1
00:00:00,050 --> 00:00:02,500
The following content is
provided under a Creative

2
00:00:02,500 --> 00:00:04,010
Commons license.

3
00:00:04,010 --> 00:00:06,350
Your support will help
MIT OpenCourseWare

4
00:00:06,350 --> 00:00:10,720
continue to offer high quality
educational resources for free.

5
00:00:10,720 --> 00:00:13,330
To make a donation or
view additional materials

6
00:00:13,330 --> 00:00:17,205
from hundreds of MIT courses,
visit MIT OpenCourseWare

7
00:00:17,205 --> 00:00:17,830
at ocw.mit.edu.

8
00:00:24,787 --> 00:00:27,120
TONIO BUONASSISI: Why don't
we go ahead and get started,

9
00:00:27,120 --> 00:00:28,980
folks.

10
00:00:28,980 --> 00:00:30,700
Just some small talk
to get us started.

11
00:00:30,700 --> 00:00:32,980
I promise to tell you the
stories about contamination

12
00:00:32,980 --> 00:00:34,800
and unintentional contamination.

13
00:00:34,800 --> 00:00:40,710
Before we dive in,
what a life, right?

14
00:00:40,710 --> 00:00:42,320
Steve Jobs.

15
00:00:42,320 --> 00:00:44,800
That was really something.

16
00:00:44,800 --> 00:00:47,210
Moments like that, I think
the best you can possibly do

17
00:00:47,210 --> 00:00:49,900
is to celebrate the person.

18
00:00:49,900 --> 00:00:53,570
And well, in honor
of his inventiveness

19
00:00:53,570 --> 00:00:56,440
and the way he really
turned Apple around,

20
00:00:56,440 --> 00:01:00,740
I just wanted to focus
a minute on that.

21
00:01:00,740 --> 00:01:03,490
Speaking of contamination and
unintentional contamination,

22
00:01:03,490 --> 00:01:08,200
the effect on processing,
the history books

23
00:01:08,200 --> 00:01:10,130
are full of this
folklore, if you

24
00:01:10,130 --> 00:01:12,100
start talking to people
who grow crystals

25
00:01:12,100 --> 00:01:15,310
and who manufacture solar cells.

26
00:01:15,310 --> 00:01:18,140
The growth of the crystals
that are used to actually make

27
00:01:18,140 --> 00:01:21,060
the wafers that ultimately
wind up being solar cells

28
00:01:21,060 --> 00:01:23,550
is a little bit of
a-- how would you

29
00:01:23,550 --> 00:01:26,940
say-- a little bit of an art.

30
00:01:26,940 --> 00:01:29,320
It is being codified
rather well now.

31
00:01:29,320 --> 00:01:31,630
And there's some strong
science behind it.

32
00:01:31,630 --> 00:01:36,419
But by and large,
up to very recently,

33
00:01:36,419 --> 00:01:37,960
it was more of an
art than a science.

34
00:01:41,000 --> 00:01:43,750
The best crystal growers are
people who are very observant

35
00:01:43,750 --> 00:01:48,180
it and who are able to look
out and see correlations

36
00:01:48,180 --> 00:01:51,470
where they weren't
immediately visible to others.

37
00:01:51,470 --> 00:01:54,850
So in one particular
factory, they

38
00:01:54,850 --> 00:01:57,600
started noticing that they had
yield losses during the winter

39
00:01:57,600 --> 00:01:58,380
time.

40
00:01:58,380 --> 00:02:00,690
And in particular they
had extreme yield losses

41
00:02:00,690 --> 00:02:02,910
whenever there was
a big snowstorm.

42
00:02:02,910 --> 00:02:06,830
And it was somebody, one of
the shift leads, I believe,

43
00:02:06,830 --> 00:02:09,240
who traced it back
to the salt that

44
00:02:09,240 --> 00:02:11,170
was marching in
on people's boots

45
00:02:11,170 --> 00:02:12,760
as they were coming
into the factory.

46
00:02:12,760 --> 00:02:16,180
And that sodium was
contaminating the silicon melt

47
00:02:16,180 --> 00:02:19,670
and resulting in the poor
quality of the crystals that

48
00:02:19,670 --> 00:02:20,880
were being grown.

49
00:02:20,880 --> 00:02:24,020
Flash forward several years
at Fraunhofer Institute

50
00:02:24,020 --> 00:02:26,680
for Solar Energy Systems
in Freiburg, Germany

51
00:02:26,680 --> 00:02:31,360
where the observant cell
manufacture started noticing,

52
00:02:31,360 --> 00:02:34,760
well, gee, our efficiencies
are always lower on Fridays.

53
00:02:34,760 --> 00:02:35,830
Why is that?

54
00:02:35,830 --> 00:02:39,470
And eventually, they traced
the problem down to the fact

55
00:02:39,470 --> 00:02:41,700
that, on Fridays,
the technicians

56
00:02:41,700 --> 00:02:43,630
would go out to lunch
at a Chinese restaurant.

57
00:02:43,630 --> 00:02:46,490
And they would come back, and
their breath and their hands

58
00:02:46,490 --> 00:02:48,330
would be MSG.

59
00:02:48,330 --> 00:02:50,960
And that was, again,
the sodium was downing

60
00:02:50,960 --> 00:02:52,670
performance of the devices.

61
00:02:52,670 --> 00:02:54,080
As so it's really interesting.

62
00:02:54,080 --> 00:02:57,640
This kind of sounds like
something you might see in a TV

63
00:02:57,640 --> 00:03:01,990
show where that scientist in
the lab coat waltzes into a room

64
00:03:01,990 --> 00:03:03,910
and says, it must be the sodium.

65
00:03:03,910 --> 00:03:04,410
Right?

66
00:03:04,410 --> 00:03:05,330
And then they test it.

67
00:03:05,330 --> 00:03:06,370
And oh, my gosh,
it's the sodium.

68
00:03:06,370 --> 00:03:07,885
Now, problem solving
the real world

69
00:03:07,885 --> 00:03:09,135
doesn't quite happen that way.

70
00:03:09,135 --> 00:03:11,650
It involves a very
methodical approach

71
00:03:11,650 --> 00:03:14,670
to testing a variety
of hypotheses,

72
00:03:14,670 --> 00:03:16,690
first brainstorming in
a structured manner,

73
00:03:16,690 --> 00:03:18,430
identifying the most
likely candidates,

74
00:03:18,430 --> 00:03:21,010
and going about
solving the problem,

75
00:03:21,010 --> 00:03:22,940
performing a series
of design experiments

76
00:03:22,940 --> 00:03:26,140
to really get to the root cause.

77
00:03:26,140 --> 00:03:28,340
So it's an interesting story.

78
00:03:28,340 --> 00:03:29,770
It's an interesting aside.

79
00:03:29,770 --> 00:03:32,620
Solving contamination problems,
though, is very difficult.

80
00:03:32,620 --> 00:03:34,340
And the best thing
you can possibly

81
00:03:34,340 --> 00:03:36,460
do if you're growing
crystals is to keep

82
00:03:36,460 --> 00:03:39,770
your system more clean than
you think you need it to be.

83
00:03:39,770 --> 00:03:41,830
That's the best
advice I can give.

84
00:03:41,830 --> 00:03:44,390
I was just talking with some
folks in Caltech the other day.

