J. S. Kwak, O. H. Nam, and Y. Park, Appl. Phys. Lett. 80, 3554 (2002).

86. J. S. Kwak, O. H. Nam, and Y. Park, J. Appl. Phys. 95, 5917 (2004).

87. M.W. Cole, F. Ren, and S. J. Pearton, J. Electrochem. Soc. 144, 275 (1997).

88. K. Shiojima, S. Sugitani, and S. Sakai, Appl. Surf. Sci. 190, 318 (2002).

89. T. Okumura, Jpn. J. Appl. Phys. 24, L437 (1985).

90. M. Passlack, N. E. J. Hunt, E. F. Schubert, G. J. Zydzik, M. Hong, J. P. Mannaerts, R. L.
Opila, and R. J. Fischer, Appl. Phys. Lett. 64, 2715 (1994).

91. M. Lenzlinger and E. H. Snow, J. Appl. Phys. 40, 278 (1969).

92. Z. Weinberg, J. Appl. Phys. 53, 5052 (1982).

93. H. Zhang, E. J. Miller, and E. T. Yu, J. Appl. Phys. 99, 023703 (2006).









94. Introduction to Nitride Semiconductor Blue Lasers and Light-Emitting Diodes, edited by S.
Nakamura and S.F. Chichibu (Taylor and Francis, London, 2000).

95. M. Meneghini, L. -R. Trevisanello, U. Zehnder, T. Zahner, U. Strauss, G. Meneghesso and E.
Zanoni, IEEE Trans. Electron. Dev. 53, 2981 (2006).

96. S. Bychikhin, D. Pogany, L. K. J. Vandamme, G. Meneghesso, and E. Zanoni, J. Appl. Phys.
97, 123714 (2005).

97. M. Pavesi, F. Rossi, and E. Zanoni, Semicond. Sci. Technol. 21, 138 (2006).

98. F. Rossi, M. Pavesi, M. Meneghini, G. Salviati, M. Manfredi, G. Meneghesso, A. Castaldini,
A. Cavallini, L. Rigutti, U. Strass, U. Zehnder, and E. Zanoni, J. Appl. Phys. 99, 053104
(2006).

99. X. A. Cao and S.D. Arthur, Appl. Phys. Lett. 85, 3971 (2004).

100. X. A. Cao, S. F. LeBoeuf, M. P. D'Evelyn, S. D. Arthur, J. Kretchmer, C. H. Yan, and Z. H.
Yang, Appl. Phys. Lett. 84, 4313 (2004).

101. X. A. Cao, J. M. Teetsov, M. P. D'Evelyn, D. W. Merfeld, and C. H. Yan, Appl. Phys. Lett.
85, 7 (2004).

102. S. Nakamura, M. Senoh, S. Nagahama, N. Iwasa, T. Yamada, T. Matsushita, Y. Sugimoto
and H. Kikyou, Jpn. J. Appl. Phys., Part 2 36, L1059 (1997).

103. C. Y. Hsu, W. H. Lan and Y. S. Wu, Appl. Phys. Lett. 83, 2447 (2003).

104. A. P. Zhang, B. Luo, J. W. Johnson, F. Ren, J. Han and S. J. Pearton, Appl. Phys. Lett. 79,
3636 (2001).

105. T. Maeda, Y. Koide and M. Murakami, Appl. Phys. Lett. 75, 4145 (1999).

106. L. -C. Chen, J. -K. Ho, C. -S. Jong, C. C. Chiu, K. -K. Shih, F. -R. Chen, J. -J. Kai and L.
Chang, Appl. Phys. Lett. 76, 3703 (2000)

107. D. Qiao, L. S. Yu, S. S. Lau, J. Y. Lin, H. X. Jiang and T. E. Haynes, J. Appl. Phys. 88,
4196 (2000).

108. H. S. Venugopalan, S. E. Mohney, J. M. DeLucca, B. P. Luther and G. E. Bulman, J. Vac.
Sci. Technol. A 16, 607 (1998).

109. C. -F. Chu, C. C. Yu, Y. K. Wang, J. Y. Tsai, F. I. Lai and S. C. Wang, Appl. Phys. Lett.
77, 3423 (2000).

110. L. F. Voss, R. Khanna, S. J. Pearton, F. Ren and I. I. Kravchenko, Appl. Phys. Lett. 88,
012104 (2006).









111. L. F. Voss, R. Khanna, S. J. Pearton, F. Ren, and I. I. Kravchenko, Appl. Surf. Sci. 253,
1255 (2006).

112. W. Zagozdzon-Wosik, D. Radhakrishnan, C. Dame, I. Rusakova, P. van der Heide, Z.-H.
Zhang, J. Bennett, L. Trombetta and P. Majhi, Rev. Adv. Mater. Sci. 8, 185 (2004).

113. Y. D. Qi, H. Liang, D. Wang, Z. D. Lu, W. Tang and K. M. Lau, Appl. Phys. Lett. 86,
101903 (2005).

114. X. Guo and E. F. Schubert, Appl. Phys. Lett. 78, 3337 (2001).

115. A. Chitnis, J. Sun, V. Mandavilli, R. Pachipulusu, S. Wu, M. Gaevski, V.Adivarahan, J. P.
Zhang, M. AsifKhan, A. Sarua and M. Kuball, Appl. Phys. Lett. 81, 3491 (2002).

116. P. G. Eliseev, P. Perlin, J. Furioli, P. Sartori, J. Mu, and M. Osinski, J. Electron. Mater. 26,
311 (1997).

117. G. Franssen, E. Litwin-Staszewska, R. Piotrzkowski, T. Suski, and P. Perlin, J. Appl. Phys.
94, 6122 (2003).

118. J. M. Shah, Y. L. Li, T. Gessmann and E. F. Schubert, J. Appl. Phys. 94, 2623 (2003).

119. S. H. Wang, S. E. Mohney, and R. Birkhahn, J. Appl. Phys. 91, 3711 (2002).









BIOGRAPHICAL SKETCH

Lars Voss was born in Pittsburgh, Pennsylvania in 1982. He spent most of his life in Erie,

Pennsylvania, until graduating from McDowell Senior High School in 2000. He then enrolled at

The Pennsylvania State University where he earned a Bachelor of Science in Engineering

Science, with a minor in electronic and photonic materials, in May 2004. While there, he had the

opportunity to work for Prof Paul Koch in plastics engineering technology at Penn State Erie

during the summer and with Prof. P.M. Lenahan in his Semiconductor Spectroscopy Laboratory.

After his undergraduate education, Lars enrolled at the University of Florida in the

Materials Science and Engineering Department and began his graduate study under Prof.

Stephen J. Pearton. During the summers of 2006 and 2007, he worked as an intern at Sandia

National Laboratories under the direction of Drs. Albert G. Baca and Randy J. Shul.





PAGE 1

1 THERMALLY STABLE OHMIC AND SCHOTTKY CONTACTS TO GaN By LARS FREDRIK VOSS 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 Lars Fredrik Voss

PAGE 3

3 To my grandparents.

PAGE 4

4 ACKNOWLEDGMENTS First and forem ost I would lik e to thank my advisor Prof. Stephen J. Pearton for his guidance and for all of the support and opportunities he provided. I would also like to thank my other supervisory committee members, Fan Re n, Cammy Abernathy, David Norton, and Rajiv Singh, for their help and their time. Thanks go to Dr. Patrick M. Lenahan for provi ding me with my first experience working in the semiconductor field while I was an undergraduate at Penn State as well as all of his advice and encouragement. I thank the members of the Pearton, Ren, and Abernathy re search groups with whom I have had the opportunity to work, including Luc Stafford, Kelly Ip, Jon Wright, Wantae Lim, Hungta Wang, Sam Kang, Travis Anderson, Soohwa n Jang, Brent Gila, Jerry Thaler, Jennifer Hite, Mark Hlad and many, many more. I would al so like to thank Ivan Kravchenko for his support in the UF Nanofabrication Facility. Thanks also go to all the people at Sandi a National Laboratories who gave me the opportunity to work there and with them for two enjoyable summers. I want to say thank you to my mentors, Randy J. Shul, Albert G. Baca, and Jeff E. Stevens, my managers, Charles Sullivan and Dale Hetherington, as well as all of the pe ople I had the pleasure to work with while at Sandia including Carlos Sanchez, David Torre s, Melissa Cavaliere, Karen Cross, Mark Overberg, Michael Cich, and many others. Thanks also go to all of my family and friends as well.

PAGE 5

5 TABLE OF CONTENTS page ACKNOWLEDGMENTS...............................................................................................................4 LIST OF TABLES................................................................................................................. ..........7 LIST OF FIGURES.........................................................................................................................8 ABSTRACT...................................................................................................................................11 CHAP TER 1 INTRODUCTION................................................................................................................... ..13 2 BACKGROUND..................................................................................................................... ..18 2.1 Gallium Nitride Properties................................................................................................ 18 2.1.1 Fundamental Properties.......................................................................................... 18 2.1.2 Electronic Properties..............................................................................................19 2.1.3 Crystal Structure..................................................................................................... 19 2.2 Properties of the Contact Materials to be Studied ............................................................19 2.2.1 Borides....................................................................................................................19 2.2.2 Nitrides...................................................................................................................20 2.2.3 Iridium.................................................................................................................. ..20 2.3 Electrical Contacts........................................................................................................ ....21 2.3.1 Ohmic Contacts...................................................................................................... 22 2.3.1.1 Ohmic contacts to p-GaN............................................................................. 23 2.3.1.1 Ohmic contacts to n-GaN............................................................................. 24 2.3.2 Schottky Contacts...................................................................................................25 2.4 Experiments......................................................................................................................27 2.5 Characterization Techniques............................................................................................ 29 2.5.1 Current-Voltage......................................................................................................29 2.5.2 Capacitance-Voltage...............................................................................................30 2.5.3 X-ray Photoelectron Spectroscopy......................................................................... 31 2.5.4 Auger Electron Spectroscopy................................................................................. 31 3 THERMALLY STABLE OHMIC CONTACTS TO p-GaN.................................................... 46 3.1 Ohmic Contacts............................................................................................................. ...46 3.1.1 Fabrication of Ohmic Contacts............................................................................... 46 3.1.2 Nitride-Based Contacts...........................................................................................47 3.1.2.1 Experiment and discussion........................................................................... 47 3.1.2.2 Summary......................................................................................................50 3.1.3 Tungsten Boride and Chromium Bori de-Based Contacts and Long Term Thermal Aging of Borides........................................................................................... 50 3.1.3.1 Experiment and discussion........................................................................... 50

PAGE 6

6 3.1.3.2 Summary......................................................................................................52 3.1.4 Contact Resistance for Othe r Boride-based C ontacts............................................ 52 3.1.4.1 Titanium boride-based contacts................................................................... 52 3.1.4.2 Zirconium Boride-based contacts.................................................................54 3.1.4.3 Gallium Nitride//Tungsten Boride-based contacts....................................... 55 3.1.5 Iridium-Based Contacts..........................................................................................57 3.1.5.1 Experiment and discussion........................................................................... 57 3.1.5.2 Summary......................................................................................................59 3.2 Conclusions...............................................................................................................59 4 OHMIC CONTACTS TO n-GaN.............................................................................................. 89 4.1 Experiment........................................................................................................................89 4.2 Results and Discussion..................................................................................................... 90 4.3 Conclusions.......................................................................................................................92 5 BORIDE-BASED SCHOTTKY CONTACTS TO p-GaN ...................................................... 103 5.1 Introduction............................................................................................................... ......103 5.2 Experimental Details......................................................................................................104 5.3 Results and Discussion................................................................................................... 106 5.4 Conclusions.....................................................................................................................110 6 BORIDE AND IR BASED CONTACTS FOR LIGHT EMITTING DIODES ...................... 123 6.1 Introduction............................................................................................................... ......123 6.2 Experimental............................................................................................................... ....124 6.3 Results and Discussion................................................................................................... 125 6.4 Conclusions.....................................................................................................................127 7 CONCLUSION........................................................................................................................132 LIST OF REFERENCES.............................................................................................................137 BIOGRAPHICAL SKETCH.......................................................................................................144

PAGE 7

7 LIST OF TABLES Table page 2-1 Bulk GaN properties........................................................................................................ ..33 2-2 Properties of common semiconductors.............................................................................. 34 2-3 Properties of the borides.................................................................................................. ..35 2-4 Properties of the nitrides................................................................................................. ...36 2-5 Properties of Ir...................................................................................................................37 3-1 Concentration of elements detected on the as-received surface (in atom %)..................... 60 3-2 Concentration of elements detected on the as-received surfaces (in atom%) .................... 61 3-3 Concentration of elements detected on the as-received surfaces (in atom%) .................... 62 3-4 Summary of specifi c contact resistances ........................................................................... 63 4-1 Percent change in specific cont act resistance during therm al aging.................................. 94 5-1 Comparison of different barrier height calculations ........................................................112 6-1 Influence of long-term aging at 200C and 350C on the turn-on voltage and reverse current of InGaN/GaN MQW -LEDs............................................................................... 127

PAGE 8

8 LIST OF FIGURES Figure page 1-1 Market forecast for GaN-based devices............................................................................. 17 2-1 Intrinsic carrier concentr ation of GaN, GaAs, and Si ........................................................38 2-3 Flat band diagram for a p-type Ohm ic contact.................................................................. 40 2-4 Flat band diagram for a p-type Schottky contact ...............................................................41 2-5 LED cross section (a) before and (b) after processing ....................................................... 42 2-6 Linear transmission line pattern......................................................................................... 43 2-7 Resistance vs. pad spacing plot.......................................................................................... 44 2-8 Schottky contact schematic................................................................................................45 3-1 Specific contact resistance and sheet resist ance under the contact of Ni/A u/ X/ Ti/Au contacts as a function of anneal temperature..................................................................... 64 3-3 Scanning electron microscopy images of Ni/Au/TaN/Ti/Au contact s (a) as deposited (b) annealed at 600 o C (c) annealed at 700oC and aged at 200 oC until the contacts became non-Ohmic and (d) annealed at 1000 oC...............................................................66 3-6 Specific contact resistance ve rsus m easurement temperature...........................................69 3-7 Specific contact resistance and sheet resistance under the contact as a function of long term thermal aging at 350oC......................................................................................70 3-8 Depth profiles of W2B-based contacts (a) as depo sited (b) annealed at 600 oC (c) annealed at 700 oC and aged at 350oC and (d) annealed at 1000 oC..................................71 3-9 Specific contact resistivity of Ni/Au/TiB2/Ti/Au Ohmic contacts and p-GaN sheet resistance under the contact as a function of annealing temperature................................. 72 3-10 Secondary electron images of Ni/Au/TiB2/Ti/Au contact pads on p-GaN as-deposited (top) or after annealing at either 800(center) or 900C (bottom)......................................73 3-11 Surface scans of Ni/Au/TiB2/Ti/Au Ohmic contacts on p-GaN as a function of anneal temperature. The as-deposited sample is at top, that annealed at 800C at center and that at 900C at bottom .....................................................................................................74 3-12 Depth profiles of Ni/Au/TiB2/Ti/Au Ohm ic contacts on p-GaN as a function of anneal temperature. The as-deposited sample is at top, that annealed at 800 pC at center, and that at 900 oC at bottom.................................................................................. 75

PAGE 9

9 3-13 Specific contact resistance of Ni/Au/ZrB2/Ti/Au and ZrB2/Ti/Au Ohmic contacts and p-GaN sheet resistance under the contact as a function of annealing temperature............ 76 3-14 Surface scans and depth profiles of Ni/Au/TiB2/Ti/Au Ohmic contacts on p-GaN as a function of anneal temperature.......................................................................................... 77 3-15 Scanning electron microscopy images of Ni/Au/TiB2/Ti/Au contact pads on p-GaN as-deposited (top) or after annealing at either 750 (cente r) or 800C (bottom)................. 78 3-16 Surface scans and depth profiles of ZrB2/Ti/Au Ohmic contacts on p-GaN as a function of anneal temperature.......................................................................................... 79 3-17 Elemental maps obtained from scanning AES of ZrB2/Ti/Au Ohmic contacts pads on p-GaN.................................................................................................................................80 3-18 Specific contact resistivity of W2B/Ti/Au Ohmic contacts and measured p-GaN sheet resistance under the contact as a function of annealing temperature................................. 81 3-19 AES surface scans of W2B/Ti/Au Ohmic contacts on p-Ga N as a function of anneal temperature.................................................................................................................... ....82 3-20 Depth profiles of W2B/Ti/Au Ohmic contacts on p-Ga N as a function of anneal temperature.................................................................................................................... ....83 3-21 Current-voltage curves for Ni/Au/Ir/Au contacts.............................................................. 84 3-22 Current-voltage curves for Ni/Ir/Au contacts.................................................................... 85 3-23 Depth profiles for Ni/Au/Ir/A u contacts (a) annealed at 300 oC (b) annealed at 500 oC and (c) annealed at 700 oC............................................................................................86 3-24 Depth profiles for Ni/Ir/Au contacts (a) annealed at 300 oC (b) annealed at 500 oC and (c) annealed at 700 oC.................................................................................................87 3-25 Scanning electron microscopy im ages of Ni/Au/Ir/Au contacts ........................................88 4-1 Specific contact resistance as a function of anneal tem perature........................................95 4-2 Scanning electron microscopy images of annealed contacts............................................. 96 4-3 Depth profiles of Ti/Al/TaN/Ti/Au contacts (a) as deposited (b) annealed at 600oC. (c) annealed at 800oC and (d) annealed at 800oC and aged at 350oC................................ 97 4-4 Depth profiles of Ti/Al/ TiN/Ti/Au contacts (a) as de posited (b) annealed at 600oC. (c) annealed at 800oC and (d) annealed at 800oC and aged at 350oC................................ 98 4-5 Depth profiles of Ti/Al/ ZrN/Ti/Au contacts (a) as de posited (b) annealed at 600oC. (c) annealed at 800oC and (d) annealed at 800oC and aged at 350oC................................ 99

PAGE 10

10 4-7 Specific contact resistance as a function of anneal tim e.................................................. 101 4-8 Specific contact resistance as a f unction of long term thermal aging.............................. 102 5-1 XPS spectra without (top) and with ( bottom ) a boride overlayer. The left-hand spectrum in the top figure corresponds to th e Ga 3d core level wh ereas the right-hand panel presents the spectrum of the valence band region.................................................. 113 5-3 Forward current-voltage characteristic of W2B-based (top) and W2B5-based (bottom) Schottky diodes as a function of annealing...................................................................... 115 5-4 Influence of the annealing temperature on the characteristic energy related to the tunneling probability. D ashed and dotte d lines correspond to the values of E0 for NA~1019 and 5 1019cm-3 respectively..............................................................................116 5-5 Influence of the annealing temperatur e on the apparent Schottky barrier height derived from IV m easurements........................................................................................ 117 5-6 Dependence of the apparent Schottky barrier height on the param eter defined as the difference between the valence band maximum and the position of the Fermi level. Low and high...................................................................................................................118 5-7 As-measured and after oxide correction dependence of C-2 versus V of Au/Pt/W2B/pGaN Schottky diodes. The measurement frequency was set to 1 kHz............................ 119 5-8 Reverse current-voltage characteristic of W2B-based Schottky diodes as a function of measurement temperature................................................................................................ 120 5-9 Influence of the annealing te m perature on the breakdown voltage................................. 121 5-10 Depth profiles of W2B/Pt/Au contacts and W2B5/Pt/Au rectifying contacts (a,b) before and (c,d) after annealing at 600 C....................................................................... 122 6-1 Optical micrograph of an as-fabricated MQW -LED. The p-contact at the center of the diode is 80 m in diameter.........................................................................................128 6-2 L I characteristics of M QW-LE Ds with Ni/Au, Ni/Au/TiB2/Ti/Au, and Ni/Au/Ir/Au p-Ohmic contacts. The inset shows emission spectra from as-fabricated LEDs at various injection currents................................................................................................. 129 6-3 Influence of long-term aging at 250C and 350C on the I-V characteristics of LEDs with (a) Ni/Au, (b) Ni/Au/TiB2/Ti/Au, and (c) Ni/Au/Ir/Au p-Ohm ic contacts............. 130 6-4 Image of aged LEDs with (a) Ni/Au and (b) Ni/A u/TiB2/Ti/Au p-Ohmic contacts. In (a), the picture was taken for a forward bias of 10 V ( I = 80 A), while the forward voltage in (b) was 4.5 V ( I = 300 A)............................................................................. 131

PAGE 11

11 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 THERMALLY STABLE OHMIC AND SCHOTTKY CONTACTS TO GaN By Lars Fredrik Voss May 2008 Chair: Stephen J. Pearton Major: Materials Science and Engineering This dissertation is focused on the developmen t of Ohmic and Schottky contacts to both nand p-type Gallium Nitride for use in microelectr onic and optical devices. The goal is to develop low resistance contacts with greater thermal budgets and superior thermal aging ch aracteristics to those commonly in use today as well as to unde rstand the mechanisms by which these contacts may fail. In addition, p-type Ohmic contacts ha ve been used to fabricate light emitting diodes (LEDs) which display far superi or aging properties than those made with conventional Ni/Au contacts. Ohmic contacts to p-GaN were fabricated using a variety of refractory materials. The materials examined were of three basic types: bo ride, nitride, and the refractory metal Ir. The boride family includes W2B, W2B5, CrB2, ZrB2, and TiB2. The nitrides examined were TaN, TiN, and ZrN. Contacts based on these ma terials were fabricated using either a GaN//Ni/Au/X/Ti/Au, GaN//X/Ti/A u, or GaN//Ni/X/Au scheme, where X is the refractory material. Contact resistances as low as ~1 x 10-4 /cm2 were consistently achieved after annealing at temperatures from 500-1000oC for 60 s in N2 using these materials for p-GaN with a carrier concentration of ~1 x 1017 cm-3. In addition, high temperature thermal aging was

PAGE 12

12 performed on selected schemes as well as on devi ces in order to attempt to estimate long term performance. Nest, nitride-based Ohmic cont acts to n-GaN are examined. The nitride was used to replace the conventional Pt, Ni, or Mo diffusion barrier in Ti/A l based contacts, for a contact scheme of GaN//Ti/Al/X/Ti/Au. These contacts ac hieve a similar specific contact resistance of ~1 x 10-4 /cm-2 for samples with a carrier concentration of 1 x 1017 cm-3 as that achieved with a Ti/Au/Pt/Au contact. The contacts are also ex amined as a function of aging and are found to display less intermixing of layers than t hose fabricated with a Ni diffusion barrier. Schottky contacts to p-GaN were also fabricated using the family of boride based materials with the scheme GaN//X/Pt/Au. These were f ound to exhibit tunneling transport through the metal-semiconductor junction and barri er heights of >3.5 eV were de termined from the fit of the current-voltage (IV) curves. X-ray Photoelect ron Spectroscopy was used to determine a true barrier height for these borides on p-GaN of 2.7 eV, close to that expected from the SchottkyMott model. Capacitance-voltage (CV) measuremen ts confirm the IV barrier height, but reveal a thin interfacial layer likely arising from oxidation or defects at the surface.