85
00:03:44,390 --> 00:03:46,260
They were running into
contamination issues.

86
00:03:46,260 --> 00:03:50,320
And that advice goes a long way.

87
00:03:50,320 --> 00:03:50,820
OK.

88
00:03:50,820 --> 00:03:53,620
Well, let's go ahead and get
started in the topic today.

89
00:03:53,620 --> 00:03:59,030
Back from Phoenix and eyes are
red from irritation of having

90
00:03:59,030 --> 00:04:00,630
been awake for too many hours.

91
00:04:00,630 --> 00:04:02,240
So that could only
mean one thing.

92
00:04:02,240 --> 00:04:04,047
Contacts, right?

93
00:04:04,047 --> 00:04:06,130
So we're going to be talking
about contacts today,

94
00:04:06,130 --> 00:04:09,560
charge extraction, as well.

95
00:04:09,560 --> 00:04:12,940
And to situate us
on the roadmap,

96
00:04:12,940 --> 00:04:15,040
we have our
fundamentals right here.

97
00:04:15,040 --> 00:04:17,600
We're about to jump forward
into the technology, which

98
00:04:17,600 --> 00:04:18,839
is really exciting.

99
00:04:18,839 --> 00:04:20,709
We're about out of
the woods here, folks.

100
00:04:20,709 --> 00:04:22,610
And finally, into the
cross-cutting themes.

101
00:04:22,610 --> 00:04:25,370
And so, again, every
photovoltaic device

102
00:04:25,370 --> 00:04:28,346
must obey this general
equation right here

103
00:04:28,346 --> 00:04:30,220
where the output energy
over the input energy

104
00:04:30,220 --> 00:04:31,910
equals a conversion efficiency.

105
00:04:31,910 --> 00:04:34,770
And for most solar cells, this
is represented by this right

106
00:04:34,770 --> 00:04:35,270
here.

107
00:04:35,270 --> 00:04:39,550
And we have now tackled every
single one of those, at least

108
00:04:39,550 --> 00:04:42,290
in a very fundamental way.

109
00:04:42,290 --> 00:04:45,510
And now, finally, we are focused
on the output, the charge

110
00:04:45,510 --> 00:04:49,510
collection, in other words,
the contacting of the device.

111
00:04:49,510 --> 00:04:51,257
Again, the total
cell efficiency is

112
00:04:51,257 --> 00:04:53,590
going to be the product of
each of these individual cell

113
00:04:53,590 --> 00:04:55,470
efficiencies here.

114
00:04:55,470 --> 00:04:59,150
And contacts are a
very, very easy way

115
00:04:59,150 --> 00:05:02,180
to kill your device
in a variety of ways.

116
00:05:02,180 --> 00:05:04,900
And we'll get to some of
these points right here.

117
00:05:04,900 --> 00:05:06,970
As a matter of
fact, this s back,

118
00:05:06,970 --> 00:05:08,820
this surface
recombination velocity

119
00:05:08,820 --> 00:05:10,390
on the backside
of the device, is

120
00:05:10,390 --> 00:05:12,347
a contact-related phenomenon.

121
00:05:12,347 --> 00:05:14,430
And that's exactly where
the water is spilling out

122
00:05:14,430 --> 00:05:17,540
of this bucket right here.

123
00:05:17,540 --> 00:05:19,490
Learning objectives.

124
00:05:19,490 --> 00:05:21,920
The idea is to start
out with describing

125
00:05:21,920 --> 00:05:25,750
the purpose of contacts,
and their most common types,

126
00:05:25,750 --> 00:05:28,526
then to describe the impact
of good and poor contacts

127
00:05:28,526 --> 00:05:30,150
on I-V characteristics,
in other words,

128
00:05:30,150 --> 00:05:32,457
to describe the device
impact of contacts.

129
00:05:32,457 --> 00:05:34,040
So we're convinced
that this is really

130
00:05:34,040 --> 00:05:36,582
something we should be spending
a lot of time thinking about.

131
00:05:36,582 --> 00:05:39,164
Then we're going to sketch the
I-V characteristics of Schottky

132
00:05:39,164 --> 00:05:40,100
and Ohmic contacts.

133
00:05:40,100 --> 00:05:42,640
We'll learn what those mean.

134
00:05:42,640 --> 00:05:44,180
I hope you know
what ohmic means.

135
00:05:44,180 --> 00:05:47,410
Describe what fundamental
material parameters determine

136
00:05:47,410 --> 00:05:50,430
I-V characteristics of a
contact/semiconductor junction,

137
00:05:50,430 --> 00:05:53,390
sketch common band alignments,
and sketch common solar cell

138
00:05:53,390 --> 00:05:55,067
device architectures,
once we're done.

139
00:05:55,067 --> 00:05:57,400
And then we can look back and
gaze over the fundamentals

140
00:05:57,400 --> 00:06:00,620
and say, Ha, ha,
how far we've come.

141
00:06:00,620 --> 00:06:02,880
And now that you look
at a solar cell device,

142
00:06:02,880 --> 00:06:05,240
hopefully, you'll look at it
in a much more profound way

143
00:06:05,240 --> 00:06:07,840
than you did in the
first day of class.

144
00:06:07,840 --> 00:06:10,110
The more you know,
the more you see.

145
00:06:10,110 --> 00:06:11,840
So contacts, why
do we need them?

146
00:06:11,840 --> 00:06:13,909
We need to extract the
carriers from the device.

147
00:06:13,909 --> 00:06:15,950
We need to prevent the
back-diffusion of carriers

148
00:06:15,950 --> 00:06:18,817
into the device.

149
00:06:18,817 --> 00:06:21,400
These contacts, in general, have
been studied very extensively

150
00:06:21,400 --> 00:06:22,810
in the semiconductor industry.

151
00:06:22,810 --> 00:06:25,072
Why is that?

152
00:06:25,072 --> 00:06:27,280
Why do we need contacts in
the semiconductor industry

153
00:06:27,280 --> 00:06:30,850
to make integrated circuits?

154
00:06:30,850 --> 00:06:31,990
Obvious answer.

155
00:06:31,990 --> 00:06:32,822
We need wires.

156
00:06:32,822 --> 00:06:35,280
We need to extract the charges
from our transistors, right?

157
00:06:35,280 --> 00:06:38,080
So we're injecting charge,
pulling out charge.

158
00:06:38,080 --> 00:06:40,380
So contacts have been
pretty extensively studied

159
00:06:40,380 --> 00:06:41,630
in the semiconductor industry.

160
00:06:41,630 --> 00:06:43,620
And we're going to
leverage that a lot.

161
00:06:43,620 --> 00:06:46,740
We're not going to reinvent the
wheel where we don't need to.

162
00:06:46,740 --> 00:06:49,290
Contacts are
semiconductor-specific.

163
00:06:49,290 --> 00:06:53,380
Fundamentals apply
broadly, but the specifics

164
00:06:53,380 --> 00:06:56,600
pertain to-- or the devil's in
the details, in other words.

165
00:06:56,600 --> 00:06:59,170
So there are very
specific effects

166
00:06:59,170 --> 00:07:03,420
that occur, depending on the
precise semiconductor metal

167
00:07:03,420 --> 00:07:04,890
combination.