PAGE 13

13 CHAPTER 1 INTRODUCTION Se miconductor technology and its rapid improvements in device speed, yield, and density have enabled much of the technology and advan cements the world has experienced over the past half century. Beginning with the first transistor deve loped in 1947 at AT&T Bell Labs and then in 1957 with the first integr ated circuit by Jack Kilby, the pace of improvement in semiconductors and integrated circuitry is unmat ched by any other techno logical development in human history, and few have been more influential. Today, what once seemed ridiculously optimistic is possible and billions of transistors are present in a single integrated circuit [1]. Silicon based technology has come to dominate an overwhelming share of the semiconductor market, and for good reason. The main advantage of Si-based technology is the overwhelming knowledge base that exists as well as the f act that is possesses a wide range of acceptable properties. It possesses a higher melting point th an Ge, from which the first devices were made, as well as a stable oxide which di splays low leakage current. Furt her, nearly all of its properties fall into the acceptable range necessary for use in nearly all applications, including a reasonable band gap, carrier mobility, and breakdown field. For certain areas of application, however, Si is unacceptable. For applications requiring higher speed electronics, GaAs dominates the mark et. This is due to a much higher electron mobility, greater than 6 times that of Si. In a ddition, Si is also inappr opriate for optoelectronic applications due to its direct band gap. GaAs-based devices are also useful in this area. However, for high power, high temperature electronics as well as emission of wavelengths in the blue to ultra-violet range, GaAss properties are insufficient. It is therefore necessary to develop fabrication techniques for alternatives to Si and GaAs [2,3]. For these applications GaN is the material of choice. Its wide, direct band gap of 3.4 eV allows for emission in the UV range,

PAGE 14

14 while the use of InGaN and AlGaN allow for tu nable emission over the range 1.92 to 6.2 eV. [437] Development of GaN-based electronics has been necessitated due to the limitations of SiC, another semiconductor used for high power and high temperature electronics. SiC MOSFETs are limited by primarily thermal oxides, as the ga te contact degrades and becomes leaky at higher operating temperatures. In addition, th e electron mobility is low, only ~400 cm2/V-s, resulting in low power-added efficiencies of less than 30% for operation between 1 a nd 5 GHz. Because of its nature, GaN is a superior choice due to its wi der bandgap as well as greater chemical stability, leading to improved performance at higher operating temperatures. GaN based electronics are already in pr oduction, most notably high electron mobility transistors (HEMTs). Other devi ces of interest include heterojunction bipolar transistors (HBTs) and metal oxide field effect transistors (MOSFETs). In order to fully realize the potential of GaN-based electronics, improvements in their pr ocessing must be achieved in addition to the significant improvements in material quality that are necessary. This includes improvements in high temperature processing and operation, one area of which is the development of both Ohmic and Schottky contacts which are able to withstand high temperatures without degradation of device characteristics or contact structure [4]. This becomes especially important as the quality of the materials improves and fa ilure is due to the contacts. While electronic applications for GaN are important, optoelectronic applications are currently dominant. As Figure 11 indicates, most of the current and projected market are for optical applications. The first bright blue LED using GaN to be put into production was demonstrated by Shuji Nakamura in 1993 at Nichia Corporation. Since that time, a great deal of progress has been made in the attempt to realize the full potential of GaN for light emitting

PAGE 15

15 applications. The promise of solid state lighti ng is one that grows more important with each passing year, as the worlds insatiable hunger fo r energy grows while reserves of conventional fossil fuels struggle to keep up. Because produc tion cannot indefinitely keep up with demand, focus should and is on increasing the efficiency of consumption. Nearly 10% of world energy consumption is used for lighting. Most of this is currently done with extremely inefficient incandescent bulbs. While compact fluorenscent lighting has grown more popular and offers a substantial improvement in energy efficiency, it cannot compete with that possible with the wide spread use of LEDs. Further, CFBs present a hazard due to the non-trivial amount of mercury present in each bulb. Because of this, any large scale use of the bulbs must go hand in hand with proper disposal of them in order to prevent co ntamination of the environment. Solid state lighting does not present this problem. While many challenges still exist to realiz ing commercially viable solid state lighting sources, including significant improvement in the efficiency of the green emitting alloys as well as possible improvements in the phosphors used to emit the yellow and white light desired, it is still important to develop improve d contact structures in order to allow for longer life times and operation in environments with higher temperatures For p-type GaN especially, existing contact schemes could be improved to offer more stable performance. Failure of GaN-based devices due to contact failure should be minimized or eliminated. The goal of this work is to develop impr oved contact schemes fo r both electronic and optoelectronic applications that will offer acceptable specific c ontact resistances and barrier heights which will be stable over a wide range of processing temperatures as well as during long term high temperature aging. Chapter 2 deals wi th the basic properties of GaN as well as an overview of the processing and char acterization techniques that were utilized during this work.

PAGE 16

16 The rest of this dissertation is devoted to the ex periments and analysis. First, Ohmic contacts to p-GaN using a variety of high temperature material s are discussed. The ne xt section deals with Ohmic contacts to n-GaN. A section on Schottky contacts to p-GaN follows. The final chapter discusses the use of the Ohmic contact s to p-GaN in light emitting diodes.

PAGE 17

17 Figure 1-1 Market forecas t for GaN-based devices MARKET FORECAST FOR GAN-BASED

PAGE 18

18 CHAPTER 2 BACKGROUND 2.1 Gallium Nitride Properties 2.1.1 Fundamental Properties The m aterials properties of GaN make it desirabl e for a variety of applications. For optical electronics, GaN holds great promise. Because it is a direct band gap semiconductor, it is an efficient light emitter. GaN fills the need for a bl ue and ultraviolet wavelength emitter that is left by other materials due to its ba nd gap of 3.475 eV. Its exciton binding energy is 28 meV. This is not as high as another blue emitters, ZnO, but GaN technology is more developed. The first GaN LED was introduced in 1993; however, fu rther development of the GaN system is necessary for better LEDs and LDs. In addition to optical applications, GaN is robust and therefore a candidate for high temperature and high power appli cations, with potentially superi or electronic properties when compared with SiC [38]. The reason for this is due to the relatively wide band gap when compared to Si and GaAs. Figure 2-1 displays th e intrinsic carrier concentration as a function of temperature of GaN, Si, and GaAs. Because of the wide band gap, the intrinsic carrier concentration is much lower; 1015 cm-3 is reached at 300C for Si, 500C for GaAs, and at 1000C for GaN. Because of this, GaN can be operated at much higher temperatures before device performance degrades and breaks down. Table 2-1 contains some basic physical prope rties of GaN [39,40]. A key problem in device operation is the presence of the lattice mismatch of GaN with the commonly used sapphire (13%) and SiC (3%) substrates, which results in a large number of lattice defects which hinder electrical conduction.

PAGE 19

19 2.1.2 Electronic Properties The properties GaN possesses are advantageous f o r a variety of reasons. They, as well as those of other common semiconductors, are disp layed in Table 2-2 [40,41,42]. It has a high breakdown field, greater than 50 time s that of Si or GaAs, which allows for its use in high power electronic applications. The high electron mobility for its 2D elect ron gas and saturation velocity allows its use in high speed electronics. In a ddition, heterostructures su ch as AlGaN/GaN allow for the manufacture of interesting and high sp eed devices such as HEMTs. Recent work suggests that the 1.5 W/cm may be a lower limit for the thermal conductivity of GaN. GaN also shows excellent resist ance to irradiation. 2.1.3 Crystal Structure The therm odynamically stable phase of GaN is the hexagonal wurtzite structure, referred to as the -phase. In addition, a metastable -phase with a zinc blende structure exists. These two phases only differ in the stackin g sequence of Ga and N; their co existence in epitaxial layers is possible due to stacking faults [43]. The lattice parameters are a = 3.189 and c = 5.185 The wurtzite structure of Ga-face a nd N-face GaN is shown in Figure 2-2. 2.2 Properties of the Contact Materials to be Studied 2.2.1 Borides The fa mily of borides to be studi ed includes the refractory compounds W2B, W2B5, CrB2, TiB2, and ZrB2. Previous work has shown the promise of using refractory materials such as W as contacts to p-GaN. These borides also possess many of the requisite characterstics for use as electrical contacts fo r semiconductor devices. Table 2-3 summarizes known properties of the borides to be studied. Each of the borides possesses a high melting temperature, good thermal and electrical conductivities, high work functions, and large heats of formation. Each of these properties is important when considering a materi al for use in electrical contacts. The high

PAGE 20

20 melting temperature is necessary in order to ensure that the material is able to withstand the elevated annealing temperatures which are generally necessary for Ohmic contact formation. A high melting temperature is an indication of mate rial stability as well. A large electrical conductivity is necessary so as to achieve a low re sistance within the contact itself as well as to avoid too much self heating, which can lead to inconsistent performance and diffusion between contact layers during operation. A large therma l conductivity is necessary for this reason as well. For a p-type material, a la rge work function is desirable in order to achieve the proper band bending for thermionic carrier transpor t or at least to minimize the barrier for tunneling transport. Finally, a large heat of formation is desira ble in order to minimize reactions at the semiconductor-contact interface during operation. These materials have previously been examined for use in contacts to n-GaN and have displayed good characte ristics as a diffusion barrier in a Ti/Al-based contac t compared to more conventional metals such as Pt or Ni. 2.2.2 Nitrides The second class of m aterials to be studied is the nitrides. Th e materials chosen were TaN, TiN, and ZrN. Table 2-4 summarizes the properti es of these nitrides. They possess many of the same characteristics as the borid es and are therefore also worthwhile to study. These include a high melting temperature, good thermal and electri cal conductivities, a wo rk function, and large heats of formation. These nitrides have been stud ied for use in thin film resistors, wear-resistant coatings on tools, thermal printer heads, ga te electrodes, and diffusion barriers in Cu interconnection, which suggest their suitability as diffusion barrier s for GaN-based electronics. They will be examined in this case for Ohmic contacts to both nand p-type GaN. 2.2.3 Iridium The f inal material to be studied for use in contacts to p-GaN will be Iridium. Properties of Ir are summarized in Table 2-5. Note that Ir ha s the required high melting temperature, large

PAGE 21

21 work function, and reasonable thermal and electri cal conductivities. Ir has previously been studied for use as a contact to p-GaN. Here, it will be examined for use in Ni-based contacts to p-GaN as an intermediate layer. It has already been demonstrated that Ir acts a superior diffusion barrier in Ti/Al-based contact s to n-GaN compared to Ni. 2.3 Electrical Contacts High quality electrical contacts to sem iconductors are cri tical for the manufacture of all types of devices. While the requirements for spec ific devices and applications may differ, it is always necessary to connect these devices to the outside world. Generally, a contact that displays the lowest specific contact resistan ce with minimal drift during extended operation is the most desirable, although for certain applications other properties may be more critical. For instance, in light emitting devices it is often desi rable to have either transparent contacts or contacts with high reflectivities in the range of emission wavelengths in order to maximize light output. For high temperature applic ations it is critical that cont acts are stable during operation at high temperature; thus materials which displa y little intermixing and minimal reactions are necessary in addition to the requirement of good electrical and thermal conductivities. Electrical contacts to semiconductors are generally understood as th e junction between the semiconductor and a metal or other contact materi al, as well as any layers of metal or other material above this. The electrical properties of the contact are controlle d by the materials at the junction and it is often necessary to anneal the contacts at elevated temperatures in order to achieve the required properties th rough elimination of defects, re moval of compensators such as Hydrogen, or formation of low resistance intermetal lic phases. In addition, the layers above the junction itself can also influence the performance of the contact. It may be necessary to use an overlayer such as Au in order to promote current spreading so that the conduction is spread evenly over the area of the cont act as well as to prevent oxidation of the contact at room

PAGE 22

22 temperature. Other materials may be used as di ffusion barriers in order to minimize intermixing between various contact layers which may lead to the formation of undesirable secondary phases, segregation of layers when it is desirable that they be interm ixed, or even to prevent outdiffusion of the semiconductor to the surface of the c ontact. Two types of c ontacts exist, and it is important to understand what mate rials properties may be important in order to fabricate each. 2.3.1 Ohmic Contacts Ohm ic contacts are those which obey Ohms Law: V = IR(2.1) Note that this is a linear rela tionship and thus current is allowe d to flow into and out of the semiconductor without distortion of the signal. In general, the most critical component of an Ohmic contact is a low specific co ntact resistance. This will mi nimize the power consumption of the contact, which will also serve to minimize the internal heating of the contact and thus allow for prolonged operation with predic table performance. An ideal Ohmic contact would have no noticeable effect on device performance, as it would consume none of the power in the system nor would it cause any rise in temperature during operation. In theory, fabrication of Ohmi c contacts should be straightfo rward. For semiconductors with moderate doping, the dominant mechanism of current flow is generally thermionic emission, which is governed by Equation 2-2: kT q TqA kb c exp**(2.2) where k is Boltzmanns constant, q the electronic charge, A** is the Rich ardsons constant, T the temperature, and b the barrier height. In this model, el ectrons are excited over a barrier. It is clear, then, that the barrier height should be sm all in order to produce a low contact resistance. For p-type conduction, a metal with a work function larger than that of the semiconductor should

PAGE 23

23 be sufficient. The desired flat band diagram for th is situation is shown in Figure 2-3 [40]. It is clear from this diagram that holes may pass eas ily between the metal and the contact at the junction. However, for some semiconductors it can be di fficult to find materials systems that produce an Ohmic contact through thermionic emission, especia lly for p-type semiconductors. A wide band gap and large electron affinity, su ch as with GaN, effectively limits the choices to materials which have a very high work function. A second approach for formation of Ohmic contacts to semiconductors utilizes another form of carrier tr ansport: tunneling. This type of transport is characterized by the carriers t unneling through the potential barri er at the semiconductor-contact junction. Equation 2-3 describes the specific contact resistance for this type of transport: D bs cN m *2 exp (2.3) The relevant terms to consider in this equati on are the barrier height and the doping density, ND or NA for p-type semiconductors. It is obvious that in order to achieve a low contact resistance the barrier height must be minimized and the dop ing density must be maximized. Barrier height is determined by material choice. Doping density can be determined during growth of the material, with ion implantation, or use of contacts which can increase doping through interactions with the semiconductor. 2.3.1.1 Ohmic contacts to p-GaN High quality, low resistance Ohmic contacts to pGaN are difficult to achieve. This is due to several factors: the relative dearth of appropriate metal systems due to the large work function of GaN, low hole densities due to the difficu lty in achieving a high dopi ng concentration during growth, the presence of compensating hydrogen interstitials, and the pr esence of compensating

PAGE 24

24 nitrogen vacancies. Despite this, several ma terials systems have been found to produce Ohmic behavior. These are generally based on a high work function metal such as Ni, Pd, or Pt with an overlayer of Au and have been found to achieve specific contact resistances in the range of 1 x 10-4 -cm2 for hole concentrations of ~1017 cm-3 [44-48]. Ni/Au contacts are the most common. The Ni serves to remove hydrogen from the near surface region of the p-GaN at elevated annealing temperatures, as the Ni hydride is thermodynamically favor able. In addition, at these elevated temperatures Au begins to diffuse in towards the GaN and it has been found that the Au can form intermediate phases with th e Ga in the form of AuGa and AuGa2 [49]. This can serve to further increase the near surface doping thro ugh the formation of Ga vacancies in the GaN which act as acceptors. However, at these anneal temperatures of > 500oC, the Au and Ni may completely exchange places at the GaN surface [44] This leads to a severe roughening of the contact surface and degraded performance and limits the temperature at which anneals may take place as well as the lifetime of the contacts when operated at high temperature or, equivalently, high powers. Because one of the most attractive f eatures of GaN is its abili ty to operate reliably under these conditions, it is important to find a contact system that will perform better. 2.3.1.1 Ohmic contacts to n-GaN Ohmic contacts to n-GaN are more studied, better understood, and easier to achieve compared to those to p-GaN. The most common materials system used is GaN//Ti/Al/X/Au, where X is a material chosen to act as a diffusi on barrier to separate the Al and Au. This is necessary, as Al and Au will react to form AlAu4 [50]. This phase is viscous at low temperatures and lateral flow can create problems, especially as the gate/source separation becomes small. Because of this, there is great inte rest in the use of high temperature metals. The Ti is present at the junction in order to form TiN through reac tion with the GaN, which creates nitrogen

PAGE 25

25 vacancies and thus increases the donor density. The Au overlayer is present in order to prevent oxidation as well as to promote cu rrent spreading in the contact. 2.3.2 Schottky Contacts The other type of electrical contact required for electronic devices are known as Schottky contacts or diodes. Schottky contacts allow l ittle or no current to flow through them until a critical voltage is reached, above which large amount s of current flow. This is called the forward biased region. In the reverse bias region, vol tage is applied in th e opposite direction and no current flows initially. However, if the voltage is pushed high enough then breakdown occurs, a huge amount of current flows through the contact and the contact is generally destroyed. Ideal Schottky contacts are formed in the abse nce of any surface states. When a metal and semiconductor are brought together, the Fermi levels must line up. In order for this to occur, a charge exchange occurs between the two materi als. When this occurs, a space charge or depletion region is formed and a barrier to current flow is creat ed. For p-type semiconductors, the flat band diagram is shown in Figure 2-4. The barrier height, bp, for a p-type semiconductor is given by the equation qbp = g -q ( m ) (2.4) where Eg is the band gap, m is the metal work function, and is the electron affinity of the semiconductor. Invariably, however, surface states are present. This makes the prediction of barrier height using the equation above inaccurate. In order to de termine barrier height in this case, it is necessary to fabricate a specimen to test. The predominant carrier in Schottky contacts is the majority carrier unlike a pn junction diode. Two mechanisms for current flow exist: thermionic emission (TE) and thermionic field emission (TFE) or tunneling. While some authors have reported values for barrier heights on p-

PAGE 26

26 GaN using the TE model, most of the work has de monstrated that TFE is the dominant transport mechanism. Reported barrier heights on p-GaN ra nge from ~0.5 to 2.9 eV even for the same metallization scheme because the mechanism of cu rrent flow is not definitively established [5165]. In the case of TE, electrons or holes are exc ited over the potential barrier and the forward current is given by the following expression: *2expexpB F BBeeV JAT kTnkT (2.5) where A*=103.8 Acm-2K-2 is the effective Richardsons constant for p-GaN, T is the absolute temperature, e is the electronic charge, B is the SBH, kB is the Boltzmanns constant, and V is the applied voltage. In the case of TFE, quantum tunneling of char ge carriers through the ba rrier occurs and the forward current can be described by 0 0expFqV JJ E (2.6) where the saturation current density J0 is given by 0.5 00 0 00exp coshB B BBBBATEqV q q J kEkTkTkT (2.7) In Eq. (2.7), =( EFEV)/q is the difference between the valence band maximum and the position of the Fermi level and 00000cothBEEEkT is the characteristic energy related to the tunneling probability.