168
00:07:04,890 --> 00:07:09,050
And lastly, the
contacts are heavily

169
00:07:09,050 --> 00:07:11,590
influenced by the interface
between the semiconductor

170
00:07:11,590 --> 00:07:12,780
and the metal.

171
00:07:12,780 --> 00:07:15,490
We're going to be talking
about that as well.

172
00:07:15,490 --> 00:07:18,450
So typical materials
used for contacts

173
00:07:18,450 --> 00:07:22,620
include metals, transparent
conducting oxides.

174
00:07:22,620 --> 00:07:26,040
And also, we can find heavily
doped organic materials

175
00:07:26,040 --> 00:07:28,390
as well.

176
00:07:28,390 --> 00:07:29,360
Metals, we understand.

177
00:07:29,360 --> 00:07:32,440
OK, they're optically opaque
and electrically conductive.

178
00:07:32,440 --> 00:07:34,770
That means they conduct
electricity very nicely.

179
00:07:34,770 --> 00:07:36,900
And so our series resistance
along the metal wire

180
00:07:36,900 --> 00:07:37,775
should be low.

181
00:07:37,775 --> 00:07:39,650
But they're optically
opaque, meaning they're

182
00:07:39,650 --> 00:07:41,150
resulting in a shading loss.

183
00:07:41,150 --> 00:07:43,004
So if we cover entire
front side with metal,

184
00:07:43,004 --> 00:07:45,170
we're not going to have a
very efficient solar cell,

185
00:07:45,170 --> 00:07:48,330
because our absorption of
light is going to be very poor.

186
00:07:48,330 --> 00:07:50,960
On the other hand,
transparent conducting oxides

187
00:07:50,960 --> 00:07:53,800
are optically transparent and
electrically conductive, not

188
00:07:53,800 --> 00:07:56,910
quite as conductive as
metals, but pretty close.

189
00:07:56,910 --> 00:07:59,990
So let me think about
that for a minute.

190
00:07:59,990 --> 00:08:03,570
A material that is
optically transparent--

191
00:08:03,570 --> 00:08:06,410
that means it must have
a very large band gap

192
00:08:06,410 --> 00:08:08,600
to let the light through,
does interact with it,

193
00:08:08,600 --> 00:08:10,170
but is electrically conductive.

194
00:08:10,170 --> 00:08:13,440
In other words, it has a high
concentration of free carriers

195
00:08:13,440 --> 00:08:16,030
that can move around the
material and conduct charge.

196
00:08:16,030 --> 00:08:18,562
How is that possible?

197
00:08:18,562 --> 00:08:20,020
I hope you're asking
that question,

198
00:08:20,020 --> 00:08:22,810
because this perplexed me
for a long time as well.

199
00:08:22,810 --> 00:08:25,250
And we'll answer that
in a couple of slides.

200
00:08:25,250 --> 00:08:27,970
First, the properties of TCOs.

201
00:08:27,970 --> 00:08:31,080
What makes a transparent
conducting oxide a transparent

202
00:08:31,080 --> 00:08:31,980
conducting oxide?

203
00:08:31,980 --> 00:08:34,355
I figure I'll spend a couple
slides on this, since metals

204
00:08:34,355 --> 00:08:35,570
are pretty self-evident.

205
00:08:35,570 --> 00:08:37,240
But TCOs, this might
be the first time

206
00:08:37,240 --> 00:08:38,409
you're encountering them.

207
00:08:38,409 --> 00:08:41,020
They're present in a
variety of devices.

208
00:08:41,020 --> 00:08:44,230
If you have an iPhone,
or if you've ever

209
00:08:44,230 --> 00:08:49,140
seen certain types of military
aircraft, TCOs are involved.

210
00:08:49,140 --> 00:08:51,920
So the material is
very transparent.

211
00:08:51,920 --> 00:08:55,300
You can see the transparency
over quite a broad wavelength

212
00:08:55,300 --> 00:09:01,160
range here, cutting off only
in the several eV range here.

213
00:09:01,160 --> 00:09:03,150
Transparency begins to
drop, because you're

214
00:09:03,150 --> 00:09:05,510
able to excite carries
across the band gap.

215
00:09:05,510 --> 00:09:07,280
So it's a very large
band gap material,

216
00:09:07,280 --> 00:09:10,100
and yet, the conductivity
is very good.

217
00:09:10,100 --> 00:09:13,980
So here, we can see that the
conductivity of ITO, that's

218
00:09:13,980 --> 00:09:17,420
indium tin oxide, that's a
particular type of transparent

219
00:09:17,420 --> 00:09:19,720
conducting oxide, is
almost as good as silver.

220
00:09:19,720 --> 00:09:21,900
Not quite, but it's
approaching it.

221
00:09:21,900 --> 00:09:26,770
And so again, the conductivity,
which is 1 over the resistivity

222
00:09:26,770 --> 00:09:29,400
is related to the carrier
concentration and the mobility,

223
00:09:29,400 --> 00:09:30,880
as we've seen before.

224
00:09:30,880 --> 00:09:34,450
And therefore, there must
be a lot of free carriers.

225
00:09:34,450 --> 00:09:38,210
And they must be fairly mobile,
to move around the material.

226
00:09:38,210 --> 00:09:40,130
So how does that work?

227
00:09:40,130 --> 00:09:43,350
Band gap greater than 3.1
eV, transmittance very high,

228
00:09:43,350 --> 00:09:46,760
as a result, but still a
large number of carriers.

229
00:09:46,760 --> 00:09:49,930
Well, if we consider our
band gap right here--

230
00:09:49,930 --> 00:09:52,270
and we'll consider floating
Fermi energy right now,

231
00:09:52,270 --> 00:09:54,600
we'll explain why that's
pegged there in a minute--

232
00:09:54,600 --> 00:09:58,099
we assume that the
band gap is very large,

233
00:09:58,099 --> 00:10:00,640
so that light can go through,
and it's optically transparent.

234
00:10:00,640 --> 00:10:04,280
But now, by doping the material
with a specific type of dopant

235
00:10:04,280 --> 00:10:07,150
that forms mid-gap
states, we're going

236
00:10:07,150 --> 00:10:12,350
to create a new energy level
here that is partially filled.

237
00:10:12,350 --> 00:10:16,090
We can then excite, with
relatively little energy,

238
00:10:16,090 --> 00:10:19,850
carriers into the excited
states within that orange band,

239
00:10:19,850 --> 00:10:22,705
and they can transport charge
throughout the material.

240
00:10:22,705 --> 00:10:24,830
On the other hand, if we
come in with a high energy

241
00:10:24,830 --> 00:10:27,580
photon or a moderate
energy photo,

242
00:10:27,580 --> 00:10:29,240
let's say, in the
visible, it doesn't

243
00:10:29,240 --> 00:10:32,520
have quite enough energy to
excite from that orange band

244
00:10:32,520 --> 00:10:35,060
into the conduction band,
nor from the valence band

245
00:10:35,060 --> 00:10:37,760
into the orange band, the
intermediate band there.

246
00:10:37,760 --> 00:10:40,960
And that's the general principle
of how transparent conducting

247
00:10:40,960 --> 00:10:43,210
oxides function.

248
00:10:43,210 --> 00:10:46,930
You can have indium doped
tin oxide, for instance, ITO,

249
00:10:46,930 --> 00:10:48,920
as a classic example.

250
00:10:48,920 --> 00:10:51,440
And this is the basic premise.