PAGE 27

27 2.4 Experiments The goal of this work is to find suitable cont act systems for use at high temperatures on nand p-GaN. Towards this end, the previously disc ussed materials will be investigated in order to determine which of them allow good quality Ohmi c and Schottky contacts. In addition, a study will be made of devices which use p-GaN, mainly LEDs. The contacts will be aged at high temperatures and compared with more conventional schemes. A variety of techniques will be used to characterize the contacts and devices in order to determine their performance and to better understand what is happening. The main experiments that will be done are the processing of the Ohmic and Schottky contacts for electrical characterization. The following sc hemes will be produced and analyzed: 1. GaN//Ni/Au/X/Ti/Au for p-type Ohmic contacts, where X indicates the refractory material to be examined 2. GaN//X/Ti/Au for p-type Ohmic contacts 3. GaN//Ni/Ir/Au for p-type Ohmic contacts 4. GaN//Ti/Au/X/Ti/Au for n-type Ohmic contacts 5. GaN//X/Pt/Au for p-type Schottky contacts In case (1), the Ni/Au should act as an Oh mic contact while the boride layer provides a diffusion barrier to prevent breakdown of the meta llization at elevated temperatures. The Ti layer is present in order to promote adhesion of the Au to the rest of the contact. The Au overlayer is to prevent oxidation an d act as a current spreading layer. In case (2), it is hoped that at least some of the borides may provide an Ohmic contact on p-GaN. Case (3) may provide some insight into the importance of Au at the GaN surface for these types of contacts. Case (4) is based on standard Ohmic contacts to n-GaN, with a nitride used as the di ffusion barrier. Case

PAGE 28

28 (5) will be used to fabricate Schottky contacts, as the Pt should not be as reactive as Ti should it reaches the GaN surface and is therefore used as the adhesion layer. Samples will be prepared on p-GaN and nGaN wafers grown in an MOCVD reactor. PGaN will be 3 m thick Mg-doped and grown on c-plane Al2O3. Hole concentration after activation obtained from Hall measurements is ~1 x 17 cm-3. The same doping will be used for both Ohmic and Schottky contacts. This is because it is extremely difficult to obtain a higher doping level for use in Ohmic contacts. At lo wer doping concentrations, it is not possible to obtain an Ohmic contact for use in Schottky contacts. For Ohmic contacts to n-GaN, epilayers 3 m thick Si-doped and grown on c-plane Al2O3 with an electron concentration of ~1 x 1017 cm-3 will be used. In order to fabricate the samples, photolit hography will be performed using a Karl Suss MJB3 aligner using positive photoresists. Samples for Ohmic contacts will then be etched in a Unaxis ICP tool in order to isolate devices and pr event current spreading. Prior to patterning for the metal layers, p-GaN samples will be dipped in a 1M KOH solution for one minute in order to remove surface oxide. Following patterning, su rface treatments may be necessary to improve performance. A 30 s RIE O2 plasma exposure to remove carbon followed by a 10% HCl dip for one to ten minutes immediately prior to insert ion into the deposition chamber will be employed to ensure a clean surface. Initial Ni/Au layers for contacts to p-GaN will be deposited using an e-beam evaporator in order to minimize surface damage. For deposition of all other layers, a Kurt Lesker sputterer will be employed. This is necessary, as the refractory materials cannot be evaporated. In addition, it is necessary to depos it all overlayers in the sputterer in order to minimize oxidation. Following lift off of the metal layers, rapid thermal annealing will be

PAGE 29

29 performed at a range of temperatures from 3001000C in order to determine how the properties of the contacts are affected at a variety of annealing temperatures. Following the development of these contacts, de vices will be fabricated. While p-GaN is unsuitable for use in HEMTs due to its low hole mobility, it is necessary for use in LEDs, LDs, and HBTs. LEDs will be fabricated on wafers w ith the structure shown in Figure 2-5. These will be compared to devices using more traditional metal schemes such as Ni/Au. They will then be thermally aged in order to determine the perfor mance relative to Ni/Au when maintained at an elevated temperature. 2.5 Characterization Techniques A variety of techniques will be utilized in or der to determine device characteristics as well as to observe the physical and chemical structure of the contacts. The techniques to be employed are 1. Current-voltage measurements 2. Capacitance-voltage measurements 3. X-ray Photoelectron Spectroscopy 4. Auger Electron Spectroscopy 5. Scanning Electron Microscopy 2.5.1 Current-Voltage Current-voltage (IV) measurements will be performed using an Agilent 4156 Semiconductor Parameter Analyzer. Two probe meas urements will be perfor med when possible. However, if problems which make it impossible to extract contact resistance arise due to high sheet resistance and the resistance of the probes, a four probe measurement technique will be employed. Using this method, current will be in jected with two outer probes while the voltage drop is measured between two inner probes. Th is allows for measurement of only the voltage

PAGE 30

30 drop between pads. IV measurements will be the main technique employed to determine the suitability of both Ohmic and Schottky contacts. The IV curves allow for the extraction of contact resistance for the former and barrier height and reverse breakdown voltage for the latter. IV measurements can be performed using either a circular or linear tr ansmission line method. The CTLM is useful for materials which are di fficult to etch and therefore may demonstrate problems with current spreading. In this study, linear TLM patterns will be used, as shown in Figure 2-6. Data from these curves allow for the creation of a plot such as that shown in Figure 2-7. Characterization of the pads is then performed using the relation S C C SC TR WR W L RRR222(2.8a,b) Figure 2-8 shows a schematic for Schottky cont acts. Schottky pads are formed using an inner Schottky contact with an outer Ohmic contact. 2.5.2 Capacitance-Voltage Capacitance-voltage (CV) measurements will be employed on Schottky contacts in order to further confirm results of IV measurements as well as to provide a deeper understanding of the barrier. If an interstitial la yer exists between the contact and the p-GaN, utilizing a CV measurement can reveal this. This is a distinct possibility that could arise either from an interstitial oxide layer forming or from def ects at the surface possibly caused by sputtering damage. Measurements were performed using an Agilent 4284A precision LCR meter in the parallel mode. The modulation frequency was set to 1 kHz. Barrier height is determined using

PAGE 31

31 lnBintABVEkTqg(2.9) where Vint is the extrapolated intercept volta ge of the reverse bias in the 1/ C2 versus V plot, EA 0.12 eV is the activation energy of Mg dopants in p-GaN, and g=2 is the degeneracy factor for acceptors. Note that in our case, the series resistance (~2 k ) and the junction conductance (~10-8-10-9) were low enough so that the measured capacitance corresponds to the proper junction capacitance. 2.5.3 X-ray Photoelectron Spectroscopy X-ray Photoelectron Spectroscopy (XPS) measurements will be further used to confirm the barrier height of Schottky cont acts. XPS functions by bombardi ng a specimen with soft x-ray radiation to examine core energy levels. When an x-ray collides with an atom, an electron is ejected from its shell. The elec tron is then collected and its energy measured. Each element has a characteristic binding energy for its core electro ns, and thus each element has a characteristic spectrum. Comparison of the spectrums of a bare specimen, such as GaN, and a specimen with a thin film coating, such as GaN//W2B, can yield the barrier height of the surface. The barrier height can be determined from the binding energy of the Ga 3d core level EB and the energy difference between that core le vel and the valence band maximum EVC according to BBVCEE (63,64)(2.10) 2.5.4 Auger Electron Spectroscopy Auger Electron Spectroscopy (AES) will be used to determine the elemental composition versus depth of the contacts. This will allow fo r an understanding of how diffusion plays a role in the behavior of the contacts. AES functions by bombarding the surface of a specimen with a

PAGE 32

32 beam of focused electrons. These electrons pene trate into the sample and collide with the atoms of the specimen. A core shell electron is then ejected, creating a vacancy. Following this, an electron from an outer shell relaxes and fills this lower energy state, releasing energy. This excess energy then results in the ejection of an el ectron from an outer shell. This Auger electron possesses an energy characteristic of the element from which it is ejected, allowing for the composition of the specimen to be determined. Using an ion gun in combination with the electron beam allows for information about th e composition versus depth to be collected.

PAGE 33

33 Table 2-1: Bulk GaN properties Property Value Lattice parameters at 300 K (nm) ao = 0.3189 nm co = 0.5185 nm Density (g-cm-3) 6.095 g-cm-3 Stable phase at 300K Wurtzite Melting point (oC) 2500 Thermal conductivity (W/cm-K) 1.3, 2.2 +/0.2 for thick, free-standing GaN Linear thermal expansion coefficient Along ao = 5.59x10-6 K-1 Along co = 7.75x10-6 K-1 Static dielectric constant 8.9 Refractive index 2.67 at 3.38 eV Energy bandgap (eV) Direct, 3.45 Exciton binding energy (meV) 26 Electron effective mass 0.20 Hole effective mass 0.59

PAGE 34

34 Table 2-2: Properties of common semiconductors Si GaAs GaN AlN 6H-SiC Bandgap (eV) @300oC 1.1 indirect 1.4 direct 3.4 direct 6.2 direct 2.9 indirect Electron mobility (cm2/V-s), RT 1400 8500 1000 (bulk) 2000 (2DEG) 135 600 Hole mobility (cm2/V-s), RT 600 400 30 14 40 Saturation velocity (cm/s), 107 1 2 2.5 1.4 2 Breakdown field (V/cm) x 106 0.3 0.4 >5 4 Thermal conductivity (W/cm) 1.5 0.5 1.5 2 5 Melting temperature (K) 1690 1510 >1700 3000 >2100

PAGE 35

35 Table 2-3: Properties of the borides Properties TiB2 ZrB2 W2B W2B5 CrB2 Melting Point (oC) 2980 ~3225 3040 ~3200 ~2670 ~2385 2200 Structure Hexagonal Hexagonal Hexagonal Hexagonal Thermal expansion coefficients x 106 (/deg) 4/6 5.9 10.5 Phonon component of heat conduction (wt/m-deg) 20.6 18.9 10.4 Elastic modulus E x 10-6 (kg/cm2) 5.6 4.3 2.5 Characteristic Temperature (K) 1100 765 726 Density of electronic state g x 10-21 (eV-1 cm-1) 4.5 4.76 54.6 Work function (eV) 4.19(?) 3.94(?) 3.18(?) Heat of Formation (kcal/mol) 71.4 76 31 Lattice constant (A) 3.028 3.169 2.982 2.969 Thermal conductivity (W/m-K) 26 80 Unknown 32 Electrical resistivity (uOhm-cm) 28 4.6 19 21

PAGE 36

36 Table 2-4: Properties of the nitrides Properties TaN TiN ZrN Melting point (K) 3633 3203 3253 Structure FCC FCC FCC Thermal expansion coefficients x 106 (/deg) 3.6 9.4 8.7 Work function (eV) 4.7 3.74 4.6 Lattice constant () 4.3 4.32 4.574 Thermal conductivity (W m-1 K-1) 57.5 19.2 20 Electrical resistivity ( cm) 15.52 25 13.6

PAGE 37

37 Table 2-5: Properties of Ir Properties Ir Melting point (K) 2719 Structure FCC Thermal expansion coefficients x 106 (/deg) 6.4 Work function (eV) 5.27 Lattice constant () 3.8 Thermal conductivity (W m-1 K-1) 147 Electrical resistivity ( cm) 4.71

PAGE 38

38 Figure 2-1 Intrinsic carrier concen tration of GaN, GaAs, and Si 1.52.02.53.03.54.0 10-2010-1010010101020 400300200100 0 GaN GaAs SiIntrinsic Conc. (cm-3)1000 / T (K-1) 25Temperature (oC)

PAGE 39

39 Figure 2-2: Stable wurtzite GaN crystal

PAGE 40

40 Figure 2-3: Flat band diagram for a p-type Ohmic contact

PAGE 41

41 Figure 2-4: Flat band diagram for a p-type Schottky contact

PAGE 42

42 (a) Sapphire GaN buffer n-GaN InGaN/GaN MQW p-GaN p-AlGaN n-AlGaN Sapphire GaN buffer n-GaN InGaN/GaN MQW p-GaN p-AlGaN n-AlGaN (b) Sapphire GaN buffer n-GaN InGaN/GaN MQW p-GaN p-AlGaN n-AlGaN p-contact n-contact Sapphire GaN buffer n-GaN InGaN/GaN MQW p-GaN p-AlGaN n-AlGaN p-contact n-contact Figure 2-5: LED cross section (a) before and (b) after processing

PAGE 43

43 Figure 2-6: Linear tran smission line pattern Metal Pads Semiconductor film L1L2L3W Metal Pads Semiconductor film L1L2L3W

PAGE 44

44 Figure 2-7: Resistance vs. pad spacing plot Distance LResistance R L1L2L3Slope=RS/W 2Rc Distance LResistance R L1L2L3Slope=RS/W 2Rc

PAGE 45

45 Figure 2-8: Schottky contact schematic Schottky contact Ohmic contact Semiconductor film Schottky contact Ohmic contact Semiconductor film

PAGE 46

46 CHAPTER 3 THERMALLY STABLE OHMIC CONTACTS TO P-GAN 3.1 Ohmic Contacts In this section, two approaches to fabricat ing Ohmic contacts using the boride family of materials will be discussed. The first is as an overlayer on conventional Ni/Au contacts and the second as the actual contact material. Because th ese borides have a large work function, it is expected that both approaches will be successf ul in yielding Ohmic be havior, provided that sputter damage and oxygen incorporation in to the borides duri ng deposition does not compromise the contacts, especially the latter. 3.1.1 Fabrication of Ohmic Contacts The samples used were light-emitting diode wafers with a 0.1 um Mg-doped p-GaN layer on top of a 0.1 um p-AlGaN layer followed by a 0.3 um InGaN/GaN superlattice on top of a ntype GaN layer grown on 2 um-thick undoped buffers on c-plane Al2O3 substrates. Hole concentrations in the p-GaN layer were 1017 cm-3 obtained from Hall measurements. Mesas for the linear transmission line method ( LTLM) method were etched using a Cl2/Ar inductively coupled plasma for pad isolation and to minimi ze current spreading. Samples were dipped in 1 M KOH for 1 minute and then rinsed in acetone a nd ethanol prior to photo lithography. Prior to insertion into the evaporation chamber, the samples were exposed to a 30 s O2 RIE plasma in order to remove any residual ca rbon and then dipped in 10:1 H2O:HCl for 5 minutes, which serves to remove any native ox ide. A Ni/Au contact of 500 / 800 was deposited using an ebeam evaporator in order to minimize surface damage. After unloading, the samples were put into a sputtering system for deposition of the X/Ti/Au laye r of thickness 500 /200 /800 where X is the refractory metal used. Au is n ecessary to decrease the sheet resistance while the Ti is necessary to prevent peeling of the Au fr om the surface due to the large lattice mismatch.

PAGE 47

47 In the case of Ir, this Ti layer is unnecessary. Contacts were also fabricat ed of the form Ni/Ir/Au, in order to confirm the importance of the Au n ear the GaN surface, and of the form X/Ti/Au to investigate the use of the refractory materials without the increased dop ing present due to the Ni/Au layer. Sputtering was perf ormed using an Ar plasma-assisted rf sputter at 5 mTorr and rf (13.56 MHz) power of 150 W. The sputter rates were 5 A/s for Au, 0.5 A/s for Ti, and were 0.82 A/s for the refractory materials. Following lift-off, the contacts were annealed in a flowing N2 ambient for 60 s in a rapid thermal ann ealing furnace at temperatures up to 1000 oC. 3.1.2 Nitride-Based Contacts 3.1.2.1 Experiment and discussion Ohmic contacts based on Ni/Au but with a nitr ide overlayer were fabricated in order to improve the temperature stability. Figure 3-1 show s the specific contact resistivity of the nitridebased contacts as a function of annealing temperature, as well as the sheet resistance of the pGaN under the contacts. At annealin g temperatures of less than 500 oC, contacts were rectifying. The contacts transition to Ohmi c behavior at higher temperatures. The specific contact resistance was not a function of temperature in the range 25-175oC, indicating that tunneling is the dominant transport mechanism. The specific c ontact resistance is fair ly stable on annealing up to 1000 oC, with a minimum of ~2 x 10-4 -cm2 at 600oC for the Ni/ Au/ ZrN/ Ti / Au contacts. This indicates that th ere is minimal change in the dopi ng of the near surface region of the GaN at these elevated anneal temper atures. The formation of AuGa and AuGa2 occurs at 600oC at the interface 49]. A sharp maximum in the sheet resistance is observed at 800 oC, with a drop at higher temperatures. Th is is likely due to increased in-diffusion of metal to the GaN without complete alloying, resulti ng in interstitial atom incorporat ion into the lattice. The drop in sheet resistance at higher temperatures is an indication that these interstitials are increasingly incorporated into the GaN. The magnitude of the increase in sheet resistance is related to the

PAGE 48

48 size of the metallic atom in the nitride, with the largest increase seen with the smallest atom Ti and vice versa for Ta, suggesting the atoms causing the spike are from the nitride. The reason the smaller atoms cause a more pronounced increase is the greater ease of diffusion into the GaN as well as a likely higher solid solubility. Figure 3-2 shows the AES depth profiles for the TaN-based contacts under a variety of conditions. Upon annealing at intermediate temper atures, most of the Ti has migrated to the surface and oxidized while Ni and Au are both pres ent at the GaN interface. In addition, the nitride layer is present at the GaN surface, sugges ting that the increase in sheet resistance is due to incorporation of the metallic atoms from the nitride into the GaN. After annealing at 700 oC and aging at 200 oC or annealing at 1000 oC, there is a large amount of intermixing between the metals and the GaN. Note that the upon annealing at 600oC the Ta and N peaks are still correlated, indicating that TaN ha s not reacted a great deal with the GaN yet. However, in the profile of the thermally aged c ontact there is no longer a correlat ion between the peaks of the Ta and N. This is an indication that the nitride is breaking down during the thermal aging. In fact, the profile of the aged contacts bears a much stronger resemblance to that of those annealed at much higher temperatures. SEM micrographs are shown in Figure 3-3 for the TaN-based contacts. The contacts show a smooth mor phology upon deposition, with an increasingly rough surface as anneal temperature incr eases due to the increased interm ixing of the contacts as well as oxidation of the surface. TiN and ZrN contact s display the same type of profile structure upon annealing and after thermal aging. As mentioned previously, th e Ni/Au/GaN structure is pr eferred due to the greater difference in electronegativities between Au and Ga than between Ni and Ga. By incorporating a layer on top of the Au, the dynamics of this relationship may change The electronegativities of

PAGE 49

49 Ta, Ti, and Zr are 1.5, 1.5, and 1.33, respectively. The values for Au, Ni, and Ga are 2.4, 1.8, and 1.6, respectively. From this, it is apparent that the differences between Au and either Ti, Ta, Zr, or Ga are roughly the same, indicating no preferred bonding to any of them. Various equilibrium phases exist for Au-Ti, Au-Zr, and Au-Ta, but none for Au-N, supporting the conjecture that Au is bonded to the metal. Thus neither configuration is strongly preferred. From the Au-Ni phase diagram, a miscibil ity gap exists with a maximum of 810.3 oC at around 70 at% Au. At higher and lower Au concentrat ions, the temperature above which Au and Ni intermix freely is lower; for exam ple, at 10 at% Au, it is below 500 oC and at 30 at% Au below 700 oC. Since there is no driving force in terms of greater bond strength to cause Au and Ni to switch places when using these diffusion barriers, the Ni and Au should each be present at the surface of the GaN. It is expected that the Ni and Au would separate with slow enough cooling but the layered structure of pure Ni/Au contacts may not form. The AES depth profiles confirm this, displaying significant intermixing at the GaN interface. We also see that the boundary between the GaN and the Ni-Au is not well-defined. This is consistent with the formation of the Au-Ga phases. From the Au-Ga phase diagra m, these phases have melting points of 461.3 oC and 491.3 oC, respectively. These are lower than th e temperatures required to obtain Ohmic behavior, suggesting it is beneficial to have so me Ni present at the GaN interface in order to maintain contact integrity. [67] Figure 3-4 shows specific contac t resistance and sheet resistance under the contacts as a function of thermal aging at 200 oC. This simulates long term operation of these contacts in devices. These contacts display poo r stability upon aging at this te mperature, lasting for nearly 10 days before the I-V characteristics are no l onger completely linear. From the AES depth profiles of the contacts that faile d, there was significant in-diffusi on of metals into the GaN. The

PAGE 50

50 temperature during aging is not sufficient for the formation of the Au-Ga phases and as a result the metal in-diffused to the GaN is in the form of interstitials, which hinder electrical conduction. In addition to this, the presence of the nitride on the surface of the GaN will lead to a breakdown the near surface GaN layer. Failure is attributed to the fact that the nitrides will intermix with the GaN. Since no change in chemical composition is expected to take place, only for Ga to exchange places with a metal such as Ti, there is no contribution to a change in the Gibbs free energy from formation or decomposition of comp ounds. In fact, it is expected that the intermixing may be favorable, as it will serve to increase the entropy of the system. From the equation G = H T S it is apparent that an increase in entropy leads to a lower free energy, and is therefore favorable. This leads to decomposition of the GaN. This is in agreement with the sheet resistance under the contacts increasing in the failed cont acts. It is also consistent with the finding in the AES profiles that the aged contacts display a breakdown in the nitride barrier layer. 3.1.2.2 Summary In conclusion, nitride-based contacts display low specific contact resistance and are stable upon anneals up to 1000 oC, with reasonable stability upon agi ng at elevated temperatures. Due to the nature of the contacts, they display extremely poor stability during aging and are therefore unsuitable for long term device operation. 3.1.3 Tungsten Boride and Chromium Boride-B ased Contacts and Long Term Thermal Aging of Borides 3.1.3.1 Experiment and discussion Figure 3-5 displays the specific contact resistan ce of the contacts as a function of anneal temperature as well as the sheet resistance under the contact. Th e contacts transition to Ohmic behavior at annealing temperatures of > 600oC. This likely correspond s to the creation of Ga

PAGE 51

51 vacancies due to the formation of AuGa and AuGa2 at the GaN surface and thus a localized increase in the hole concentration. This im plies that the current transport mechanism is tunneling. Figure 3-6 supports th is conclusion, as the contact resistance is not dependent on measurement temperature in the measured range 25-140 oC. The sheet resistance is also relatively stable as a function of annealing temperature, suggesting th at there is little intermixing between the boride and the GaN, even dur ing elevated anneal temperatures. Figure 3-7 displays the specific contact resistance of the W2B and CrB2 contacts as a function of long term thermal aging at 350oC. Both W2B and CrB2 display excellent stability for the duration of the aging, 23 days. While the nitrides provided good stability over the same range of anneal temperatures, the contacts fa iled after approximately 9 days at a lower temperature. This is attributed to the fact that the nitrides will inte rmix with the GaN. In the case of the borides, however, changes in the compounds must occur for significant intermixing to take place. This could take several forms, includ ing the reactions XB + GaN => XN + GaB or XB + GaN => XGa + BN. GaB phases are not present in phase diagrams [51]. Full thermodynamic evaluation of these is not possibl e, as most of the values fo r the Gibbs free energy are not available. However, it would be expected th at these reactions could only proceed at very elevated temperatures, as the compounds involve d are each very stable while any Ga-metal phases have relatively low melting te mperatures and thus can be assu med to be less stable [51]. For example, the maximum melting temperatur es of selected Ga-metal phases are 1670oC for GaTi, 1863oC for GaCr, and 1855oC for GaZr. These are much lower than the melting temperatures of either GaN or the borides or nitr ides. Thus, while a reaction of this nature may proceed during very high temperature anneals, during the thermal aging it will only proceed very slowly, if at all.

PAGE 52

52 Figure 3-8 displays the AE S depth profiles for the W2B-based contacts as deposited, annealed at 600 oC, annealed at 700 oC and aged at 350 oC, and annealed at 1000 oC. Upon annealing at 600 oC, the Ni and Au show strong mixing at the GaN interface. However, the boride barrier prevents most in termixing between the Ni/Au and the Ti/Au overlayer. The boride has minimal reaction with the GaN surface and the surface had excellent morphology at 600C. Note that the boride peak remains strongly co rrelated, indicating the W and B atoms remain together. This is the case even at 1000oC anneals and after long term thermal aging at 350 oC, indicating that the bo ride is not chemically ac tive in these contacts. At higher temperatures, the AES profiles were not entirely reliable due to non-uniformity depending on where the profile was taken. However, the IV characteristics we re still Ohmic and the su rface still had reasonable morphology. At temperatures of 1000C and higher, the contact layers became totally intermixed, although the peaks of W and B still co rrespond to one another. Similar results were obtained with CrB2-based contacts. 3.1.3.2 Summary These boride-based contacts display excellent thermal stability both upon high temperature anneals and during long term thermal aging at very high temperature. This is due to the fact that there is little intermixing between the boride and the GaN itself, due to the large change in Gibbs Free Energy that would be necessary for any chem ical reaction to occur. This is supported by the depth profiles obtained from AES, which s how that the boride has not broken down during experiments. 3.1.4 Contact Resistance for Other Boride-based Contacts 3.1.4.1 Titanium boride-based contacts Contacts with a TiB2 diffusion barrier for enhancing the thermal stability of Ni/Au Ohmic contacts on p-GaN were fabricat ed. Figure 3-9 shows the specifi c contact resistivity of the

PAGE 53

53 Ni/Au/TiB2/Ti/Au/p-GaN structure as a function of a nnealing temperature, along with the sheet resistance of the p-GaN under the contacts, extr acted from the TLM measurements. The contacts are rectifying below an anneal temperature of 500C but transition to Ohmic behavior for higher temperatures. The specific contact resistivity improves steadily with temperature above 700C and reaches a minimum of ~2 x10-4 .cm-2 after annealing at 800 C fo r 60 secs. Note that the contact resistance stays at a similar value even for temperatures up to 950C, even though the higher annealing temperatures produced increases in the sheet resistance of the GaN and less reproducible contact properties due to roughening of the contact morphology. The mechanism for the increase in sheet resistance may incl ude loss of N (producing nitrogen vacancy-related donors) or a reduction in the solu ble Mg concentration. This sp ecific contact resistance is comparable to that achieved on the same sample s with Ni/Au metallization annealed at 500C. The incorporation of the TiB2 diffusion barrier clearly provides a much wider range of thermal stability compared to the standard Ni/Au contacts. Figure 3-10 shows SEM pictures of the TLM contact pads as a function of annealing temperature. After 800C annealing the morphology is degraded, but is still much smoother than conventional Ni/Au contacts under th ese conditions. The darker appearance of the contacts after high temperature anneals is mainly a result of the outdiffusion of Ti, which rapidly oxidizes. Figure 3-11 shows the AES surface scans from the as-deposited sample (top) and those annealed at 800C (center) and 900C (bottom). Ti is evident on the surface after annealing at 800C and the concentration increases with a nnealing temperature. The increased oxygen concentration evident from the summary of th e data in Table 3-1 most likely comes from oxidation of this out-diffused Ti. The carbon signal in all cases comes from adventitious carbon on the surface.

PAGE 54

54 The AES depth profiles from the as-deposit ed and 800 and 900C annealed samples are shown in Figure 3-12. The as-deposited sample shows sharp interfaces be tween the metals and between the Ni and the GaN. After annealing at 800C, the Ni shows significant movement through all of the overlying layers and by 950C, mo st of the contact meta llurgy is intermixed and the Ti is essentially all removed to the surface. There is an accompanying decrease in abruptness of the TiB2 interfaces with the metals on either side. The TiB2 appears to be a barrier for Ti diffusion, so this probably excludes formati on of TiN phases at the interface of the GaN to account for the improved contact resistance. 3.1.4.2 Zirconium Boride-based contacts ZrB2-based contacts were also fabricated. Figure 3-13 shows th e specific contact resistance of the ZrB2/Ti/Au and Ni/Au/ZrB2/ Ti/Au/p-GaN structures as a function of annealing temperature, along with the sheet resistance of the p-GaN under th e contacts, extracted from the TLM measurements. The contacts are rectifying below an ann eal temperature of 750C but transition to Ohmic behavior for hi gher temperatures. The Ni/Au/ZrB2/Ti/Au contact resistance improves with higher anneal temperature, at the expense of poorer morphology. The ZrB2/Ti/Au contacts do not show low resist ance values until 800C and the value degrades at higher temperatures. The associated sheet resistance gene rally decreases with te mperature. The specific contact resistance values obtai ned with both types of ZrB2-based schemes is comparable to that achieved on the same samples with Ni/Au metallization annealed at 500C but the former have much higher thermal stability, as expected from the high melting temperature of the ZrB2. Figure 3-14 shows the AES surface scans and depth profiles from the as-deposited Ni/Au/ZrB2/ Ti/Au/p-GaN sample (left) and from the sa mple annealed at 750C (right). Ti, Ni and Zr are evident on the surface after annealing at 750C and the concentration increased with annealing temperature. The as-deposited sample shows sharp interfaces be tween the metals and

PAGE 55

55 between the Ni and the GaN. Table 3-2 summar izes the near-surface composition AES data. The increased oxygen concentration evid ent from this data most likely comes from oxidation of the out-diffused Ti. The carbon signal in all cases comes fr om adventitious carbon on the surface. After annealing at 750C, the Ni shows significa nt movement through all of the overlying layers and even the Zr diffuses out of the boride later, le aving the boron in its orig inal location. Both Zr and Au diffuse to the interface with GaN. The increased contact intermixing at higher annealing temperatures did roughen the contact morphology as shown in the SEM pictures of the TLM contact pads as a function of annealing temperature in Figure 3-15. After 800C annealing the morphology is degraded, but is still much sm oother than conventiona l Ni/Au contacts under these conditions. The darker appearance of the c ontacts after high temperature anneals is mainly a result of the outdiffusion of Ti, which then oxidizes. The same basic trends were observed with the ZrB2/Ti/Au structures. Figure 3-16 shows the AES surface scans and surface scans of ZrB2/Ti/Au Ohmic contacts on p-GaN as a function of anneal temperature. The as-deposited sample shows abrupt interfaces but after annealing at 900C, the Zr shows a very broad distribution (more so than the B). The c ontacts after annealing showed the presence of reacted islands, as s hown in the elemental maps of Figure 3-17. The islands contain both Au and Ga a nd show that the GaN epi layer ha s begun to dissociate at this temperature, at least under the contact metallurgy. Note that the areas with Au and Ga overlap, indicating that they are indeed forming th e Au-Ga phases reported in the literature. 3.1.4.3 Gallium Nitride//Tungsten Boride-based contacts Figure 3-17 shows the specific co ntact resistivity of the W2B/Ti/Au/p-GaN structure as a function of annealing temperat ure, along with the sheet resi stance of the p-GaN under the contacts, extracted from the TLM measurements. The contacts are rectifying below an anneal temperature of 500C but transition to Ohmic behavior at this te mperature. The specific contact

PAGE 56

56 resistivity improves steadily with temperature and a minimum of 1.7 x10-3 .cm-2 was obtained after annealing at 800 C for 60 secs. Higher anne aling temperatures produ ced sharp increases in the sheet resistance of the GaN and irreproduci ble contact properties. The mechanism for the increase in sheet resistance may include loss of N (producing nitrogen vacancy-related donors) or a reduction in the soluble Mg concentration. However the contact morphology as determined by SEM was similar over the entire annealing range used here. This specific contact resistance is comparable to that achieved on the same sample s with Ni/Au metallization annealed at 500C. Figure 3-18 shows the AES surface scans from the as-deposited sample and those annealed at 500, 700 and 800C.Ti is evident on the surface after anne aling at 500C and the concentration increases with annealing temperat ure. The increased oxygen concentration evident from the summary of the data in Table 3-3 most likely comes from oxidation of this out-diffused Ti. The carbon signal in all cases comes fr om adventitious carbon on the surface. The AES depth profiles from the as-deposited and annealed samples are shown in Figure 3-19. The as-deposited sample shows sharp in terfaces between the metals and between the W2B and the GaN. After annealing at 500C, the Ti shows significant movement through the Au layer and by 800C is essentially all removed to the surface. There is an accompanying decrease in abruptness of the W2B/GaN interface, which suggests the onset of reactions that are responsible for the improved contact resistance. Since the W2B appears to be a barrier for Ti diffusion, this probably excludes formation of TiN phases. Cole et al.[29] reported the formation of the W2N and W-N interfacial phases after annealing sputte red W on GaN [68]. The former appeared after anneals at 600C and the latter after 1000C. That work demons trated the importance of these interfacial phases in the resulting contact properties. The specific contact resistance obtained for the W2B-based contacts is lower than reported previously for pure W on p-GaN [69], likely due

PAGE 57

57 to the lack of these phases and the resulting compensation of holes by nitrogen vacancies. The absence of Ti diffusion through the W2B also suggests that an optimum metal scheme might involve a thin sandwich of the bor ide layers around the Ti, with the Au overlayer to reduce the sheet resistance. The reflectivity of pure W is around 60% at 460 nm, while that of Au is around 95% [70], so clearly the thickne ss of the boride layers should be small relative to the Au thickness. The long-term stability and reliability of the boridebased contacts must still be established on LEDs and laser diodes. In the former case, the reflectance of the multi-layered contacts at the emission wavelength of the LEDs must also be established. 3.1.5 Iridium-Based Contacts 3.1.5.1 Experiment and discussion Figures 3-21 and 3-22 show the IV plots for both types of contacts at different anneal temperatures and 20 m spacing. Only the Ni/Au/Ir/Au contacts annealed at 500 oC were found to have Ohmic characteristics. It was observed that these contacts displa yed a specific contact resistance of ~2.3 x 10-4 -cm2, comparable to results previously achieved on similar wafers for other schemes. Table 3-4 shows a summary of previous results achieved using a similar approach with other thermally stable materials as well as the anneal temperatures above which the contacts failed either due to unreas onable morphology or breakdown of Ohmic characteristics. From the IV curves for the Ni/Au/Ir/Au contacts, we see a decrease in the amount of current upon annealing at 300 oC, which likely corresponds to indiffusion into the GaN of Ni but at a temperature too low for the Ni hydride to form and increase the doping. At 500 oC the contacts transition to Ohmi c behavior due to increased near surface doping. At higher temperatures the contacts revert to rectifying behavior. For the Ni/Ir/Au contacts, the as deposited display the largest amount of current. At 300 oC the amount of current decreases

PAGE 58

58 again, only to increase at 500 oC due to removal of H by Ni. At 700 oC it begins to decrease again. Figures 3-23 and 3-24 show the Auger electron spectroscopy (AES) depth profiles for the Ni/Au/Ir/Au and Ni/Ir/Au contacts, respectively, at several different anneal temperatures. At 300oC, the contacts maintain well defined layers, with a distinct lack of intermixing. At this temperature, the contacts do not display Ohmi c characteristics because the Ni has not depassivated the H-Mg complexes. At 500 oC, the Ni/Au/Ir/Au contacts display significant interaction between the GaN and the Ni/Au underlayer. This likely corresponds to the formation of the Au-Ga phases, which serve to increase the doping. The Ni/I r/Au contacts show that the Ni has diffused towards the Au, but the Au has not diffused in towards the GaN. With little or no Au at the surface of the GaN, the hole concentrati on is not expected to increase as much as with the Ni/Au-based contact. This is the reason why Oh mic behavior is not observed in the Ni/Ir/Au contacts. In addition, we can observe that the Au seems to be essentially insoluble in the Ir at the moderate temperature anneals. At 700 oC, we can see that the layer structure has broken down and the metals are completely intermixed. Figure 3-25 displays scanning electron micr oscopy (SEM) images of the Ni/Au/Ir/Au contacts as deposited and annealed at several temperatures. The surface at 500 oC displays good morphology while that of the 700 oC anneal shows signs of the br eakdown of the layer structure. The failure of any contacts to display Ohmic behavior at high anneal temperatures is likely due to the nature of the Ir. Ir is expected to be stable on the GaN surface because it is not expected to form extensive secondary phases with the Ga or the N. While IrN has been recently been observed, it has only been synthesized at very high pressures. For the high temperature anneals, the Ir diffuses into the GaN. Howeve r, because no nitride can be formed under these

PAGE 59

59 conditions, the Ir must occupy interstitial sites and cannot be incorporat ed into the crystal structure of the GaN itself. Because the Ir atoms are very large a nd are likely sitting in interstitial sites, they can cause significant lattice distortions and it is expected that the crystallinity of the GaN would be significantly decreased. These distortions, as well as the Ir atoms themselves, would act as scattering sites for ch arge carriers. It is in this manner that the contacts fail when annealed at high temperatures. Thus, even when Au arrives at the surface in the Ni/Ir/Au contacts at 700 oC, the Ir has already caused enough disruption to the GaN that it does not matter. This same lack of chemical interaction of the Ir on the GaN may allow for improved long term operation of LEDs using the Ni/Au/Ir/Au contacts. Because the Ir is not expected to interact chemically with the GaN, at the lower temperatures which were examined during aging, and which would be present during long term ope ration, the GaN itself is not expected to decompose over a long period of time. 3.1.5.2 Summary Ir diffusion barriers for contacts to p-GaN have been fabricated. The importance of Au being in contact with the GaN for the format ion of low resistance Ohmic contacts has been confirmed, as the Ni/Ir/Au contacts do not display Ohmic characteristics. 3.2 Conclusions Ohmic contacts have been fabricated to p-GaN using three classes of materials. It has been determined that boride based contacts are superior to either Ir or nitride based ones due to both their ability to withstand a wide range of annealing temperatures and to withstand long term thermal aging. From these results, it can be not ed that it is important to choose materials for contacts which will not display a great deal of either chemical reaction or indiffusion.

PAGE 60

60 Table 3-1: Concentration of elements de tected on the as-received surface (in atom%) Filename Sample ID C(1) O(1) Ti(1) Au(3) Sensitivity factors [0.076] [0.212] [0.188] [0.049] 051829101 As deposited 36 1 nd 64 051829201 800C annealed 38 31 19 12 051829301 900C anneal 27 33 28 12 051829401 950C Anneal A 30 26 33 nd 051829401 950C Anneal B 28 26 35 nd

PAGE 61

61 Table 3-2: Concentration of elements dete cted on the as-received surfaces (in atom%) Sample ID C(1) O(1) Ti(1) Ni(1) Zr(2) Au(3) Sensitivity factors 0.076 0.212 0.188 0.227 0.043 0.049 As deposited Ni/Au/ZrB2/Ti/Au 44 4 nd nd nd 52 annealed 750C 45 19 2 1 14 19 annealed 800C 42 22 9 1 16 10 As deposited ZrB2/Ti/Au 37 4 nd nd nd 59 annealed 800C 43 23 nd nd 12 22 annealed 1000C 48 19 nd nd 14 19

PAGE 62

62 Table 3-3: Concentration of elements dete cted on the as-received surfaces (in atom%) Sample C(1) O(1) Ti(2) Au(3) [0.076] [0.212] [0.296] [0.049] As deposited 41 nd nd 59 500C annealed 43 24 5 28 700C annealed 44 34 7 15 800C annealed 38 34 15 13

PAGE 63

63 Table 3-4: Summary of specific contact resistances Contact Scheme sc,minium ( -cm2) Anneal temperature (oC) Failure temperature (oC) Ni/Au/W2B/Ti/Au 3.5 x 10-5 1000 >1000 W2B/Ti/Au 1.69 x 10-3 800 >900 Ni/Au/CrB2/Ti/Au 7 x 10-5 1000 >1000 Ni/Au/TiB2/Ti/Au 1.93 x 10-4 850 >900 Ni/Au/ZrB2/Ti/Au 1 x 10-4 900 >900 ZrB2/Ti/Au 1.8 x 10-3 800 >900 Ni/Au/TaN/Ti/Au 2.5 x 10-4 800 >1000 Ni/Au/TiN/Ti/Au 2.45 x 10-4 600 >1000 Ni/Au/ZrN/Ti/Au 2 x 10-4 600 >1000 Ni/Au/Ir/Au 2.3 x 10-4 500 >500 TaN/Ti/Au 3.7 x 10-4 800 >1000 ZrN/Ti/Au 3.1 x 10-4 600 >700

PAGE 64

64 5006007008009001000 1x10-51x10-41x10-3 0 20 40 60 80 100 TaN TiN ZrN Specific Contact Resistance(-cm2)Anneal Temperature (C) Sheet Resistance (/) TaN TiN ZrN Figure 3-1: Specific contact resistance and sheet re sistance under the contact of Ni/Au/ X/ Ti/Au contacts as a function of anneal temperature.

PAGE 65

65 Figure 3-2: Depth profiles of Ni/Au/TaN/Ti/Au contac ts (a) as deposited (b) annealed at 600 o C (c) annealed at 700oC and aged at 200 oC until the contacts became non-Ohmic and (d) annealed at 1000 oC. 05001000150020002500300035 0 0 10 20 30 40 50 60 70 80 90 100 (a) Au Ga Ta Ti Ni Au NAtomic Concentration (%)Depth ()N 010002000300040 0 0 10 20 30 40 50 60 70 80 90 100 (c)O Au Ga Ta Ti NiAtomic Concentration (%)Depth ()N 010002000300040 0 0 10 20 30 40 50 60 70 80 90 100 (d)O Au Ga Ta Ti NiAtomic Concentration (%)Depth ()N010002000300040 0 0 10 20 30 40 50 60 70 80 90 100 O Au Ga Ta Ti Ni Au NAtomic Concentration (%)Depth ()N(b)

PAGE 66

66 Figure 3-3: Scanning electron micr oscopy images of Ni/Au/TaN/Ti/Au contacts (a) as deposited (b) annealed at 600 o C (c) annealed at 700oC and aged at 200 oC until the contacts became non-Ohmic and (d) annealed at 1000 oC.