251
00:10:51,440 --> 00:10:54,570
Any questions so far, because
this is kind of important,

252
00:10:54,570 --> 00:10:56,920
before we launch into-- yeah.

253
00:10:56,920 --> 00:10:59,552
AUDIENCE: So how big
are those, that E1, E2?

254
00:10:59,552 --> 00:11:00,510
TONIO BUONASSISI: Yeah.

255
00:11:00,510 --> 00:11:07,630
So this E3 right here is around
0.4 eV, for a typical TCO.

256
00:11:07,630 --> 00:11:10,660
And E1 and E2, the
sum of both of them

257
00:11:10,660 --> 00:11:13,360
could be on the order
of three to four eV.

258
00:11:13,360 --> 00:11:15,020
Maybe higher.

259
00:11:15,020 --> 00:11:15,770
Well actually, no.

260
00:11:15,770 --> 00:11:16,842
Sorry.

261
00:11:16,842 --> 00:11:18,300
Just one of these
transitions right

262
00:11:18,300 --> 00:11:21,740
here could be on the
order of three eV.

263
00:11:21,740 --> 00:11:24,160
So the entire band gap
could be quite large.

264
00:11:24,160 --> 00:11:26,770
If you think of metal
oxides-- most metal oxides

265
00:11:26,770 --> 00:11:28,660
have very large band gaps.

266
00:11:28,660 --> 00:11:32,610
And most TCOs fall
into that category.

267
00:11:32,610 --> 00:11:35,120
They could be tin
oxide or zinc oxide,

268
00:11:35,120 --> 00:11:37,960
but heavily doped
with an element,

269
00:11:37,960 --> 00:11:40,620
like indium, or
fluorine, or aluminum,

270
00:11:40,620 --> 00:11:43,010
to provide that
intermediate band.

271
00:11:43,010 --> 00:11:45,570
AUDIENCE: What are the doping
concentrations required

272
00:11:45,570 --> 00:11:46,504
[INAUDIBLE]?

273
00:11:46,504 --> 00:11:47,420
TONIO BUONASSISI: Yes.

274
00:11:47,420 --> 00:11:51,530
So this is a deep dive,
and I'm happy to entertain

275
00:11:51,530 --> 00:11:53,185
that question.

276
00:11:53,185 --> 00:11:55,060
For those who it kind
of goes over your head,

277
00:11:55,060 --> 00:11:55,560
don't worry about it.

278
00:11:55,560 --> 00:11:57,530
We'll return, so plant
a flag post in your mind

279
00:11:57,530 --> 00:11:58,850
where we're at.

280
00:11:58,850 --> 00:12:01,734
Remember the hydrogenic donor
model for a dopant atom.

281
00:12:01,734 --> 00:12:03,150
You introduce it
into the lattice.

282
00:12:03,150 --> 00:12:05,430
And that loosely bound
electron is there, kind of

283
00:12:05,430 --> 00:12:06,710
like a hydrogenic donor.

284
00:12:06,710 --> 00:12:08,310
It has a certain radius.

285
00:12:08,310 --> 00:12:10,710
And if you begin doping at
a very high concentration,

286
00:12:10,710 --> 00:12:13,001
those electrons will begin
interfering with each other.

287
00:12:13,001 --> 00:12:14,410
They'll begin interacting.

288
00:12:14,410 --> 00:12:17,300
And so instead of forming
one isolated defect level,

289
00:12:17,300 --> 00:12:19,530
they'll start splitting
and forming a band.

290
00:12:19,530 --> 00:12:21,760
And the density at
which you have to dope

291
00:12:21,760 --> 00:12:24,854
depends on that donor radius.

292
00:12:24,854 --> 00:12:27,270
So you're typically talking
about concentrations in the 10

293
00:12:27,270 --> 00:12:31,380
to the 20 per cubic centimeter,
or one atomic percent

294
00:12:31,380 --> 00:12:32,502
or higher.

295
00:12:32,502 --> 00:12:33,960
So that's the dopant
density that's

296
00:12:33,960 --> 00:12:36,450
required to force
this transition.

297
00:12:36,450 --> 00:12:38,190
It's pretty cool.

298
00:12:38,190 --> 00:12:40,880
I mean, there are people who
do their PhDs on transparent

299
00:12:40,880 --> 00:12:41,630
conducting oxides.

300
00:12:41,630 --> 00:12:42,849
Maybe some of you are.

301
00:12:42,849 --> 00:12:45,140
You could probably come up
and give a lecture about it.

302
00:12:45,140 --> 00:12:47,800
So it's a fascinating subject.

303
00:12:47,800 --> 00:12:51,230
And because ITO, or
indium tin oxide,

304
00:12:51,230 --> 00:12:53,860
is just so amazing in
terms of its conductivity--

305
00:12:53,860 --> 00:12:55,760
it's almost up there
buttressing again

306
00:12:55,760 --> 00:13:01,210
silver-- it's very difficult
to replace with something else.

307
00:13:01,210 --> 00:13:04,100
Zinc oxide and its
variants get pretty close,

308
00:13:04,100 --> 00:13:06,290
like aluminum dope zinc
oxide, sometimes called

309
00:13:06,290 --> 00:13:07,927
AZO in the PV community.

310
00:13:07,927 --> 00:13:09,510
They get pretty
close, but they're not

311
00:13:09,510 --> 00:13:12,580
quite as good in terms
of optical transparency

312
00:13:12,580 --> 00:13:14,260
and conductivity as ITO.

313
00:13:14,260 --> 00:13:15,400
Now ITO is a problem.

314
00:13:15,400 --> 00:13:16,610
It contains indium.

315
00:13:16,610 --> 00:13:18,900
And the world supply
of indium is limited.

316
00:13:18,900 --> 00:13:21,810
It's not the most abundant
element on the Earth's crust.

317
00:13:21,810 --> 00:13:24,410
If you look at a periodic table,
it's pretty far down there.

318
00:13:24,410 --> 00:13:26,610
And as you know, the
elemental abundance,

319
00:13:26,610 --> 00:13:29,590
it has a relative decay
from-- a power law

320
00:13:29,590 --> 00:13:31,570
decay-- from light
elements to heavy elements

321
00:13:31,570 --> 00:13:33,780
with some fluctuations
in between.

322
00:13:33,780 --> 00:13:36,360
But in general, the star
dust that we have here

323
00:13:36,360 --> 00:13:38,440
is less abundant in
the heavier elements.

324
00:13:38,440 --> 00:13:40,620
And indium is one
of those elements.

325
00:13:40,620 --> 00:13:44,210
And so people are searching
for alternatives to ITO.

326
00:13:44,210 --> 00:13:47,311
And it's an active area
of research right now.

327
00:13:47,311 --> 00:13:47,810
OK.

328
00:13:47,810 --> 00:13:50,970
So for now, let's go
back to the flag post

329
00:13:50,970 --> 00:13:53,520
that we planted just
a few minutes ago.

330
00:13:53,520 --> 00:13:55,050
For everyone who
didn't quite follow

331
00:13:55,050 --> 00:13:57,591
the detailed explanation, the
important thing to keep in mind

332
00:13:57,591 --> 00:14:00,440
is that TCOs are
conductive and transparent.

333
00:14:00,440 --> 00:14:04,650
And they are so, because of
the unique band structure.