PAGE 67

67 0246810 1x10-51x10-41x10-3 0 20 40 60 80 100 120 140 160 TaN TiN ZrN Specific Contact Resistance (-cm2)Day Sheet Resistance ( TaN TiN ZrN Figure 3-4: Specific cont act resistance of Ni/Au/X/Ti/Au cont acts as a function of aging at 200 oC.

PAGE 68

68 Figure 3-5: Specific contact resistance and sheet resistance under the contact versus anneal temperature.6007008009001000 1x10-61x10-51x10-41x10-31x10-2 0 5 10 15 20 25 30 Specific Contact Resistance (-cm2)Anneal Temperature (oC) CrB2 W2B Sheet Resistance ()

PAGE 69

69 Figure 3-6: Specific contact resistance versus measurement temperature. 0255075100125150 1x10-61x10-51x10-41x10-31x10-2 Specific Contact Resistance (-cm2)Measurement Temperature (oC) 700oC 1000oC

PAGE 70

70 Figure 3-7: Specific contact resistance and sheet resistance under the contact as a function of long term thermal aging at 350oC. 0510152025 1x10-61x10-51x10-41x10-3 0 10 20 30 40 50 Specific Contact Resistance (/cm 2 )Days CrB2 W2B Sheet Resistance ()

PAGE 71

71 Figure 3-8: Depth profiles of W2B-based contacts (a) as depo sited (b) annealed at 600 oC (c) annealed at 700 oC and aged at 350oC and (d) annealed at 1000 oC 0 20004000 0 20 40 60 80 100 Ti Ni (b) N O W B Au Atomic Concentration (%)Depth () 01000200030004000 0 20 40 60 80 100 Au Au W Ga Ni Ti N B (a) Atomic Concentration (%)Depth () 0500100015002000 0 20 40 60 80 100 (b) N Ga W B Au Atomic Concentration (%)Depth () 0 10002000 0 20 40 60 80 100 Ni Ti (b) N Ga W B Au Atomic Concentration (%)Depth ()

PAGE 72

72 700750800850900950 1E-5 1E-4 1E-3 Contact Resistance Contact Resistance (ohm-cm2)Temperature0 20 40 60 80 100 Sheet Resistance Sheet Resistance Figure 3-9: Specific contact resistivity of Ni/Au/TiB2/Ti/Au Ohmic contacts and p-GaN sheet resistance under the contact as a function of annealing temperature.

PAGE 73

73 Figure 3-10: Secondary electron images of Ni/Au/TiB2/Ti/Au contact pads on p-GaN asdeposited (top) or after annealing at either 800(ce nter) or 900C (bottom).

PAGE 74

74 Figure 3-11: Surface scans of Ni/Au/TiB2/Ti/Au Ohmic contacts on p-GaN as a function of anneal temperature. The as-deposited sample is at top, that annealed at 800C at center and that at 900C at bottom 50 0 1000 1500 2000 -6000 -5000 -4000 -3000 -2000 -1000 0 1000 2000 3000 4000 Kinetic Energy (eV ) c/ s Ci O S Au Au Au Au Au Au Au Atomic % Au 3 60 0 C1 36.2 S1 2.4 O1 1.4 500 1000 1500 2000 -12000 -10000 -8000 -6000 -4000 -2000 0 2000 4000 6000 8000 Kinetic Energy (eV ) c/s C O Ti Ti Au Au Au S Bi B Atomic % C1 38.1 O1 30. 6 Ti1 18.7 Au3 7.7 B2 4.5 S1 0.4 500 100 0 1500 2000 1.5 -1 0.5 0 0.5 1 x 10 4 K inetic Energy (eV) c/ s C O B B Ti Au Au Ti Au Atomic % O1 33. 5 Ti1 27.8 C1 27.2 Au3 5.8 B 2 5 7

PAGE 75

75 Figure 3-12: Depth profiles of Ni/Au/TiB2/Ti/A u Ohmic contacts on p-GaN as a function of anneal temperature. The as-deposited sample is at top, that annealed at 800 pC at center, and that at 900 oC at bottom. 0 500 1000 1500 2000 0 10 20 30 40 50 60 70 80 90 100 Sputter Depth ( ) A tomic Concentration (%) Ti1.ls1 O1.ls2 N1.ls3 Ti2.ls 5 Ga1.ls 7 A u3.ls 8 B2.ls 9 0 500 1000 1500 0 10 20 30 40 50 60 70 80 90 100 Sputter Depth ( ) A tomic Concentration (%) Ti2.ls1 N1.ls2 B2.ls3 A u3.ls4 Ni1.ls 5 Ga1.ls6 0 500 1000 1500 2000 2500 3000 0 10 20 30 40 50 60 70 80 90 100 Sputter Depth ( ) A tomic Concentration (% ) Ti1.ls1 O1.ls2 N1.ls3 Ti2.ls 5 Ni1.ls6 Ga1.ls7 A u3.ls8 B2.ls 9 A u3.ls8

PAGE 76

76 750 800 850 900 10-410-310-2 Contact Resistance Ni/Au/ZrB2/Ti/Au ZrB2/Ti/Au Contact Resistance (-cm)Temperature (oC)3 4 5 6 7 8 9 10 11 12 Sheet Resistance Ni/Au/ZrB2/Ti/Au ZrB2/Ti/Au Sheet Resistance () Figure 3-13: Specific contact resistance of Ni/Au/ZrB2/Ti/Au and ZrB2/Ti/Au Ohmic contacts and p-GaN sheet resistance under the contact as a function of annealing temperature.

PAGE 77

77 Figure 3-14: Surface scans and depth profiles of Ni/Au/TiB2/Ti/Au Ohmic contacts on p-GaN as a function of anneal temperature. 500 1000 1500 2000 -8000 -6000 -4000 2000 0 2000 4000 Kinetic Energy (eV) c/s C O S Au Au Au Au Au Au Au A tomic % A u3 52.1 C1 44.2 S1 2.2 O1 1.5 Sample #1 As deposited 500 1000 1500 2000 -8000 -6000 -4000 -2000 0 2000 4000 6000 8000 c/s C O Al Al Ni Au Au Ti Au Au Z r Z r A tomic % C 1 44.9 O 1 18.9 A u3 15.7 Z r2 14.0 A l2 2.9 T i1 2.2 N i1 1.4 Sample #2 Annealed at 750C Kinetic Energy (eV) 0 500 1000 1500 2000 2500 3000 0 10 20 30 40 50 60 70 80 90 100 Sputter Depth ( ) Atomic Concentration (%) C O Au Z r B Ni Ti Ga Au 0 1000 2000 3000 4000 5000 0 10 20 30 40 50 60 70 80 90 100 Atomic Concentration (%) C Ti O Z r Ni Ga A u B Sputter Depth ( )

PAGE 78

78 Figure 3-15: Scanning electron mi croscopy images of Ni/Au/TiB2/Ti/Au contact pads on p-GaN as-deposited (top) or after annealing at either 750 (center) or 800C (bottom).

PAGE 79

79 Figure 3-16: Surface scans and depth profiles of ZrB2/Ti/Au Ohmic contacts on p-GaN as a function of anneal temperature. 500 1000 1500 2000 -5000 0 c/s C O S Au Au Au Au Au Au Au A tomic % A u3 59.1 C1 36.4 S1 3.1 O1 1.3 Sample #4 As deposited 5000 -10000 Kinetic Energy (eV) 500 1000 1500 2000 -6000 0 2000 4000 c/s C N O Mg Al Al Z r Z r Au Au Au A tomic % C1 52.9 O1 16.3 Zr2 12.0 A u3 8.9 M g2 4.2 A l2 4.0 N 1 1.7 Sample #6 Annealed at 900C -8000 -10000 6000 Kinetic Energy (eV) -2000 -4000 0 500 1000 1500 0 10 20 30 40 50 60 70 80 90 100 C Ti O Z r Ga Au B Atomic Concentration (%)Sputter Depth ( ) 0 500 1000 1500 2000 0 10 20 30 40 50 60 70 80 90 100 Atomic Concentration (%) Au B Ga Z r O Ti C Sputter Depth ( )

PAGE 80

80 Figure 3-17: Elemental maps obtained from scanning AES of ZrB2/Ti/Au Ohmic contacts pads on p-GaN. 0.0 130400 0.0 25900 Au3 Ga1 0.0 47.0 Ga1+Au3 151.0 191.0 SEM

PAGE 81

81 500550600650700750800 1.6 2.0 2.4 2.8 3.2 3.6 Contact Resistance Contact Resistance (10-3 -cm2)Temperature (oC)60 80 100 120 140 Sheet Resistance Sheet Resistance (/ Figure 3-18: Specific contact resistivity of W2B/Ti/Au Ohmic contacts and measured p-GaN sheet resistance under the contact as a function of annealing temperature.

PAGE 82

82 Figure 3-19: AES surface scans of W2B/Ti/Au Ohmic contacts on p-GaN as a function of anneal temperature. 500 1000 1500 2000 -6000 -5000 -4000 -3000 -2000 -1000 0 1000 2000 3000 4000 051535101 .spe Kinetic Energy (eV)c/s C N Au Au Au Au Au Au AuAtomic % Au3 58.7 C1 37.5 N1 3.9Sample #1 As deposited 500 1000 1500 2000 -10000 -8000 -6000 -4000 -2000 0 2000 4000 6000 8000 051535201.spe Kinetic Energy (eV) c/s C O Ti AuAtomic % C1 43.3 Au3 27.9 O1 24.0 Ti2 4.8Sample #2 500C annealed Surface survey 500 1000 1500 2000 -1.5 -1 -0.5 0 0.5 1 x 10 4 051535301.spe Kinetic Energy (eV)c/s C O Ti Au Au Au S TiAtomic % C1 44.0 O1 33.7 Au3 14.7 Ti2 7.1 S1 0.5Sample #3 700C annealed Surface survey 500 1000 1500 2000 -1.5 -1 -0.5 0 0.5 1 x 104 051535401.spe Kinetic Energy (eV)c/s C O S Ti Ti Au AuSanmple #4 800C annealed Surface surveyAtomic % C1 38.1 O1 33.6 Ti2 15.1 Au3 12.7 S1 0.5

PAGE 83

83 Figure 3-20: Depth profiles of W2B/Ti/Au Ohmic contacts on p-GaN as a function of anneal temperature. 0 500 1000 1500 2000 0 10 20 30 40 50 60 70 80 90 100 051535103.pr o Sputter Depth ( )Atomic Concentration (%) C1.ls1 O1.ls2 N1.ls3 Ti2.ls5 W2.ls6 B1.ls7 Au3.ls8 Ga1.ls9 0 500 1000 1500 2000 2500 0 10 20 30 40 50 60 70 80 90 100 051535203 .pro Sputter Depth ( )Atomic Concentration (%) C1.ls1 N1.ls2 O1.ls4 Ti2.ls5 W2.ls6 B1.ls7 Au3.ls8 Ga1.ls9 0 500 1000 1500 2000 2500 0 10 20 30 40 50 60 70 80 90 100 051535303.pro Sputter Depth ( )Atomic Concentration (%) C1.ls1 N1.ls2 O1.ls4 Ti2.ls5 W2.ls6 Au3.ls7 Ga1.ls8 B1.ls9 0 500 1000 1500 2000 2500 0 10 20 30 40 50 60 70 80 90 100 051535403.pro Sputter Depth ( )Atomic Concentration (%) C1.ls1 N1.ls2 O1.ls4 Ti2.ls5 W2.ls6 Au3.ls7 Ga1.ls8 B1.ls9

PAGE 84

84 -3-2-10123 1E-10 1E-9 1E-8 1E-7 1E-6 1E-5 1E-4 1E-3 0.01 0.1 As dep 300oC 500oC 700oC Current (A)Voltage (V) Figure 3-21: Current-voltage cu rves for Ni/Au/Ir/Au contacts.

PAGE 85

85 -3-2-10123 1E-9 1E-8 1E-7 1E-6 1E-5 1E-4 1E-3 0.01 As dep 300oC 500oC 700oC Current (A)Voltage (V) Figure 3-22: Current-voltage cu rves for Ni/Ir/Au contacts

PAGE 86

86 0100020003000 0 20 40 60 80 100 Au N Ga Ni Ir Au Atomic Concentraton (%)Depth ()(a) 010002000300040005000 0 20 40 60 80 100 Au N Ga Ni Ir Au Atomic Concentraton (%)Depth ()(b) 01000200030004000 0 20 40 60 80 100 O N Ga Ni Ir Au Atomic Concentraton (%)Depth ()(c) Figure 3-23: Depth profiles for Ni/Au/ Ir/Au contacts (a) annealed at 300 oC (b) annealed at 500 oC and (c) annealed at 700 oC.

PAGE 87

87 0 10002000 0 20 40 60 80 100 O N Ga Ni Ir Au Atomic Concentraton (%)Depth ()(a)0 1000 2000 0 20 40 60 80 100 N Ga Ni Ir Au Atomic Concentraton (%)Depth ()(b)0100020003000 0 20 40 60 80 100 N Ga Ni Ir Au Atomic Concentraton (%)Depth ()(c) Figure 3-24: Depth profiles for Ni/I r/Au contacts (a) annealed at 300 oC (b) annealed at 500 oC and (c) annealed at 700 oC.

PAGE 88

88 Figure 3-25: Scanning electron microsc opy images of Ni/Au/Ir/Au contacts

PAGE 89

89 CHAPTER 4 OHMIC CONTACTS TO N-GAN In this chapter, we report results on Ohm ic contacts based on these materials to n-GaN. Specific contact resistance has been studied as a function of annealing temperature, annealing time, and thermal aging for TaN, TiN, and ZrN diff usion barriers. All nitrides show comparable specific contact resistance to conventional Ti /Al/Pt/Au Ohmic contacts and display similar stability when aged at 350C in air. 4.1 Experiment The samples consisted of 3 m thick Si-doped GaN grown by Metal Organic Chemical Vapor Deposition on c-plane Al2O3 substrates. From Hall measurements, the electron concentration was ~3 x 1017 cm-3. Mesas 0.6 m deep were formed by Cl2/Ar inductively coupled plasma etching to provide for electrical isolation of the contact pads and to prevent current spreading. Deposition of a Ti (200 )/Al (800 )/X (500 )/Ti (200 )/Au (800 ) metallization was performed by sputtering, where X is TaN, TiN, or ZrN. Sputtering was performed using a Ar plasma-assisted system at a pressure of 5 mTorr and rf (13.56 MHz) powers of 150-350W. For comparison, a Ti (200 )/Al (800 )/Pt (400 )/Au (800 ) was deposited by e-beam evaporation. The contacts were patterned by lift off and annealed at 4001000C for 30-240s in a flowing N2 ambient in a rapid therma l annealing (RTA) furnace. Specific contact resistances were obtained using the linear transmission line (L-TLM) method. 100 m x 100 m pads with spacings of 80, 40, 20, 10, and 5 as well as 76, 36, 16, 6, and 1 m were used. The total resistance is given by RT = 2 RC + Rs ( L / W ) (4.1)

PAGE 90

90 where RC is the contact resistance, RS the sheet resistance, L the pad spacing, and W the pad width. A plot of RT vs pad spacing provides both RC, from the y-intercept, and RS, from the slope. The specific contact resistance is then calculated according to the equation c= RC 2W2/ Rs(4.2) All current-voltage plots were measured us ing an Agilent 4145B Parameter Analyzer. Auger electron spectrosc opy (AES) depth profiling of the contacts was peformed with a Physical Electronics 660 scanning Auger micr oprobe. The electron beam conditions were 10 keV, 1 A beam current at 30 from the sample nor mal. For depth profiling, the ion beam conditions were 3 keV Ar+, 2.0 A, (3 mm)2 raster. Before AES pr ofiling, scanning electron microscopy images (SEMs) were acquired from the samples. The SEMs were obtained at magnifications of 125x and 1000x and were used to locate and document analysis area locations and the surface morphology. Quantification of elements was achieved using the elemental sensitivity factors. 4.2 Results and Discussion Figure 4-1 shows the contact resistances as a fu nction of anneal temperature for the nitridebased Ohmic contacts as well as the conventio nal Ohmic contact. As deposited, the contacts display slight curvature in the current-voltage curves. Upon annealing at 600-1000C, the IV characteristics transform to Ohmic behavior. Th e specific contact resistances are stable up to 1000C anneals, except for the TiN, which show s increased specific contact resistivity at 1000C. The minimum specific contact resistance achieved was ~6 x 10-5 .cm2 for the TiNbased contacts at an anneal of 800C. This is likely due to diffusion of additional Ti to the surface to form the TiN phase. The ZrN-based c onract appears to be the most stable with increased annealing temperature, maintain ing a fairly constant value of ~1 x 10-4 for the range

PAGE 91

91 600-900C. These specific contact resi stance values are expected for the moderately low doped samples used in this study. This is because the current transport mechanism for Ti/Al-based Ohmic contacts to n-GaN is tunneling and is governed by the relation (4.3) Obviously, this relation shows a strong de pendence on the doping of the n-GaN. For doping levels an order of magnitude higher, 3 x 1018 cm-3, we would expect a decrease in the contact resistance of approximately an order of magnitude. Figure 4-2 shows the SEM images of the contacts as a function of anneal temperature. The as-deposited samples for TiN and ZrN show a smooth morphology while bubbles are present on the TaN as-deposited sample, possibly due to trapped sputter gases. At 600oC and 800oC anneals the contacts maintain a reasonable morphology. Furthermore, there appears to be little degradation of the surfac e even after aging at 350oC. All contacts show good edge acuity, a concern for devices with sm all contact separations. Figures 4-3,4, and 5 show the AES depth prof iles of the nitride-based contacts. As deposited, all contacts contain a significant amount of oxygen incor poration in the Ti and nitride layers. This is due to the relatively high base pressure of the sputte ring system of ~2 x 10-6 Torr. Note that despite this, the contact resistances are not adversely affected when compared to the Ti/Al/Pt/Au contacts, which were deposited in an electron beam evaporati on operating at a base pressure of ~3 x 10-7 Torr. As deposited, the metallic in terfaces are sharp and there is little intermixing between the layers. Upon annealing at 600C, Al has outdiffused to the surface and some Au has indiffused to the interface. At higher temperatures, the amount of interdiffusion increases. The amount of interdiffusion is much less than that seen with a standard metallic

PAGE 92

92 diffusion barrier, such as Ni. Fi gure 4-6 shows a profile for a Ti /Al/Ni/Au contact annealed at 500oC. From the profiles, the nitrides appear to be breaking down somewh at, with TiN the least affected, as the location of the Zr and Ta peaks do not correspond precisely to those of the N. Little change in the contact structure is seen with long term aging even after greater than 20 days. Because there is only minimal intermixing of the layers, little of the viscous AlAu4 phase has formed compared to conventional contact schemes. Figure 4-7 shows the dependence upon anneal time at 800C. Upon annealing for up to 180 seconds, an increase in contact resistance is s een in each of the contacts. The standard Ti/Al/Pt/Au contact exhibit the largest increas e in contact resistance, while the contacts containing TaN and TiN display more moderate in creases. The ZrN-based contact displays only a slight increase in contact resistance even upon annealing for 180 seconds. Figure 4-8 displays the specific co ntact resistance as a function of long term thermal aging. Samples placed on a hot plate at 350C display all display a gra dual increase in the specific contact resistance throughout the period. Table 4-1 shows the pe rcentage increase in specific contact resistance for each scheme. Compared to the Ti/Al/Pt/Au contacts, the ZrN and TaN contacts display better stability, with the spec ific contact resistance in creasing only about 100%. The TiN contacts displayed a larger increase. 4.3 Conclusions Ohmic contacts to n-GaN were fa bricated using a Ti/Al/X/Ti/A u scheme, where X is either TaN, TiN, or ZrN. These contacts display similar specific contact resistances to the conventional Ti/Al/Pt/Au contacts deposited on the same wafer, ~10-4 ohm-cm2 over a wide range of annealing temperatures. In a ddition, the TaN and ZrN contacts display only about half the increase in specific contact resistance that is seen in the Ti/Al/Pt/Au contacts. The AES depth profiles confirm that little intermixing has occu rred during the aging process. Because of this,

PAGE 93

93 these contacts show promise for use in high temperature applications. While the contact resistance does not seem to be better than that of conventional contacts af ter aging, there is far less mixing between layers. This is a primary concern, due to the AlAu4 phase, for devices with short spacings for high temperatur e applications. Further evaluation in devices such as HEMTs is needed in order to determine to potential of th ese contacts. In addition, the effect of oxygen incorporation in the nitride should be examined.