334
00:14:04,650 --> 00:14:07,410
So we can create contacts
to semiconductors

335
00:14:07,410 --> 00:14:10,520
and extract charge using
either a metal or a TCO.

336
00:14:10,520 --> 00:14:12,890
And right now,
we're going to dive

337
00:14:12,890 --> 00:14:15,620
into the impact of good
and bad contacts on device

338
00:14:15,620 --> 00:14:16,830
characteristics.

339
00:14:16,830 --> 00:14:20,064
So during our last-- I believe
it was two lectures ago,

340
00:14:20,064 --> 00:14:21,980
we talked about the
equivalent circuit diagram

341
00:14:21,980 --> 00:14:25,500
of a solar cell, this being the
simplest case, now corrected

342
00:14:25,500 --> 00:14:27,640
with our minus 1
term right here.

343
00:14:27,640 --> 00:14:32,380
And we can see our I-V curve in
linear, linear scale and log,

344
00:14:32,380 --> 00:14:35,220
linear scale there, log
current linear voltage.

345
00:14:35,220 --> 00:14:36,906
And because this
is an exponential,

346
00:14:36,906 --> 00:14:38,280
you would expect
a straight line,

347
00:14:38,280 --> 00:14:40,230
and you do see that indeed.

348
00:14:40,230 --> 00:14:43,365
Now the I-V curve,
in the upper part,

349
00:14:43,365 --> 00:14:44,990
you can really see
has a very nice fill

350
00:14:44,990 --> 00:14:47,260
factor as a very sharp kink.

351
00:14:47,260 --> 00:14:48,240
Pops straight up.

352
00:14:48,240 --> 00:14:51,330
And so the fill factor of that
I-V curve, when illuminated,

353
00:14:51,330 --> 00:14:52,760
will be very high.

354
00:14:52,760 --> 00:14:56,250
And hence, the solar cell
efficiency will be high.

355
00:14:56,250 --> 00:14:58,880
Not so when you introduce
a series resistance.

356
00:14:58,880 --> 00:15:01,540
So when you have that
series resistance component,

357
00:15:01,540 --> 00:15:04,100
now, at higher
bias voltages, you

358
00:15:04,100 --> 00:15:06,250
begin to be series
resistance limited,

359
00:15:06,250 --> 00:15:08,337
which causes the fill
factor to decrease.

360
00:15:08,337 --> 00:15:10,170
And you can see that
by going back and forth

361
00:15:10,170 --> 00:15:12,030
between these two slides.

362
00:15:12,030 --> 00:15:14,570
High fill factor, no
series resistance.

363
00:15:14,570 --> 00:15:16,907
Lower fill factor,
higher series resistance.

364
00:15:16,907 --> 00:15:19,490
Especially, if you focus right
here at this point right there,

365
00:15:19,490 --> 00:15:23,130
you can really see the
decrease of fill factor

366
00:15:23,130 --> 00:15:25,710
as a result of adding a
series resistance component.

367
00:15:25,710 --> 00:15:29,332
So contacts, when
improperly performed,

368
00:15:29,332 --> 00:15:31,790
can add a series resistance
component, a rather significant

369
00:15:31,790 --> 00:15:34,790
one, and drop the
performance of your device.

370
00:15:34,790 --> 00:15:36,140
Likewise, shunt resistance.

371
00:15:36,140 --> 00:15:39,730
We saw a simple effect right
here of shunt resistance.

372
00:15:39,730 --> 00:15:43,050
If the shunt resistance
goes down further

373
00:15:43,050 --> 00:15:46,990
and the saturation current
goes up even higher,

374
00:15:46,990 --> 00:15:49,140
it begins to affect the
fill factor as well.

375
00:15:49,140 --> 00:15:53,480
And so shunting can also
be impacted by contacts.

376
00:15:53,480 --> 00:15:56,760
Particularly when you fire your
contacts, in other words when

377
00:15:56,760 --> 00:16:00,060
you heat your sample up,
or you heat your stack up

378
00:16:00,060 --> 00:16:04,270
to create good contact between
the metal and the semiconductor

379
00:16:04,270 --> 00:16:07,310
underneath, or the TCO and
the semiconductor underneath,

380
00:16:07,310 --> 00:16:09,330
a couple of things can happen.

381
00:16:09,330 --> 00:16:12,430
If you don't get the chemistry
just right at the interface--

382
00:16:12,430 --> 00:16:14,490
maybe didn't clean
your sample properly,

383
00:16:14,490 --> 00:16:17,570
maybe you didn't heat
it up high enough,

384
00:16:17,570 --> 00:16:21,960
so that the atoms really started
to interdiffuse and interact--

385
00:16:21,960 --> 00:16:24,880
you can under-fire your contact.

386
00:16:24,880 --> 00:16:27,820
And that results in poor
contact with your semiconductor

387
00:16:27,820 --> 00:16:30,010
and large series resistance.

388
00:16:30,010 --> 00:16:32,560
So large series resistance
will, essentially,

389
00:16:32,560 --> 00:16:38,482
impede charge flow and
result in a fill factor loss.

390
00:16:38,482 --> 00:16:39,940
On the other hand,
if you over-fire

391
00:16:39,940 --> 00:16:42,440
your contact-- let's
say you heat it up,

392
00:16:42,440 --> 00:16:46,280
but fire it a little too long--
then the contact material

393
00:16:46,280 --> 00:16:48,260
can drive too far into
your semiconductor,

394
00:16:48,260 --> 00:16:50,160
and you wind up
shunting your device.

395
00:16:50,160 --> 00:16:51,390
What does a shunt mean?

396
00:16:51,390 --> 00:16:54,360
A shunt mean that the metal
goes straight through the PN

397
00:16:54,360 --> 00:16:58,710
junction and contacts the base
material on the other side.

398
00:16:58,710 --> 00:17:00,955
So remember, the PN junction
region of the solar cell

399
00:17:00,955 --> 00:17:03,530
is only about a micron
away from the surface.

400
00:17:03,530 --> 00:17:04,930
It's really close.

401
00:17:04,930 --> 00:17:07,641
So if you have a fast diffusing
metal species and you heat it

402
00:17:07,641 --> 00:17:09,890
up too much, you get the
metal going straight through,

403
00:17:09,890 --> 00:17:12,880
in this case, the
n+ layer into the P,

404
00:17:12,880 --> 00:17:14,650
you'll shunt your device.

405
00:17:14,650 --> 00:17:16,150
So this is the fine
balance that you

406
00:17:16,150 --> 00:17:18,349
have to walk when
you're putting contact

407
00:17:18,349 --> 00:17:19,810
metallization on the device.

408
00:17:19,810 --> 00:17:22,630
You can't over-fire it,
because we wind up with shunts.

409
00:17:22,630 --> 00:17:24,640
You can't under-fire
it either, because you

410
00:17:24,640 --> 00:17:27,640
might wind up with a very
large contact resistance.

411
00:17:27,640 --> 00:17:31,020
So it's a very tricky
thing to get right.

412
00:17:31,020 --> 00:17:34,780
We just finished installing a
contact metallization printer

413
00:17:34,780 --> 00:17:35,350
in our lab.

414
00:17:35,350 --> 00:17:37,560
And Joe, how many
human hours total

415
00:17:37,560 --> 00:17:41,036
have been spent trying to
optimize the firing process?

416
00:17:41,036 --> 00:17:43,270
JOE: Probably
about 30 or 40 now?

417
00:17:43,270 --> 00:17:44,990
TONIO BUONASSISI: 30
or 40 human hours?

418
00:17:44,990 --> 00:17:46,602
So it's still a
work in progress,

419
00:17:46,602 --> 00:17:48,310
but it takes a while
to really nail that.

420
00:17:48,310 --> 00:17:49,530
Yes, Ashley?

421
00:17:49,530 --> 00:17:53,190
ASHLEY: So would
you want your TCO

422
00:17:53,190 --> 00:17:58,003
to melt at a lower temperature
that your semiconductor?

423
00:17:58,003 --> 00:18:01,190
Is that another consideration?

424
00:18:01,190 --> 00:18:03,500
TONIO BUONASSISI: So
let's back up a step.

425
00:18:03,500 --> 00:18:06,010
For the metal case, it's
pretty straightforward.

426
00:18:06,010 --> 00:18:07,610
This is a homogeneous element.

427
00:18:07,610 --> 00:18:10,750
Let's call it a unary material,
meaning one element comprising

428
00:18:10,750 --> 00:18:12,900
that metal, typically.

429
00:18:12,900 --> 00:18:16,270
And it's very simple for us
to see, OK, if we keep it up

430
00:18:16,270 --> 00:18:19,610
and it reacts with our
material, it drives in.

431
00:18:19,610 --> 00:18:21,590
For a TCO, it's a little
less straightforward,

432
00:18:21,590 --> 00:18:24,420
because you typically
have a multinary compound,

433
00:18:24,420 --> 00:18:27,350
meaning several elements
comprising your TCO.

434
00:18:27,350 --> 00:18:29,060
The formation of this
intermediate band

435
00:18:29,060 --> 00:18:34,300
right here that forms, really,
the conductivity of the TCO,

436
00:18:34,300 --> 00:18:36,920
is predicated upon
the principle that you

437
00:18:36,920 --> 00:18:39,070
have these dopant atoms
homogeneously spaced

438
00:18:39,070 --> 00:18:40,350
throughout your material.

439
00:18:40,350 --> 00:18:41,460
If you heat it up
too high and they

440
00:18:41,460 --> 00:18:42,834
begin to cluster,
in other words,

441
00:18:42,834 --> 00:18:44,910
precipitate out of
phase, you can wind up

442
00:18:44,910 --> 00:18:47,970
destroying your TCO, or
increasing the resistivity,

443
00:18:47,970 --> 00:18:51,690
maybe even decreasing
the optical transparency.

444
00:18:51,690 --> 00:18:54,440
And so it really is
material-specific

445
00:18:54,440 --> 00:18:55,815
when we start
talking about TCOs.

446
00:18:55,815 --> 00:18:57,356
I'd rather not
generalize about them.

447
00:18:57,356 --> 00:18:58,975
About metals is more
straightforward.

448
00:18:58,975 --> 00:18:59,474
ASHLEY: OK.

449
00:18:59,474 --> 00:19:02,286
I guess then, for
metals, you would

450
00:19:02,286 --> 00:19:07,064
have to choose a metal that
melts before your semiconductor

451
00:19:07,064 --> 00:19:07,980
melt-- or [INAUDIBLE].

452
00:19:07,980 --> 00:19:09,896
TONIO BUONASSISI:
Typically, you wouldn't have

453
00:19:09,896 --> 00:19:12,470
your mental melting, per se.

454
00:19:12,470 --> 00:19:15,150
What you're looking for is
a chemical reaction here

455
00:19:15,150 --> 00:19:16,780
at this interface.

456
00:19:16,780 --> 00:19:20,080
You're looking for
the metal, most often,

457
00:19:20,080 --> 00:19:23,460
to form a binary compound
between your metal

458
00:19:23,460 --> 00:19:25,490
and the semiconducting
material underneath.

459
00:19:25,490 --> 00:19:27,889
Let's say you have
nickel and silicon.

460
00:19:27,889 --> 00:19:29,930
You'd be looking for it
to form a nickel silicide

461
00:19:29,930 --> 00:19:30,700
at that interface.

462
00:19:30,700 --> 00:19:31,230
ASHLEY: OK.

463
00:19:31,230 --> 00:19:31,620
TONIO BUONASSISI:
And that's where

464
00:19:31,620 --> 00:19:33,453
you go to your phase
diagrams and figure out

465
00:19:33,453 --> 00:19:38,120
what temperature those, say,
intermetallics should form.

466
00:19:38,120 --> 00:19:40,450
And then you're targeting
that particular temperature

467
00:19:40,450 --> 00:19:41,600
during your ramp.

468
00:19:41,600 --> 00:19:42,100
ASHLEY: OK.

469
00:19:42,100 --> 00:19:42,670
TONIO BUONASSISI:
And of course, it's

470
00:19:42,670 --> 00:19:44,044
not only phase
diagrams which are

471
00:19:44,044 --> 00:19:47,310
under equilibrium conditions,
kinetics are involved as well.

472
00:19:47,310 --> 00:19:49,920
So it becomes a
rather tricky process.

473
00:19:49,920 --> 00:19:51,238
Yeah, question in the back.

474
00:19:51,238 --> 00:19:53,150
AUDIENCE: This
picture is drawn where

475
00:19:53,150 --> 00:19:55,580
it says doping is concentrated
underneath [INAUDIBLE].

476
00:19:55,580 --> 00:19:56,496
TONIO BUONASSISI: Yep.

477
00:19:56,496 --> 00:19:58,204
AUDIENCE: How do we
manage to concentrate

478
00:19:58,204 --> 00:20:01,419
the dopant in just that
certain section [INAUDIBLE]?

479
00:20:01,419 --> 00:20:02,960
TONIO BUONASSISI:
Very good question.

480
00:20:02,960 --> 00:20:05,457
So we'll explain toward
the end of lecture

481
00:20:05,457 --> 00:20:07,290
why dopant is concentrated
underneath there.

482
00:20:07,290 --> 00:20:08,770
For those who are a
little bit more advanced,

483
00:20:08,770 --> 00:20:11,160
it relates to a tunneling
junction effect or a field

484
00:20:11,160 --> 00:20:12,430
emission effect.

485
00:20:12,430 --> 00:20:14,740
But we'll explain what
that is for everyone else.

486
00:20:14,740 --> 00:20:18,880
The way you typically get an
enhanced dopant concentration

487
00:20:18,880 --> 00:20:20,579
right underneath the
metal, first, let's

488
00:20:20,579 --> 00:20:22,370
appreciate the length
scales involved here.

489
00:20:22,370 --> 00:20:24,340
The width of that
contact metal is probably

490
00:20:24,340 --> 00:20:26,920
on the order of
somewhere between,

491
00:20:26,920 --> 00:20:31,220
let's say, 80 and 120
microns, maybe a little more.

492
00:20:31,220 --> 00:20:33,590
And so, at those
length scales, we

493
00:20:33,590 --> 00:20:35,810
don't need photo
lithography to really nail

494
00:20:35,810 --> 00:20:37,200
a particular location.

495
00:20:37,200 --> 00:20:39,230
We're not talking about
tens of nanometers.

496
00:20:39,230 --> 00:20:41,390
It's really something
more macroscopic.