PAGE 94

94 Table 4-1: Percent change in specific contact resistance during thermal aging Contact Percent Change Ti/Al/Pt/Au 236% Ti/Al/TaN/Ti/Au 107% Ti/Al/TiN/Ti/Au 330% Ti/Al/ZrN/Ti/Au 091%

PAGE 95

95 Figure 4-1: Specific contact resistance as a function of anneal temperature 6008001000 1x10-51x10-41x10-31x10-2 Ti/Al/Pt/Au Ti/Al/ZrN/Ti/Au Ti/Al/TiN/Ti/Au Ti/Al/TaN/Ti/Au Specific Contact Resistance (-cm 2 )Anneal Temperature (C)

PAGE 96

96 TaN TiN ZrN TaN-based as deposited TiN-based as deposited ZrN-based as deposited TaN-based 600oC anneal TiN-based 600oC anneal ZrN-based 600oC anneal TaN-based 800oC anneal TiN-based 800oC anneal ZrN-based 800oC TaN-based aged TiN-based aged ZrN-based aged Figure 4-2: Scanning elect ron microscopy images of annealed contacts

PAGE 97

97 Figure 4-3: Depth profiles of Ti/Al/TaN/Ti/Au contacts (a) as deposited (b) annealed at 600oC. (c) annealed at 800oC and (d) annealed at 800oC and aged at 350oC. 0100020003000400050006000 0 20 40 60 80 100 (d) Ti Au N Ti O Au Ga O N TaAtomic Concentratioon (%)Depth ()Al0100020003000400050006000 0 20 40 60 80 100 (a) N Ti C O Au Ga O Ti N TaAtomic Concentratioon (%)Depth ()Al0100020003000400050006000 0 20 40 60 80 100 N Ti O Au Ga O N TaAtomic Concentratioon (%)Depth ()Al (c)01000200030004000500060007000 0 20 40 60 80 100 (b) Ta Ga O N Al N Al AuAtomic Concentration (%)Depth () Ti

PAGE 98

98 Figure 4-4: Depth profiles of Ti/Al/TiN/Ti/Au contacts (a) as deposited (b) annealed at 600oC. (c) annealed at 800oC and (d) annealed at 800oC and aged at 350oC. 0100020003000400050006000 0 20 40 60 80 100 (a) Ti C O Au Ga O Ti N NAtomic Concentratioon (%)Depth ()Al0100020003000400050006000 0 20 40 60 80 100 (b) N Ti O Au GaAtomic Concentratioon (%)Depth ()Al01000200030004000500060007000 0 20 40 60 80 100 (c) Ti O Au Ga O Ti N NAtomic Concentratioon (%)Depth ()Al01000200030004000500060007000 0 20 40 60 80 100 (d) Ti O Au Ga O Ti N NAtomic Concentratioon (%)Depth ()Al

PAGE 99

99 Figure 4-5: Depth profiles of Ti/Al/ZrN/Ti/Au contacts (a) as deposited (b) annealed at 600oC. (c) annealed at 800oC and (d) annealed at 800oC and aged at 350oC. 0100020003000400050006000 0 20 40 60 80 100 (a) N Ti C O Au Ga O Ti N ZrAtomic Concentratioon (%)Depth ()Al010002000300040005000 0 20 40 60 80 100 (c) N Ti O Au Ga Ti N ZrAtomic Concentratioon (%)Depth ()Al0100020003000400050006000 0 20 40 60 80 100 (d) N Ti O Au Ga Ti N ZrAtomic Concentratioon (%)Depth ()Al0100020003000400050006000 0 20 40 60 80 100 (b) Al Au N Ti C O Au Ga O Ti N ZrAtomic Concentratioon (%)Depth ()Al

PAGE 100

100 Figure 4-6: Depth profile of T i/Al/Ni/Au contact annealed at 500oC.0100020003000 0 20 40 60 80 100 Depth ()NAtomic Concentration (%)Au Ni Ga Ti Al O Ni-500oC

PAGE 101

101 Figure 4-7: Specific contact resistan ce as a function of anneal time. 6080100120140160180 1x10-51x10-41x10-3 Ti/Al/Pt/Au Ti/Al/TaN/Ti/Au Ti/Al/TiN/Ti/Au Ti/Al/ZrN/Ti/Au Specific Contact Resistance (-cm 2 )Anneal Time (s)

PAGE 102

102 Figure 4-8. Specific contact resistance as a function of long term thermal aging. 024681012141618202224 1x10-51x10-41x10-3 Ti/Al/Pt/Au Ti/Al/TaN/Ti/Au Ti/Al/TiN/Ti/Au Ti/Al/ZrN/Ti/Au Specific Contact Resistance (-cm2)Days

PAGE 103

103 CHAPTER 5 BORIDE-BASED SCHOTTKY CONTACTS TO P-GAN 5.1 Introduction The unique combination of wide bandgap, hi gh breakdown field, high saturation velocity, and ability to form high-quality heterostructur es with good transport properties make GaN an ideal candidate for several device applicati ons including light emitting diodes (LEDs) for displays, laser diodes for data storage, and hi gh electron mobility transistors (HEMTs) for highpower and high frequency electronics. An important aspect for improved reliability of GaNbased devices is the development of more therma lly stable Ohmic and Schottky contacts to both n-type and p-type GaN. This aspe ct has been one of the major issues in the operating lifetime of III-nitride lasers and LEDs and it is yet to be overcome completely. Another example is AlGaN/GaN HEMTs in advanced power amplifiers and converters for which the development of Ohmic and Schottky contacts ensuring long-time operation under uncooled a nd high temperature conditions without metal spiking or loss of edge acuity remains a challenge. One approach that has proven promising for contact formation on n-GaN is the use of very high melting temperature metals like W and WSix [24,68,71-74]. Recently, other promising metallurgies based on transition metal borides like W2B, W2B5, ZrB2, CrB2, and TiB2 have also been proposed [75-78]. The interest in borides is due to their high melting temperature (e.g. 3200C for ZrB2) and high thermodynamic stability. For p-GaN, most rectifying contact studies have been performed with simple metals such as Ni Pd, Ti, Cr, and Pt, [17, 54-66] and studies on boride-based metallurgies are still in their infancy. In this work, we report on the annealing and measurement temperature dependence of W2B and W2B5-based rectifying contacts to p-GaN. The barrier heights determined by X-Ray Photoelectron Spectroscopy (XPS) are in the rang e 2.7-2.9 eV. These contacts are found to be

PAGE 104

104 reasonably stable upon annealing to ~600-700C. While current-voltage characteristics as a function of measurement temperature suggest thermionic field emission (TFE) over a Schottky barrier as the most dominant mechanism of fo rward current flow, the bias and measurement temperature dependence of the reve rse-bias current indicates that leakage must originate from surface leakage or generation in the depletion layer. 5.2 Experimental Details The p-GaN samples were 1 m-thick Mg-doped GaN layers grown by metal organic chemical vapor deposition (MOCVD) on 1 m-thick undoped buffers on c-plane Al2O3 substrates. The hole concentration obtained fr om Hall measurements after acceptor activation annealing was ~1017 cm-3. The surface was cleaned by sequent ial rinsing in acetone, ethanol and 10:1 H2O:HCl prior to insertion in the depositio n chamber. Boride(500)/Pt(200)/Au(800), where the boride was either W2B or W2B5, was used as the Schottky metallization scheme. Au was added to lower the contact sheet resistance while Pt is a diffusion barrier. All metals or compounds were deposited by Ar plasma-assisted rf (13.56 MHz) sputtering at a pressure of 15 mTorr and rf powers of 250-400 W. The contacts were patterned by liftoff of lithographicallydefined photoresist and annealed at temperat ures up to 700C for 1 min in a flowing N2 ambient in a RTA furnace. For Ohmic contacts, we used Pt/Au annealed at 500C in O2 for 30s prior to deposition of the Schottky metallization. Ring contact geometry for the diodes was employed, with the Schottky contacts surrounded by the Ohmic contacts. The Schottky contact dots were 40 m in diameter, the surrounding was 60 m in inner diameter and 70 m in outer diameter. XPS, current-voltage (I-V), and auger electron spectroscopy (AES) were used to characterize boride-based rectifyi ng contacts. XPS measurements were taken with a Physical Electronics 5100LSci spectrometer with an al uminum x-ray source (energy 1486.6 eV). High-

PAGE 105

105 resolution spectra were acquired to determine the binding energy (i.e. chemical state) and concentration of specific elements observed in the survey spectra. Charge correction was performed by using the known posit ion of the C-(C,H) lin e in the C 1s spectra at 284.8 eV. The SBH was determined from the binding energy of the Ga 3d core level EB and the energy difference between that core le vel and the valence band maximum EVC according to BBVCEE .[62,79] The I-Vs were recorded over the temperature range 25-200C using a probe station and an Agilent 4145B parameter analyzer. For the relativ ely high doped p-GaN samp les investigated in this work, TFE over a Schottky barrier is expect ed to be the dominant mechanism of forward current flow [56,57,62,63,80]. We fit the forward I-V characteristics to the relation for TFE over a barrier [81] 0 0expFeV JJ E (5.1) where JF is the current density, e is the electronic charge, and V is the applied voltage. In Eq. (1), 00000cothB E EEkT is the characteristic energy re lated to the tunneling probability, kB and T being the Boltzmanns constant and absolute temperature, respectively. For p-GaN, 12 3 007.7510AEeV Ncm, [56,62] NA being the density of acceptors. The saturation current density J0 is given by 0.5 00 0 00 0exp coshB B BBBATEqV e e J kEkTkTE (5.2) where A*=103.8 Acm-2K-2 is the effective Richardsons constant for p-GaN, [82] B is the barrier height, and =(EF-EV)/e is the difference between the valence band maximum and the position of the Fermi level. This parameter depends on the doping level, but, as will be shown

PAGE 106

106 later, it only slightly influences the SBH of W2B and W2B5-based rectifying contacts to p-GaN. Nakayama et al. [83] have suggested that a hole concentration of 21017 cm-3 in p-GaN samples leads to =0.13 V. This hole concentration is very clos e to that achieved in the present study so that we also assume =0.13 V. Auger electron spectrosc opy (AES) was used to analyze th e depth profiles. A Physical Electronics 660 Auger Microprobe Electron Beam at 10 keV, 0.3 A, 30 from sample normal was used for the data collection while for profiling the ion beam conditions were 3 keV Ar+, 2.0 A, (4 mm)2 raster with sputter-etch rates of 80, 40, and 104 /min for the boride, Pt and Au layer respectively. A survey spectrum (a plot of the first derivative of the number of electrons detected as a function of energy) was used to determine the composition of the outer few nanometers of each sample. Quantifications we re accomplished by using elemental sensitivity factors. 5.3 Results and Discussion Figure 5-1 presents an example of the Ga 3d core level and the valence band spectrum collected on a p-GaN surface without a borid e overlayer. From Figure 5-1, we find EVC=17.54 eV, which is in reasonable agreement with the value of 17.8 eV reported previously. [62,84] After W2B deposition, Figure 5-1 shows that the binding energy of the Ga 3d core level is EB=20.25 eV, thereby yielding B=2.71 eV. Similarly, we find EB=20.41 eV for W2B5, which gives B=2.87 eV. These results are similar to the 2.68-2.78 eV values reported by Yu et al. [56] for Ni/p-GaN. Based on the Schottky-Mott model [52], the SBHs for n-GaN and p-GaN should add up to the GaN band gap 3.4 eV. For W2B and W2B5-based rectifying c ontacts to n-GaN, Khanna et al. [76,77] have reported B~0.55 eV, thereby suggesting B~2.85 eV for p-GaN. Our XPS results are clearly in excellent agreem ent with the predictions of this model.

PAGE 107

107 The I-V characteristics as a function of the measurement temperature for the W2B-based diodes are shown in Figure 5-2. All log I vs V curves measured between 25 and 200C exhibit a linear region at low forward bi as (up to ~2.0 V) and a less steep region at higher voltages. Furthermore, all log I vs V curves are parallel. As noted in refs. [56,57,62,63] for p-GaN samples with relatively high doping levels (>1017 cm-3), this latter result is inconsistent with the thermionic emission model but typical for car rier transport with a dominant tunneling component. Note that Shirojima et al. [61] have reported accurate determination of the SBH using the TE model. However, the acceptor conc entration in their p-GaN samples was relatively low (~1016 cm-3). In our case, the Mg concentration is expected to be in the 1019 cm-3 range, thus confirming that the forward current cannot be an alyzed using the thermionic emission model. Figure 5-3 displays the I-V characteristics obtained from th e diodes annealed at different temperatures. The extracted characteristic ener gy related to the tunneling probability as a function of annealing temperature is shown in Figure 5-4. For W2B, the value of E0 is seen to remain stable in the 120 meV range and then increases beyond ~500C. This value of E0 corresponds to an acceptor density of about 21020 cm-3, which is higher than that expected from the Mg concentration alone (E0~25meV for NA~1019 cm-3). This can probably be attributed to the presence of acceptor-like deep level defects induced by both the high Mg doping [63,85,86] and the sputter-deposition proce ss [29,87]. In the case of W2B5, a slight decrease of E0 is observed at intermediate annealing temperatur e which can probably be attributed to the annealing of sputterdamage [29,87]. The influence of annealing on the SBH derived from the I-V measurements of W2B/Pt/Au and W2B5/Pt/Au contacts is presented in Figure 5-5. Over the whole range of temperatures investigated, the SBH only slightly decreases with increasing annealing temperature. However,

PAGE 108

108 these values are unphysical, being larger than th e GaN band gap, and are inconsistent with both our XPS measurements and the pred ictions of the Schottky-Mott mode l. One may argue that this discrepancy results from the value of used in our calculations. However, it is shown in Figure 5-6 that over the wide range of values investigated, this parameter only slightly influences the SBH. The discrepancy between XPS and I-V measurements is more likely related to the presence of an interfacial defect layer or due to the pres ence of an interfacial oxide layer. Indeed, it was shown by Shiojima et al. [88] using high-temperature isothe rmal capacitance transient spectroscopy that the de fects induced by the high Mg doping in p-GaN are essentially located at the surface vicinity and that this region acts as a series capacitance [88,89]. Therefore, when such defect or oxide layer is present, the carriers have to tunnel through an additional barrier, thus resulting in an apparent SBH given by 0 BBB where 0 B is the true SBH between the contact metal and p-GaN and where B is the increment of the SBH due to the presence of the thin defect or oxide layer. Our sepa rate capacitance-voltage measurements (C-V) have indeed shown that by accounting for such def ect layer, the SBH obtained from the C-V characteristics becomes similar to that determined by XPS [80]. We have calculated a corrected junction capacitance Ccorr for W2B/p-GaN from the equi valent circuit using the simple relation 111corr m oxCCC where Cm and Cox are the as-measured and oxide capacitances respectively. From our XPS meas urements, the true SBH for W2B/p-GaN is 0~2.7eVB. Using our C-V measurements, Figure 5-7 shows that the value of Cox required to obtain a SBH similar to that determined by XPS is Cox~0.70 nF. Assuming a contact area of ~2 10-9 m2 and a relative permittivity of ~10 for the oxide layer [90], we estimate an oxide layer thickness of ~0.25 nm, i.e. about one monolayer. For 20 m-diameter diodes, the value of Cox required to fit the XPS

PAGE 109

109 data was 4 times lower than that for 40 m-diameter diodes, as expected from the surface area ratio. In addition, the density of acceptors obtained from the corrected C-Vs was 1.8 1020 cm-3, which is very similar to the value determined above from the I-Vs. Table 5-1 summarizes the calculated barrier heights from XPS, IV with both thermionic and thermionic field emission, and CV measurements. Figure 5-8 presents the influence of the m easurement temperature on the reverse-bias current for the W2B-based diodes. All log I vs V curves measured between 25 and 200C exhibit a linear region at high reverse bias and a steeper variation at lowe r reverse voltages. In addition, as the measurement temperature increases, the re verse leakage current also increases. Given the large barrier height typical for rectifying contacts to p-GaN, one can assume that thermionic emission over the Schottky barrier makes only a ne gligible contribution to the reverse-bias current flow. This assumption is furt her supported by the f act that all log I vs V curves in Figure 5-8 are parallel, which is clearly inconsistent with the thermionic emission model. On the other hand, in the tunneling approach the reverse current density JL is given by the Fowler-Nordheim tunneling expression 2expLbbJCEED where C=C(B) and D=D(B) are parameters independent of the measurement temperature [91-93] For tunneling, the re verse-bias current is thus independent of the measurement temperature, which again is inconsistent with the experimental data displayed in Figure 5-6. Leakage must therefore originate from other mechanisms such as generation in the depletion layer or surface leakage. The large band gap of GaN makes the intrinsic carrier c oncentration in a depletion regi on very small, suggesting that contributions to the reverse-bias leakage from generation in the depletio n layer are negligible. However, given the large number of deep-level de fects located at the surf ace vicinity [88] as a result of both the high Mg doping [29,85,86] and the sputter-deposition process [29,87], one may

PAGE 110

110 expect these defects to act as generation centers when the carrier density is below its equilibrium value as in the reverse-bias regime of a Schottky diode. The breakdown voltage, VB, of the boride-based rectif ying diodes as a function of annealing temperature is shown in Figure 5-9 VB shows the opposite tren d to the characteristic energy related to the tunneling probability displayed in Figure 5-4, decreasing where E0 increases. As E0 is directly related to the density of acceptor-like deep-level defects, this observation is consistent with our expectations that leakage esse ntially results from a defectmediated mechanism. As observed for boride-base d rectifying contacts to n-GaN [75,76], the improvement of breakdown voltage at intermedia te annealing temperatures is likely to result from the annealing of sputter-damage in the near-surface region of GaN. At high annealing temperature, VB is seen to decrease, indicating that the diodes become very leaky. AES depth profiles of the as-deposited and a nnealed contacts are shown in Figure 5-10. For both diodes, the as-deposited layers exhibit relatively sharp interfaces, consistent with the featureless surface morphology observed with both optical and scanning electron microscopy. After high-temperature annealing, W2B-based contacts show the onset of metallurgical reaction in the contact scheme whereas W2B5 presents significant interdiffusion of layers. 5.4 Conclusions In summary, our XPS measurements have shown that W2Band W2B5-based rectifying contacts to p-GaN produce barrier he ight in the 2.7-2.9 eV range, in excellent agreement with the predictions of the Schottky-Mott model. By comparison, the SBH determined from the I-Vs using the TFE model are unphysically large and much higher than those deduced from XPS due to the presence of a defect layer acting as an add itional barrier to carrier transport. The apparent SBH only slightly decrease s upon annealing to 600-700 C due to the onset of metallurgical

PAGE 111

111 reaction with the GaN. This suggests that that the boride-based metallurgies may be promising for rectifying contacts to p-GaN where thermal stability is a critical issue. The bias and measurement temperature of the reverse leakage current indicate s that leakage must originate from surface leakage or generation in the de pletion region through deep-level defects.

PAGE 112

112 Table 5-1: Comparison of differe nt barrier height calculations XPS TE TFE C-V B = 2.7 eV B = 1.2 eV n = 4.4 B = 3.8 eV E0= 110 meV B = 4.1 eV

PAGE 113

113 24222018164 0 2 17.54 eV Valence band region (X5) Ga 3d Intensity (a.u.)Binding Energy (eV) 26242220181614 2 W2B5/ p-GaN W2B / p-GaN Ga 3d Intensity (a.u.)Binding Energy (eV) Figure 5-1: XPS spectra without (top) and with (bottom) a boride overlayer. The left-hand spectrum in the top figure corresponds to th e Ga 3d core level wh ereas the right-hand panel presents the spectrum of the valence band region.

PAGE 114

114 0.00.51.01.52.02.53.03.54.0 10-61x10-51x10-410-310-2 Current (A)Voltage (V) 200oC 150oC 100oC 50oC 25oC Figure 5-2: Forward current-vo ltage characteristic of W2B-based Schottky diodes as a function of measurement temperature.

PAGE 115

115 0123456 10-710-61x10-51x10-410-3 W2B/Pt/Au Current (A)Voltage (V) As-deposited 300oC anneal 600oC anneal 0123456 10-710-61x10-51x10-410-3 W2B5/Pt/Au Current (A)Voltage (V) As-deposited 300oC anneal 600oC anneal Figure 5-3: Forward current-vo ltage characteristic of W2B-based (top) and W2B5-based (bottom) Schottky diodes as a function of annealing.