497
00:20:41,390 --> 00:20:43,660
So we can do it in a few ways.

498
00:20:43,660 --> 00:20:47,040
On the high end, we could
use an ion implantation tool

499
00:20:47,040 --> 00:20:50,050
to pattern our
material with a mask.

500
00:20:50,050 --> 00:20:56,970
On the low end, we might diffuse
in the emitter very deeply,

501
00:20:56,970 --> 00:20:59,900
to create a highly doped
region everywhere, and then

502
00:20:59,900 --> 00:21:04,420
create a mask using not exactly
a photo lithography process.

503
00:21:04,420 --> 00:21:07,170
It could be a much simpler
contacting process,

504
00:21:07,170 --> 00:21:09,890
and then etch away some of
the emitter in other regions

505
00:21:09,890 --> 00:21:14,190
to make it more lightly doped
or, even in silicon's case,

506
00:21:14,190 --> 00:21:16,700
create a poor silicon
layer and etch that off,

507
00:21:16,700 --> 00:21:19,000
to distinguish between
the heavily doped

508
00:21:19,000 --> 00:21:20,420
and the lightly doped regions.

509
00:21:20,420 --> 00:21:23,130
There are a few different
ways to skin that cat.

510
00:21:23,130 --> 00:21:27,430
And that's one of the beauties
of solar cell processing

511
00:21:27,430 --> 00:21:29,880
is that it really pertains
to the specific material

512
00:21:29,880 --> 00:21:31,109
system involved.

513
00:21:31,109 --> 00:21:33,150
But the length scales are
such that you're really

514
00:21:33,150 --> 00:21:37,900
not typically limited
by photolithography.

515
00:21:37,900 --> 00:21:40,440
You can probably get around
that and use other techniques

516
00:21:40,440 --> 00:21:41,672
for masking it off.

517
00:21:41,672 --> 00:21:43,440
Mm-hm.

518
00:21:43,440 --> 00:21:44,840
Great.

519
00:21:44,840 --> 00:21:48,710
I see folks are awake today,
despite the P-set due.

520
00:21:48,710 --> 00:21:52,820
Sketch the I-V characteristics
of Schottky and Ohmic contacts.

521
00:21:52,820 --> 00:21:55,650
OK, so what we're
going to do here

522
00:21:55,650 --> 00:22:00,885
is talk about how Schottky and
Ohmic contacts come into being.

523
00:22:00,885 --> 00:22:02,090
And first off, define them.

524
00:22:02,090 --> 00:22:03,640
So let's define them.

525
00:22:03,640 --> 00:22:08,530
An Ohmic contact is one in
which I sweep my voltage,

526
00:22:08,530 --> 00:22:11,650
and I obtain a linear
current response.

527
00:22:11,650 --> 00:22:13,410
So this is pretty
straightforward.

528
00:22:13,410 --> 00:22:16,680
It's Ohm's law, follows
Ohm's law, so V equals IR.

529
00:22:16,680 --> 00:22:18,480
This is a linear
relationship then

530
00:22:18,480 --> 00:22:20,730
between voltage and
current, the slope of which

531
00:22:20,730 --> 00:22:23,160
is dictated by the resistance.

532
00:22:23,160 --> 00:22:25,490
A Schottky contact,
on the other hand,

533
00:22:25,490 --> 00:22:28,480
follows this beautiful
exponential curve

534
00:22:28,480 --> 00:22:31,180
that we've come to know and love
from our ideal diode equation.

535
00:22:31,180 --> 00:22:33,900
As a matter of fact, it
follows that same expression

536
00:22:33,900 --> 00:22:34,880
rather nicely.

537
00:22:34,880 --> 00:22:38,110
There's an exponential relation
between voltage and current.

538
00:22:38,110 --> 00:22:39,994
So whenever we have an
exponential relation

539
00:22:39,994 --> 00:22:41,660
between voltage and
current, now that we

540
00:22:41,660 --> 00:22:43,740
know how a PN
junciton works, and we

541
00:22:43,740 --> 00:22:46,430
recall that it's really that
diffusion current that's

542
00:22:46,430 --> 00:22:50,110
driving the forward
current right here

543
00:22:50,110 --> 00:22:51,780
under forward bias conditions.

544
00:22:51,780 --> 00:22:54,710
We've learned that, when
the barrier's too large,

545
00:22:54,710 --> 00:22:56,700
the electrons just
can't make it over.

546
00:22:56,700 --> 00:22:58,900
But as a barrier
begins dropping,

547
00:22:58,900 --> 00:23:01,100
an exponentially increasing
number of electrons

548
00:23:01,100 --> 00:23:03,630
can jump over that
barrier and move

549
00:23:03,630 --> 00:23:05,662
from the region of
high concentration

550
00:23:05,662 --> 00:23:07,120
to the region of
low concentration,

551
00:23:07,120 --> 00:23:10,050
or from the n-type material
to the p-type material.

552
00:23:10,050 --> 00:23:14,250
So when we see an exponential
current voltage response,

553
00:23:14,250 --> 00:23:17,170
we should immediately think
about some barrier involved

554
00:23:17,170 --> 00:23:18,460
in our system.

555
00:23:18,460 --> 00:23:22,510
And as we begin
biasing this device--

556
00:23:22,510 --> 00:23:25,170
in this case, a semiconductor
metal junction--

557
00:23:25,170 --> 00:23:28,510
as we begin biasing the
semiconductor versus the metal,

558
00:23:28,510 --> 00:23:32,210
we begin seeing an
exponential current response.

559
00:23:32,210 --> 00:23:33,680
We should imagine
that there must

560
00:23:33,680 --> 00:23:37,560
be some barrier in between the
metal and the semiconductor.

561
00:23:37,560 --> 00:23:40,150
And we'll explain how that
barrier comes into being.

562
00:23:40,150 --> 00:23:43,250
So again, just to recap, making
sure we hit all the points.

563
00:23:43,250 --> 00:23:44,960
Ohmic, a linear I-V curve.

564
00:23:44,960 --> 00:23:47,570
And Ohmic contacts
are typically used

565
00:23:47,570 --> 00:23:49,800
when charge separation
is not the goal

566
00:23:49,800 --> 00:23:51,040
for your metallization.

567
00:23:51,040 --> 00:23:53,380
Let's say you've already
achieved your charge separation

568
00:23:53,380 --> 00:23:55,160
through the PN junction,
and now, you just

569
00:23:55,160 --> 00:23:56,576
want to contact
your semiconductor

570
00:23:56,576 --> 00:23:59,160
to extract the charge,
you'd use an Ohmic contact.

571
00:23:59,160 --> 00:24:01,100
But you would use
a Schottky contact

572
00:24:01,100 --> 00:24:05,350
when you want to enforce
some charge separation.

573
00:24:05,350 --> 00:24:07,320
Typically, not quite
as good as, say,

574
00:24:07,320 --> 00:24:10,080
a PN junction for
separating charge,

575
00:24:10,080 --> 00:24:11,780
not quite as high voltage.

576
00:24:11,780 --> 00:24:16,890
But it's a useful tool,
for example, in research.

577
00:24:16,890 --> 00:24:20,430
So we're going to describe
what fundamental material

578
00:24:20,430 --> 00:24:23,640
parameters determine the
I-V characteristics of both

579
00:24:23,640 --> 00:24:27,830
the Ohmic and Schottky
junctions, more generally,

580
00:24:27,830 --> 00:24:30,650
a contact
semiconductor junction.