PAGE 116

116 0100200300400500600700800 0 20 40 60 80 100 120 140 160 180 200 220 240 260 NA~5x1019cm-3 NA~1019cm-3 W2B/Pt/Au W2B5/Pt/Au E0 (meV)Anneal temperature (oC) Figure 5-4: Influence of the annealing temperat ure on the characteristic energy related to the tunneling probability. Dashed and dotte d lines correspond to the values of E0 for NA~1019 and 51019cm-3 respectively.

PAGE 117

117 0100200300400500600700800 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0 W2B/Pt/Au W2B5/Pt/Au B (V)Anneal temperature (oC) Figure 5-5: Influence of the annealing temper ature on the apparent Sc hottky barrier height derived from IV measurements.

PAGE 118

118 0.080.120.160.200.24 3.0 3.5 4.0 4.5 5.0 Lightly-doped Highly-doped W2B/Pt/Au W2B5/Pt/Au B (eV) (eV) Figure 5-6: Dependence of th e apparent Schottky barrier height on the parameter defined as the difference between the valence band maximum and the position of the Fermi level. Low and high

PAGE 119

119 -4-3-2-1012345 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 Voltage (V)1/C2 (1021 F-2) As-measured After oxide correction Figure 5-7: As-measured and afte r oxide correction dependence of C-2 versus V of Au/Pt/W2B/pGaN Schottky diodes. The measurement frequency was set to 1 kHz.

PAGE 120

120 -4.0-3.5-3.0-2.5-2.0-1.5-1.0-0.5 10-810-710-61x10-51x10-4 Current (A)Voltage (V) 200oC 150oC 100oC 50oC 25oC Figure 5-8: Reverse currentvoltage characteristic of W2B-based Schottky diodes as a function of measurement temperature.

PAGE 121

121 0100200300400500600700800 0 1 2 3 4 5 6 7 8 9 W2B/Pt/Au W2B5/Pt/Au VB (V)Anneal temperature (oC) Figure 5-9: Influence of the annea ling temperature on the breakdown voltage.

PAGE 122

122 0 50 100150200 0 10 20 30 40 50 60 70 80 90 100 W2B/Pt/Au as-deposited Ga N O Pt Au W B Atomic concentration (%)Sputter depth (nm)(a) 0 50 100150200 0 10 20 30 40 50 60 70 80 90 100 W2B5/Pt/Au as-deposited Ga N O Pt Au W B Atomic concentration (%)Sputter depth (nm)(b) 0 50 100150200 0 10 20 30 40 50 60 70 80 90 100 W2B/Pt/Au annealed @700oC Ga N O Pt Au W B Atomic concentration (%)Sputter depth (nm)(c) 0 50 100150200 0 10 20 30 40 50 60 70 80 90 100 W2B5/Pt/Au annealed @600oC Ga N O Pt Au W B Atomic concentration (%)Sputter depth (nm)(d) Figure 5-10: Depth profiles of W2B/Pt/Au contacts and W2B5/Pt/Au rectifying contacts (a,b) before and (c,d) after annealing at 600 C.

PAGE 123

123 CHAPTER 6 BORIDE AND IR BASED CONTACTS FOR L IGHT EMITTING DIODES 6.1 Introduction InGaN/GaN multiple quantum well light-emitting diodes (MQW-LEDs) are commercially available in a broad range of wa velengths for use in applications such as full color displays, traffic signals, and exterior ligh ting. There is also interest in shorter wavelength LEDs with AlGaN active regions to excite down conversion phosphors for white light [94]. Nevertheless, to compete with fluorescent and other high-efficiency lighting sources, it is essential to drive GaNbased LEDs at very high current densities to ma ximize light output. One drawback of the high current densities is self-heating of the he terostructure which can produce Ohmic contact degradation and generation of non-radiative recombination centers [95-98]. Cao et al. [99-101] demonstrated improved heat dissipation and cu rrent spreading by reduc ing the density of dislocations in the heterostructure through the use of free-stan ding GaN instead of the more common sapphire substrates. Further reduction of self-heating effects and improved efficiency and operating lifetime can be obt ained through the use of low-resistance and thermally stable Ohmic contacts to the p-GaN layer. Indeed, pow er dissipation by Joul e heating across the pGaN/metal interface at high current injection le vels was found to produce indiffusion of the Ohmic contact elements along disl ocations in the III-nitride-bas ed epilayers, leading to an electrical short of th e pn junction. [102,103] The conventional metallization schemes used for making Ohmic contacts to p-GaN for InGaN/GaN LEDs are based on high work function me tals such as Ni, Pd, Cr or Pt with an overlayer of Au to reduce the sheet resistance. Specific contact resistances of ~10-2-10-4 .cm2 are obtained by annealing at 450-650 C [106,107]. These contacts have stability problems at high temperature. For example, the initial Au/Ni/ GaN structure transforms to Ni/Au/GaN with a

PAGE 124

124 rough Ni surface after annealing at 600C [44,108] A way to prevent exce ssive intermixing and contact morphology degradation is to use a high -melting-point diffusion barrier in the contact stack [109-111]. TiB2, with a melting temperature of ~3000C, reasonable electrical resistivity (28 .cm) and thermal conductivity (26 W.m-1.K-1), and heat of formation comparable to those for silicides or nitrides [112], s hows promise as a diffusion barrier. In this letter, we report on the long-term annealing characteristics at 200-350C of InGaN/GaN MQW-LEDs with TiB2and Ir-based p-Ohmic contacts. This high-temperature stress stimulates accelerated aging of GaN-ba sed LEDs and gives an idea of the expected reliability of the Ohmic contacts. The use of TiB2 or Ir as a diffusion barrier in Ni/Au-based contacts is found to produce superior long-term stability of turn-on voltage, leakage current and output power. 6.2 Experimental The MQW-LED structures were grown by metal organic chemical vapor deposition on cplane sapphire substrates. The layer structure consisted of a low-temperature GaN buffer, 3 mthick n-GaN, 0.1 m n-AlGaN clad, three period undoped InGaN/GaN MQW active, 0.1 m pAlGaN clad, and 0.3 m p-GaN layer. 1 m-deep mesas were fabricated using Cl2/Ar inductively coupled plasma etchi ng. Ohmic contacts to n-GaN were formed by lift-off of e-beam deposited Ti (20 nm) / Al(80 nm) / Pt(40 nm) / Au (80 nm) subsequently annealed at 800C for 1 min in a flowing N2 ambient in a rapid thermal ann ealing (RTA) furnace. The first pmetallization scheme investigated consisted of Ni (50 nm) / Au (80 nm) / Ir (50 nm) / Au (80 nm). The second contact scheme was Ni (50 nm) / Au (80 nm) / TiB2 (50 nm) / Ti (20 nm) / Au (80 nm). Ti was added to im prove the adherence of Au on TiB2. The Ni/Au layers were deposited by e-beam evaporat ion while the Ir/Au and TiB2/Ti/Au overlayers were deposited by

PAGE 125

125 Ar plasma-assisted rf magnetron sputtering. For comparison, devices with Ni/Au contacts were also fabricated using the same wafer. All cont acts were patterned by liftoff of lithographicallydefined photoresist and annealed at 600C for 1 min in a flowing N2 ambient. Prior to their insertion in the e-beam chamber, the surface of both nand p-type GaN were cleaned in a 1HCl:10H20 solution for 1 minute. Figure 6-1 show s a typical optical micrograph of asfabricated MQW-LEDs with Ni/Au/TiB2/Ti/Au Ohmic contacts. Similar contact morphology was achieved using the other metallization schemes. All devices were first aged for a period of 10 days at 200C on a heater plate in air. The sa mples were then removed from the heater block, allowed to cool to room temperature, and charac terized before being returned for further 35 days aging at 350C. MQW-LEDs were analy zed by luminescence-current-voltage (L-I-V) measurements using a probe station and an Agilent 4145B parameter analyzer. The light output power was measured using a Si photodetector located at ~2 cm from the sample surface. Electroluminescence (EL) spectra were acq uired by a fiber optic spectrometer. 6.3 Results and Discussion Figure 6-2 presents the 300 K L-I characteristics from each as-fabricated device. Typical EL spectra measured for different forward currents are also shown in the inset. For an injected current of 500 A, the EL peak wavelength of the InGaN/GaN MQW-LEDs is 459 nm with a full width at half maximum (FWHM) of 26 nm. As the injected curren t increases, the peak exhibits a blue shift while the FWHM remains fa irly constant. This behavior results from the decrease by carrier screening of the quantum-confined Stark effect induced by the built-in piezoelectric field in the strain ed InGaN/GaN QWs [113]. The L-I characteristics are similar for all metallization schemes investigated, showing that the TiB2/Ti/Au or Ir/Au overlayers did not reduce the light output with respect to the mo re conventional Ni/Au c ontacts. The non-linear

PAGE 126

126 increase and relatively low output power can pr obably be attributed to enhanced current crowding and self-heating effects [114,115]. These effects are expected to be particularly important for the LED structures investigated in th is work due to the high density of dislocations in InGaN and GaN epilayers grown on sapphire as well as the low th ermal conductivity of sapphire [99-101]. As the dislocation density or LED junction temperature increases, confinement in the MQW becomes less efficient, l eading to premature satu ration of the emission power. I-V characteristics from MQW-LEDs with Ni/Au, Ni/Au/TiB2/Ti/Au, and Ni/Au/Ir/Au Ohmic contacts to the p-GaN layer are shown in Figure 6-3. For all diode s, the ideality factor was >2, consistent with other data reporte d previously [95-101,116,117]. There is still no agreement whether such high values result from tunnelling [101,117], recombination in the depletion layer [117], or rectifyi ng junctions present in heterost ructure diodes [118]. The turn-on voltage (voltage at which the current reaches 0. 05 mA) and reverse leakage current (measured at -5V) are given in Table 6-1. The uncertainty on each value was estimated from the deviation between different devices on the same wafer. Fo r as-fabricated devices, the turn-on voltage was 3.2-3.6V and the reverse leakage current was ~10-6 A. Such high leakage currents are commonly observed in GaN-based LEDs grown on sapphire substrates and esse ntially result from the high density of dislocations in such samples [101]. After 10 days agi ng at 200C, an increase of the turn-on voltage to ~4.1 V and a decr ease of the leakage current to ~10-9 was observed for all devices. After aging at 350C for an additional 30 days, MQW-LEDs with TiB2and Ir-based Ohmic contacts show turn-on volta ge and leakage current similar to those in the as-deposited state whereas there was serious degradation in LEDs fabricated with Ni/Au contacts. This reduction of injected current at a given forward voltage had a large impact on the EL output as

PAGE 127

127 shown in Figure 6-4. While the brightness of MQW-LEDs with Ni/Au Ohmic contacts at 10 V (I = 80 A) was low, optical images of aged MQW-LEDs with boride-based Ohmic contacts show much stronger emission, even at a lower applied voltage of 4.5 V (I = 300 A). Several mechanisms can be invoked to explain the drastic degradation of Ni/Au contacts upon aging, including formation of an islanded contact morphology, formation of reactions with the GaN resulting in a modification of the doping profile, and excessive intermixing of Ni and Au leading to oxidation on the rough Ni surface and an increase of th e series resistance [44,109,119]. Although the dominant degradation mechanism remains unknown, the use of TiB2/Ti/Au or Ir/Au overlayers on the Ni/Au-based Ohmic contacts clearly improves the long-te rm thermal stability of InGaN/GaN MQW-LEDs. 6.4 Conclusions In conclusion, TiB2 and Ir diffusion barriers in pOhmic contacts on GaN LEDs produced less change in turn-on voltage, leakage current, and output power after long-term annealing at 200C and 350C for 45 days. These schemes look pr omising for high temperature applications where improved stability over Ni/Au is mandatory. Table 6-1: Influence of long-term aging at 200 C and 350C on the turn-on voltage and reverse current of InGaN/GaN MQW-LEDs. Turn-on voltage (V) Reverse current @ -5V (A) Contact to p-GaN Day 0 Day 10 @ 200C Day 45 @ 350C Day 0 Day 10 @ 200C Day 45 @ 350C Ni/Au 3.30.3 4.90.4 9.10.5 (2)10-6 (6)10-9 (5)10-5 Ni/Au/TiB2/Ti/Au 3.50.2 4.10.5 4.00.2 (5)10-6 (2)10-9 (2)10-8 Ni/Au/Ir/Au 3.30.2 4.30.3 3.90.3 (3)10-6 (2)10-9 (3)10-7

PAGE 128

128 Figure 6-1: Optical micrograph of an as-fabricat ed MQW-LED. The p-cont act at the center of the diode is 80 m in diameter.

PAGE 129

129 0.0 0.2 0.4 0.6 0.8 1.0 0.0 0.2 0.4 0.6 0.8 1.0 -4-20246810 0.0 0.2 0.4 0.6 0.8 1.0 (a) Day 0 Day 10 @ 200oC Day 45 @ 350oC(b) (c) Current (mA)Voltage (V) Figure 6-2 : L-I characteristics of MQW-LEDs with Ni/Au, Ni/Au/TiB2/Ti/Au, and Ni/Au/Ir/Au p-Ohmic contacts. The inset shows emissi on spectra from as-fabricated LEDs at various injection currents.

PAGE 130

130 0.00.51.01.52.0 0.0 0.1 0.2 0.3 0.4 0.5 0.6 440460480500 0 4 8 12 16 20 Intensity (a.u)Wavelength (nm) 1.5 mA 1 mA 750 A 500 A Luminescence (W)Current (mA) Ni/Au Ni/Au/TiB2/Ti/Au Ni/Au/Ir/Au Figure 6-3: Influence of long-term aging at 250 C and 350C on the I-V ch aracteristics of LEDs with (a) Ni/Au, (b) Ni/Au/TiB2/Ti/Au, and (c) Ni/Au/Ir/Au p-Ohmic contacts.

PAGE 131

131 (a) (b) Figure 6-4: Image of aged LEDs with (a) Ni/Au and (b) Ni/Au/TiB2/Ti/Au p-Ohmic contacts. In (a), the picture was taken for a forward bias of 10 V (I = 80 A), while the forward voltage in (b) was 4.5 V (I = 300 A).

PAGE 132

132 CHAPTER 7 CONCLUSION Im proved device processing is necessary in orde r to realize the full potential of GaN based electronics and optoelectronics. While improved material quality is critical, especially for pGaN as well as related alloys such as InGaN to increase the efficiency of green emitting devices, it is not the only challenge. Developing relia ble, stable, low resistance Ohmic and Schottky contacts to both nand p-type GaN is still a challenge and hence remains of interest to the GaN community at large. Without contacts that can withstand elevated temperatures and other harsh environments, many of the properties that ma ke GaN a unique and desirable semiconductor mean nothing. The goal of this work was to develop such cont acts to both nand p-GaN. In order to improve upon existing contact schemes, it is necessary to achieve at least comparable contact resistances or barrier heights. In addition, the contact s should be able to withstand more stressful processing conditions, such as elevated anneali ng temperatures, while maintaining predictable and stable characteristics. Perhaps the most cr itical factor is the abil ity of the contacts to withstand long periods in a harsh environment, simulated here by placement on a hot plate at temperatures of 200oC and 350 oC. A final consideration is the amount of intermixing of the contact layers, especially in Ohmic contacts to n-GaN, as large amounts of intermixing can lead to undesirable phases, such as AlAu4 and issues with lateral flow. In addition, if any undesirable phases may form at device operation temperatures, or if the contacts themselv es interact with the GaN at these temperatures, devices will prove unreliable during prolonged operation. To tackle these issues, three separate materials system s were examined: borides, nitrides, and Ir.

PAGE 133

133 The first section of this dissertation involved the fabrication of Ohmic contacts to p-type GaN. All three material systems mentioned we re examined for this use. Contacts were fabricated of the following structures: 1. Nickel / Gold / X / Titanium / Gold, where X is a chosen boride or nitride 2. X / Titanium / Gold, where X is a chosen boride of nitride 3. Nickel / Gold / Ir / Gold 4. Nickel / Ir /Gold Each of these was then subjected to a nnealing at temperat ures ranging from 300oC to 1000oC in a flowing N2 ambient for 60 s. Schemes 1 and 3 consistently produced Ohmic contacts in the range of 500-1000 oC, 2 only inconsistently at specific temperatures, 3 at only 500oC and 4 not at all. Schemes 1 and 3 produced specific contact resistance values similar to those reported for Nickel / Gold contacts in the literature. Contacts of scheme 1 were subjected to long term thermal aging on a hot plate at both 200oC and 350oC. Nitride-based contacts failed early in aging, however boride based contacts displayed stable specific contact resistances throughout. Auger Electron Spectroscopy showed breakdown of the nitride structure during aging due to severe intermixing with the GaN. The borides showed minimal intermixing with the GaN, accounting for their stability. AES depth profiles of scheme 3 revealed severe intermixing of the contacts with the GaN at anneal temperatures above 500oC, accounting for their failure at elevated anneal temperatures. Contacts fabri cated with scheme 4 were not Ohmic due to the absence of Au at the surface to promote increased hole concentration as well as severe intermixing at high temperatures. The second section of the work dealt with the examination of nitride based contacts to nGaN. Contacts using the borides and Ir have previously been examined. Contacts of the

PAGE 134

134 structure Titanium / Aluminum / X / Titanium / Gold were fabricated using standard semiconductor processing procedures, where X is either TaN, TiN, or ZrN. For comparison, conventional contacts of the form Titanium / Aluminum / Platinum / Gold were fabricated as well. The contacts were then subjected to a range of anneal temperatures between 500oC and 1000oC in a flowing N2 ambient for 60 s, as well as anneals at 800oC for up to 180 s, in order to determine if the nitrides would allow for an increased thermal budget. Contacts were also aged on a hot plate at 350oC and aged for a period of 24 days. Current-voltage measurements, Auger Electron Spectroscopy, and Scanni ng Electron Microscopy were used to characterize the contacts. Current-voltage measurements of th e nitride contacts did not show any improvement over the performance offered by the conventional Pt-b arrier contacts in any of the experiments. However, AES depth profiles revealed that th e nitride based contac ts displayed far less intermixing than Ni-barrier contacts, even wh en the nitrides were annealed at higher temperatures. Further, the profiles of therma lly aged nitride contacts displayed no noticeable difference from unaged contacts. Thus, while electrical performance of these contacts was essentially unchanged from that of the conve ntional ones, they do offer the benefit of dramatically reducing the amount of intermixing between layers even after being subjected to harsh long term aging. This is significant, as on e of the sources of failure for small gate width GaN devices is lateral flow arising due to th e mixing of the Al underlayer and Au overlayer, leading to the formation of the viscous AlAu4 phase. This leads to short circuiting and thus failure of devices. Chapter five dealt with the fabrication of Sc hottky contacts to p-GaN. For this purpose, W2B and W2B5 sputter targets were chosen. Nitrides we re ignored due to the severe intermixing expected between them and the GaN. It is also expected Ir would diffuse a great deal into the

PAGE 135

135 GaN at elevated annealing temperatures, also ma king it unsuitable for use as a Schottky contact. Contacts were fabricated with the scheme X / Plat inum / Gold, where X is the boride. Platinum was chosen as the adhesive layer instead of T itanium in order to eliminate formation of TiNx phases. Current-voltage measurements were used to evaluate the barrier he ight of the contacts. Because IV curves measured at different temperat ures were parallel, it was determined that the current transport mechanism present was thermioni c field emission. Unphysical barrier heights of approximately 4 eV were observed for anneals up to 700oC, with good stability. A slight decrease is seen at increased annealing temperat ures. This likely was the result of either an increased near surface defect concentration due to sputtering or due to incorporation of an interfacial oxygen layer between the GaN and the contact. X-ray Photoelectron Spectroscopy measurements of thin boride la yers on GaN reveal a true barrier height of 2.85 eV, in close agreement with that predicted by the Schottky-Mo tt model. Capacitance-voltage measurements confirm the large barrier height in agreement with the IV results, but if an interfacial layer of 1-2 monolayers is included the corr ected barrier height would be in agreement with XPS measurements. Chapter six deals with the application of bor ide and Ir-based Ohmic contacts to p-GaN for light emitting diodes. Contacts to LEDs were fabricated with the structures Nickel / Gold, Nickel / Gold / Titanium Boride / Titanium / Gold, and Nickel / Gold / Iridium / Gold. They were then annealed at 500oC, in order to prevent damage of the LED structure as well as metal spiking into the active layers. Initial IV a nd photoluminescence measurements revealed all contacts to have similar propert ies and performance. Contacts were then aged initially at 200oC, after which all displayed a slight degradation of performance. After further aging at 350oC, the

PAGE 136

136 Ni/Au contacts had degraded severely while the TiB2 and Ir-based contacts maintained good performance. This increased device stability confirms the earlier work. In conclusion, contacts to nand p-GaN were fabricated. Nitride diffusion barrier contacts to n-GaN show much less intermixing that that present with a Ni di ffusion barrier, although no improvement in the IV characterist ics either as a functi on of annealing or duri ng long term aging. Ohmic contacts to p-GaN using th e borides displayed both the abil ity to withstand a wide range of anneal temperatures, indica ting their increased thermal budget as compared to conventional Ni/Au contacts, as well as far superior long term aging characteristics due to the stability of the borides on GaN. This has implications for processing of devices using p-GaN which may be used at elevated temperatures. While the specific Ohmic contacts may not be completely suitable for immediate use in light emitting a pplications due to thei r poor reflectivity and opacity, they show great promise. Development of transparent contacts using very thin layers of these borides may be possible, as are improveme nts in the reflectivity by experimenting with different metallization schemes and low anneal temperatures. In addition, if MOSFETs are to be achieved using GaN, low resistance, stable Ohmi c contacts will be necessary and these may fit the requirements.

PAGE 137

137 LIST OF REFERENCES 1. S. A. Campbell, The Science and Engineer ing of Microelectronics Fabrication 2nd ed., Oxford University Press, Oxford, UK, 2001. 2. J. C. Bean, in High-Speed Semiconductor Device s, edited by S.M. Sze, John Wiley & Sons, New York, NY, 1990. 3. S. M. Sze, Physics of Semiconductor Devices, John Wiley & Sons, New York, NY, 1981. 4. D.C. Look, Mater. Sci. Eng. B80, 383 (2001). 5. 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). 6. 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). 7. 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). 8. V. Kumar, L. Zhou, D. Selvanathan, and I. Adesida, J. Appl. Phys. 92, 1712 (2002) 9. A. Motayed, R. Bathe, M.C. Wood, O.S. Diouf R. D. Vispute, and S. Noor Mohammad, J. Appl. Phys. 93, 1087 (2003) 10. J. Burm, K. Chu, W. A. Davis, W. J. Schaff, L.F. Eastman, and T. J. Eustis, Appl. Phys. Lett. 70, 464 (1997) 11. M. E. Lin, Z. Ma, F. Y. Huang, Z. F. Fan, L. H. Allen, and H. Morko, Appl. Phys. Lett. 64, 1003 (1994). 12. B. P. Luther, S. E. Mohney, T. N. Jackson, M. Asif Khan, Q. Chen, and J. W. Yang, Appl. Phys. Lett. 70, 57 (1997). 13. Z. Fan, S. N. Mohammad, W. Kim, Aktas, A. E. Botchkarev, and H. Morko, Appl. Phys. Lett. 68, 1672 (1996) 14. S. Ruvimov, Z. Liliental-Weber, J. Washburn, D. Qiao, S. S. Lau, and Paul K. Chu, Appl. Phys. Lett. 73, 2582 (1998). 15. A. N. Bright, P. J. Thomas, M. Weyland, D. M. Tricker, C. J. Humphreys, and R. Davies, J. Appl. Phys. 89, 3143 (2001). 16. N. A. Papanicolaou, M. V. Rao, J. Mittereder and W. T. Anderson, J. Vac. Sci. Technol. B 19, 261 (2001). 17. Q. Z. Liu and S. S. Lau, Solid-S tate Electronics, 42, 677-691 (1998).

PAGE 138

138 18. R. Li, S. J. Cai, L. Wong, Y. Chen, K. L. Wang, R. P. Smith, S. C. Martin, K. S. Boutros, J. M. Redwing, IEEE Electron De vice Letters, 20, 323 (1999). 19. S. T. Sheppard, K. Doverspike, W. L. Pribble, S. T. Allen, J. W. Palmour, L. T. Kehias, T. J. Jenkins, IEEE Electron Devi ce Letters, 20, 161 (1999). 20. D. Qiao, Z. F. Guan, J. Carlton, S. S. Lau, and G. J. Sullivan, Appl. Phys. Lett. 74, 2652 (1999) 21. V. Tilak, R. Dimitov, M. Murphy, B. Green, J. Smart, W. J. Schaft, J. R. Shealy and L. F. Eastman, Mat. Res. Soc. Symp. Proc., 622, T7.4.1(2000). 22. S. -H. Lim, J. Washburn, Z. Liliental-Weber, and D. Qiao, Appl. Phys. Lett. 78, 3797 (2001). 23. M. Asif Khan, M. S. Shur and Q. Chen, Appl. Phys. Lett., 68, 3022(1996). 24. S. Murai, H. Masuda, Y. Koide, and M. Murakami, Appl. Phys. Lett. 80, 2934 (2002) 25. C. Lu, H. Chen, X. Lv, X. Xie, and S. N. Mohammad, J. Appl. Phys. 91, 9218 (2002) 26. J. S. Foresi and T. D. Mousta kus, Appl. Phys. Lett., 62, 2859(1993). 27. M. W. Fay, G. Moldovan, P. D. Brown, I. Harri son, J. C. Birbeck, B. T. Hughes, M. J. Uren, and T. Martin, J. Appl. Phys. 92, 94 (2002). 28. K. O. Schweitz, P. K. Wang, S. E. Mohney, and D. Gotthold, Appl. Phys. Lett. 80, 1954 (2002). 29. M. W. Cole, D. W. Eckart, W. Y. Han, R. L. Pfeffer, T. Monahan, F. Ren, C. Yuan, R. A. Stall, S. J. Pearton, Y. Li and Y. Lu, J. Appl. Phys. 80, 278 (1996). 30. A. Zeitouny, M. Eizenberg, S. J. Pearton, and F. Ren, J. Appl. Phys. 88, 2048 (2000). 31. D. Selvanathan, F. M. Mohammed, A. Tesfayesus and I. Adesida, J.Vac.Sci.Technol. B22 2409(2004). 32. B. Luo, F. Ren, R. C. Fitch, J. K. Gillespie, T. Jenkins, J. Sewell, D. Via, A. Crespo, A.G. Baca R. D. Briggs, D. Gotthold, R. Birkhahn, B. Peres and S. J. Pearton, Appl. Phys. Lett. 82, 3910 (2003). 33. H. W. Jang and J. -L. Lee, J.Appl.Phys.93,5416(2003). 34. R. C. Fitch, J. K. Gillespie, N. Moser, T. Jenkins, J. Sewell, D. Via, A. Crespo, A.M. Dabiran, P. P. Chow, A. Osinsky, J. R. LaRoch e, F. Ren and S. J. Pearton, Appl.Phys.Lett. 84, 1495 (2004). 35. R. C. Fitch, J. K. Gillespie, N. Moser, T. Jenkins, J. Sewell, D. Via, A. Crespo, A.M. Dabiran, P. P. Chow, A. Osinsky, J. R. LaRoche, F. Ren and S. J. Pearton J.Vac.Sci.Technol. B22, 619(2004).

PAGE 139

139 36. D. Selvanathan L. Zhou, V. Kumar, I. Adesida and N. Finnegan, J.Electron.Mater.32 335(2003). 37. X. A. Cao, S. J. Pearton, G. Dang, A. P. Zh ang, F. Ren, and J. M. Van Hove, Appl. Phys. Lett. 75, 4130 (1999). 38. H. Morkoc, S. Strite, G. B. Gao, M. E. Lin, B. Svedlov, and M. Burns, J. Appl. Phys. 76 (2), 1363 (1994). 39. S. Nakamura and G. Fasol, The Blue Laser Diode ~Springe r, Heidelberg,1997. 40. S. M. Sze, Physics of Semiconductor De vices, John Wiley & Sons, New York (1981). 41. http://www.onr.navy.mil/sci_tech/informa tion/312_electronics/ncsr/properties.asp 42. D. W. Palmer, www.semiconductors.co.uk, 2004a. 43. O. Ambacher, J. Phys. D: Appl. Phys. 31, 2653 (1998). 44. H. Omiya, F. A. Ponce, H. Marui, S. Tanaka, and T. Mukai ,Appl. Phys. Lett. 85, 6143 (2004). 45. M. R. Park and W. A. Anderson, Mat. Res. Soc. Symp. Proc. 680E, E6.5.1 (2001). 46. C. H Lin, D. L. Hibbard, A. Au, H. P. Lee, Z. J. Dong, F. J. Szalkowski, J. Chen, and C. Chen, Mat. Res. Soc. Symp. 639, G4.8.1 (2001). 47. T. Gessmann, Y.-L. Li, E.L. Waldron, J.W. Graff, E.F. Schubert, J.K. Sheu, Appl. Phys. Lett. 80, 986 (2002). 48. J. K. Kim, J.-L. Lee, J. W. Lee, H. E. Shin, Y. J. Park, and T. Kim, Appl. Phys. Lett. 73, 2953 (1998). 49. V. R. Reddy, Mater. Chem.Phys. 93 286 (2005). 50. D. Selvanathan, L. Zhou, V. Kumar, I. Adesida and N. Finnegan, J.Electron. Mater. 32, 335 (2003). 51. G. -T. Chen, C. -C. Pan, C. -S.Fang, T. -C. Huang, J. -I.Chyi, M. -N. Chang, S. -B. Huang and J. -T.Hsu, Appl. Phys. Lett. 85, 2797 (2004). 52. K. A. Rickert, A. B. Ellis, J. K. Kim, J. L. Lee, F. J. Himpsel, F. Dwikusuma, and T. F. Kuech, J. Appl. Phys. 92, 6671 (2002). 53. Q. Z. Liu and S. S. Lau, Solid -State Electron. 42, 667 (1998). 54. M. W. Wang, J. O. McCaldin, J. F. Swenberg T. C. McGill, and R. J. Hauenstein, Appl. Phys. Lett. 66, 1974 (1995)

PAGE 140

140 55. T. Mori, T. Kozawa, T. Ohwaki, Y. Taga, S. N aagai, S. Yamasaki, S. Asami, N. Shibata, and M. Koike, Appl. Phys. Lett. 69, 3537 (1996). 56. L. S. Yu, D. Qiao, L. Jia, S. S. Lau, Y. Qi and K. M. Lau, Appl. Phys. Lett. 79, 4536 (2001). 57. L. S. Yu, L. Jia, D. Qiao, S.S. Lau, J. Li, J.Y. Lin, and H.X. Jiang, IEEE Trans. Electron. Devices 50, 292 (2003). 58. X. A. Cao, S. J. Pearton, G. Dang, A. P. Zh ang, F. Ren, and J. M. Van Hove, Appl. Phys. Lett. 75, 4130 (1999). 59. P. J. Hartlieb, A. Roskowski, R. F. Davis, W. Platow, and R. J. Nemanich, J. Appl. Phys. 91, 732 (2002). 60. D. L. Hibbard, R. W. Chuang, Y. S. Zhao, C. L. Jensen, H. P. Lee, Z. J. Dong, R. Shih, and M. Bremser, J. Electron. Mater. 29, 291 (2000). 61. K. Shiojima, T. Sugahara, and S. Sakai, Appl. Phys. Lett. 74, 1936 (1999). 62. Y. J. Lin, Appl. Phys. Lett. 86, 122109 (2005). 63. K. Shiojima, T. Sugahara, and S. Sakai, Appl. Phys. Lett. 77, 4353 (2000). 64. T. Sawada, Y. Ito, N. Kimura, K. Imai, K. Suzuki, and S. Sakai, Appl. Surf. Sci. 190, 326 (2002). 65. C. K. Tan, A. Abdul Aziz, and F. K. Yam, Appl. Surf. Sci. 252, 5930 (2006). 66. Y. J. Lin and C. W. Lin, J. Electron. Mat. 33, 1036 (2004). 67. T. B. Massalski (Ed.), Binary alloy phase diagrams (ASM International, Materials Park, Ohio, 1990). 68. X. A. Cao, S. J. Pearton, F. Ren, and J. R. Lothian, Appl. Phys. Lett 73, 942 (1998). 69. X. A. Cao, F. Ren, J. R. Lothian, S. J. Pearton, C. R. Abernathy, J. C. Zolper, M. W. Cole, A. Zeitouny, M. Eizenberg, and R. J. Shul, MRS Internet Journal of Nitride Semiconductor Research. 4 (1999). 70. Handbook of the Physiochemical Properties of the Elements, edited by G.V. Samsonov, IFIPlenum, New York, (1968). 71. M. W. Cole, F. Ren, and S. J. Pearton, J. Electrochem. Soc. 144, L275 (1997). 72. A. Zeitouny, M. Eizenberg, S. J. Pearton, and F. Ren, J. Appl. Phys. 88, 2048 (2000). 73. J. Kim, F. Ren, A. G. Baca, and S. J. Pearton, Appl. Phys. Lett. 82, 3263 (2003).

PAGE 141

141 74. B. Luo, F. Ren, R. C. Fitch, J. K. Gillespie, T. Jenkins, J. Sewell, D. Via, A. Crespo, A. G. Baca, R. D. Briggs, D. Gotthold, R. Birkhahn, B. Peres, and S. J. Pearton, Appl. Phys. Lett. 82, 3910 (2003). 75. R. Khanna, S. J. Pearton, F. Ren, and I. Kravchenko, J. Electrochem. Soc. 11, G804 (2005). 76. R. Khanna, S. J. Pearton, F. Ren, I. Kravchenko C.J. Kao and G.C. Chi, Appl. Phys. Lett. 87, 052110 (2005). 77. R. Khanna, S. J. Pearton, F. Ren, and I. Kravchenko, Appl. Surf. Sci. 252, 5814 (2006). 78. T. N. Oder, P. Martin, J. Y. Lin, H. X. Ji ang, J. R. Williams, and T. Isaacs-Smith, Appl. Phys. Lett. 88, 183505 (2006). 79. R. T. Tung, Mater. Sci. Rep. R35, 1 (2001). 80. L. Stafford, L. F. Voss, S. J. Pearton, J. -J. Chen, and F. Ren, submitted. 81. Michael Shur, Physics of Semiconductor Devi ces (Prentice-Hall, Engelwood Cliffs, NJ, 1990). 82. C. Merz, M. Kunzer, U. Kaufmann, I. Akasaki, and H. Amano, Semicond. Sci. Technol. 11, 712 (1996). 83. H. Nakayama, P. Hacke, M.R.H. Khan, T. Dtec hprohm, K. Hiramatsu, and N. Sawaki, Jpn. J. Appl. Phys. 35, L282 (1996). 84. J. R. Waldrop and R. W. Grant, Appl. Phys. Lett. 68, 2879 (1996). 85. J. S. Kwak, O. H. Nam, and Y. Pa rk, Appl. Phys. Lett. 80, 3554 (2002). 86. J. S. Kwak, O. H. Nam, and Y. Park, J. Appl. Phys. 95, 5917 (2004). 87. M.W. Cole, F. Ren, and S. J. Pearton, J. Electrochem. Soc. 144, 275 (1997). 88. K. Shiojima, S. Sugitani, and S. Sakai, Appl. Surf. Sci. 190, 318 (2002). 89. T. Okumura, Jpn. J. Appl. Phys. 24, L437 (1985). 90. M. Passlack, N. E. J. Hunt, E. F. Schubert, G. J. Zydzik, M. Hong, J. P. Mannaerts, R. L. Opila, and R. J. Fischer, Appl. Phys. Lett. 64, 2715 (1994). 91. M. Lenzlinger and E. H. Snow, J. Appl. Phys. 40, 278 (1969). 92. Z. Weinberg, J. Appl. Phys. 53, 5052 (1982). 93. H. Zhang, E. J. Miller, and E. T. Yu, J. Appl. Phys. 99, 023703 (2006).

PAGE 142

142 94. Introduction to Nitride Semiconductor Blue Lase rs and Light-Emitting Diodes, edited by S. Nakamura and S.F. Chichibu (Taylor and Francis, London, 2000). 95. M. Meneghini, L. -R. Trevisanello, U. Zehnder, T. Zahner, U. Strauss, G. Meneghesso and E. Zanoni, IEEE Trans. Electron. Dev. 53, 2981 (2006). 96. S. Bychikhin, D. Pogany, L. K. J. Vandamme, G. Meneghesso, and E. Zanoni, J. Appl. Phys. 97, 123714 (2005). 97. M. Pavesi, F. Rossi, and E. Zanoni, Semicond. Sci. Technol. 21, 138 (2006). 98. F. Rossi, M. Pavesi, M. Meneghini, G. Salviati M. Manfredi, G. Meneghesso, A. Castaldini, A. Cavallini, L. Rigutti, U. Strass, U. Zehnder, and E. Zanoni, J. Appl. Phys. 99, 053104 (2006). 99. X. A. Cao and S.D. Arthur, Appl. Phys. Lett. 85, 3971 (2004). 100. X. A. Cao, S. F. LeBoeuf, M. P. D'Evelyn, S. D. Arthur, J. Kretchmer, C. H. Yan, and Z. H. Yang, Appl. Phys. Lett. 84, 4313 (2004). 101. X. A. Cao, J. M. Teetsov, M. P. D'Evelyn, D. W. Merfeld, and C. H. Yan, Appl. Phys. Lett. 85, 7 (2004). 102. S. Nakamura, M. Senoh, S. Nagahama, N. Iwasa, T. Yamada, T. Matsushita, Y. Sugimoto and H. Kikyou, Jpn. J. Appl. Phys., Part 2 36, L1059 (1997). 103. C. Y. Hsu, W. H. Lan and Y. S. Wu, Appl. Phys. Lett. 83, 2447 (2003). 104. A. P. Zhang, B. Luo, J. W. Johnson, F. Ren, J. Han and S. J. Pearton, Appl. Phys. Lett. 79, 3636 (2001). 105. T. Maeda, Y. Koide and M. Murakam i, Appl. Phys. Lett. 75, 4145 (1999). 106. L. -C. Chen, J. -K. Ho, C. -S. Jong, C. C. Chiu, K. -K. Shih, F. -R. Chen, J. -J. Kai and L. Chang, Appl. Phys. Lett. 76, 3703 (2000) 107. D. Qiao, L. S. Yu, S. S. Lau, J. Y. Lin, H. X. Jiang and T. E. Haynes, J. Appl. Phys. 88, 4196 (2000). 108. H. S. Venugopalan, S. E. Mohney, J. M. DeLucca, B. P. Luther and G. E. Bul man, J. Vac. Sci. Technol. A 16, 607 (1998). 109. C. -F. Chu, C. C. Yu, Y. K. Wang, J. Y. Tsai, F. I. Lai and S. C. Wang, Appl. Phys. Lett. 77, 3423 (2000). 110. L. F. Voss, R. Khanna, S. J. Pearton, F. Ren and I. I. Kravchenko, Appl. Phys. Lett. 88, 012104 (2006).

PAGE 143

143 111. L. F. Voss, R. Khanna, S. J. Pearton, F. Ren, and I. I. Kravchenko, Appl. Surf. Sci. 253, 1255 (2006). 112. W. Zagozdzon-Wosik, D. Radhakrishnan, C. Darne, I. Rusakova, P. van der Heide, Z.-H. Zhang, J. Bennett, L. Trombetta and P. Majh i, Rev. Adv. Mater. Sci. 8, 185 (2004). 113. Y. D. Qi, H. Liang, D. Wang, Z. D. Lu, W. Tang and K. M. Lau, Appl. Phys. Lett. 86, 101903 (2005). 114. X. Guo and E. F. Schubert, Appl. Phys. Lett. 78, 3337 (2001). 115. A. Chitnis, J. Sun, V. Mandavilli, R. Pachi pulusu, S. Wu, M. Gaevski, V.Adivarahan, J. P. Zhang, M. Asif Khan, A. Sarua and M. Kuball, Appl. Phys. Lett. 81, 3491 (2002). 116. P. G. Eliseev, P. Perlin, J. Furioli, P. Sartor i, J. Mu, and M. Osinski, J. Electron. Mater. 26, 311 (1997). 117. G. Franssen, E. Litwin-Staszewska, R. Piotrz kowski, T. Suski, and P. Perlin, J. Appl. Phys. 94, 6122 (2003). 118. J. M. Shah, Y. L. Li, T. Gessmann and E. F. Schubert, J. Appl. Phys. 94, 2623 (2003). 119. S. H. Wang, S. E. Mohney, and R. Birkhahn, J. Appl. Phys. 91, 3711 (2002).

PAGE 144

144 BIOGRAPHICAL SKETCH Lars Voss was born in Pittsburgh, Pennsylvania in 1982. He spent m ost of his life in Erie, Pennsylvania, until graduating from McDowell Senior High School in 2000. He then enrolled at The Pennsylvania State University where he ea rned a Bachelor of Science in Engineering Science, with a minor in electronic and photonic ma terials, in May 2004. While there, he had the opportunity to work for Prof. Paul Koch in plas tics engineering technology at Penn State Erie during the summer and with Prof. P.M. Lenaha n in his Semiconductor Sp ectroscopy Laboratory. After his undergraduate education, Lars enrolled at th e University of Florida in the Materials Science and Engineering Departme nt and began his graduate study under Prof. Stephen J. Pearton. During the summers of 200 6 and 2007, he worked as an intern at Sandia National Laboratories under the direction of Drs. Albert G. Baca and Randy J. Shul.