581
00:24:30,650 --> 00:24:34,160
And this dives into
Schottky band theory.

582
00:24:34,160 --> 00:24:38,240
This is a very idealized
version of the real world.

583
00:24:38,240 --> 00:24:39,790
It's an idealized
version, because it

584
00:24:39,790 --> 00:24:42,690
gets us started and points
us in the right direction.

585
00:24:42,690 --> 00:24:44,630
Then we'll add
layers of complexity,

586
00:24:44,630 --> 00:24:46,300
until we build up
to really understand

587
00:24:46,300 --> 00:24:48,050
what's going on inside
of these devices.

588
00:24:48,050 --> 00:24:48,883
So let's start here.

589
00:24:48,883 --> 00:24:55,410
This is a very nice, elegant
view of contact theory.

590
00:24:55,410 --> 00:24:56,910
We have our
semiconductor over here.

591
00:24:56,910 --> 00:24:57,650
Let's walk through this.

592
00:24:57,650 --> 00:24:59,270
So this is our
energy band diagram.

593
00:24:59,270 --> 00:25:01,130
We have E on the
vertical axis that

594
00:25:01,130 --> 00:25:03,260
represents the energy
of the electron,

595
00:25:03,260 --> 00:25:06,250
versus X, some real
space parameter.

596
00:25:06,250 --> 00:25:07,924
We have our semiconductor
material here.

597
00:25:07,924 --> 00:25:09,840
We have our valence band,
our conduction band,

598
00:25:09,840 --> 00:25:10,880
and our Fermi energy.

599
00:25:10,880 --> 00:25:14,630
This looks to be an n-type
semiconducting material.

600
00:25:14,630 --> 00:25:16,950
And we have our vacuum level.

601
00:25:16,950 --> 00:25:18,700
The vacuum level is
essentially the energy

602
00:25:18,700 --> 00:25:21,010
at which we remove an electron
from the semiconductor.

603
00:25:21,010 --> 00:25:24,750
If the semiconductor were
placed, let's say, in a vacuum,

604
00:25:24,750 --> 00:25:29,700
and we were to excite an
electron out of the material

605
00:25:29,700 --> 00:25:31,730
through some, let's say,
a photoemission effect,

606
00:25:31,730 --> 00:25:35,270
we would have to overcome
the energy between the Fermi

607
00:25:35,270 --> 00:25:36,850
energy and the vacuum level.

608
00:25:36,850 --> 00:25:40,430
This xi right here is
essentially the electron

609
00:25:40,430 --> 00:25:42,280
affinity of the semiconductor.

610
00:25:42,280 --> 00:25:46,100
That's the delta in
energy, or q times xi

611
00:25:46,100 --> 00:25:47,980
is the delta in energy
between the vacuum

612
00:25:47,980 --> 00:25:49,990
level and the conduction band.

613
00:25:49,990 --> 00:25:51,740
So we have all of the
variables mapped out

614
00:25:51,740 --> 00:25:52,820
for the semiconductor.

615
00:25:52,820 --> 00:25:54,920
Most of them should be
familiar for us already.

616
00:25:54,920 --> 00:25:58,600
We've added the vacuum
level for completeness.

617
00:25:58,600 --> 00:26:00,910
The metal, we also
have a Fermi energy,

618
00:26:00,910 --> 00:26:03,530
we also a chemical
potential of the metal.

619
00:26:03,530 --> 00:26:07,030
And we have a work function
of the metal, right here.

620
00:26:07,030 --> 00:26:10,550
So this work function
is defined as the energy

621
00:26:10,550 --> 00:26:13,850
necessary to remove an
electron from the metal.

622
00:26:13,850 --> 00:26:17,780
Say again, by if you shined
a light on the material,

623
00:26:17,780 --> 00:26:20,670
it has a high enough energy,
you can begin exciting electrons

624
00:26:20,670 --> 00:26:25,070
off of the metal with
a photoemission effect.

625
00:26:25,070 --> 00:26:28,030
And so we have our vacuum levels
in this particular diagram

626
00:26:28,030 --> 00:26:28,790
lined up.

627
00:26:28,790 --> 00:26:30,690
And our chemical
potentials are a little bit

628
00:26:30,690 --> 00:26:33,520
different between the metal
and the semiconductor.

629
00:26:33,520 --> 00:26:37,580
So what do you think will happen
when we put the two together?

630
00:26:37,580 --> 00:26:39,680
If we put the two together,
the chemical potential

631
00:26:39,680 --> 00:26:42,570
throughout the entire system
has to be the same, right?

632
00:26:42,570 --> 00:26:47,136
Because now, they're in good
contact with one another.

633
00:26:47,136 --> 00:26:48,510
But what about
that vacuum level?

634
00:26:48,510 --> 00:26:50,070
What's going to happen?

635
00:26:50,070 --> 00:26:51,820
Will there be
discontinuity, first of all,

636
00:26:51,820 --> 00:26:53,929
in the vacuum level?

637
00:26:53,929 --> 00:26:55,970
Can there be a discontinuity
in the vacuum level?

638
00:26:55,970 --> 00:26:59,670
If I move one little delta
x from the semiconductor

639
00:26:59,670 --> 00:27:01,350
into the metal,
do I expect there

640
00:27:01,350 --> 00:27:03,170
to be a discontinuity
in the vacuum level?

641
00:27:03,170 --> 00:27:04,680
No, probably not.

642
00:27:04,680 --> 00:27:07,930
So there has to be some smooth
change in the vacuum energy,

643
00:27:07,930 --> 00:27:10,570
relative to the Fermi energy.

644
00:27:10,570 --> 00:27:12,337
And if the vacuum
level changes, that

645
00:27:12,337 --> 00:27:14,670
means that the conduction
band and valence band energies

646
00:27:14,670 --> 00:27:17,800
are also going to change,
relative to the Fermi energy,

647
00:27:17,800 --> 00:27:20,810
because the valence band
and conduction band energies

648
00:27:20,810 --> 00:27:25,650
track with the vacuum level--
the vacuum level being defined

649
00:27:25,650 --> 00:27:27,810
as the amount of energy
necessary to remove

650
00:27:27,810 --> 00:27:29,220
the electron from the system.

651
00:27:29,220 --> 00:27:31,470
So what you get when you
put these two materials

652
00:27:31,470 --> 00:27:34,820
together is something
like this right here.

653
00:27:34,820 --> 00:27:38,029
This dashed blue line
representing the Fermi energy,

654
00:27:38,029 --> 00:27:40,320
or the chemical potential
throughout the entire system,

655
00:27:40,320 --> 00:27:41,870
is the same.

656
00:27:41,870 --> 00:27:44,690
In this particular case,
we didn't bias our device.

657
00:27:44,690 --> 00:27:49,600
We're not applying a
battery or some bias voltage

658
00:27:49,600 --> 00:27:51,500
between one side and the other.

659
00:27:51,500 --> 00:27:54,750
So we have the same chemical
potential throughout.

660
00:27:54,750 --> 00:27:56,940
If we have the same chemical
potential throughout,

661
00:27:56,940 --> 00:27:59,550
notice that we had
to push this up,

662
00:27:59,550 --> 00:28:01,639
or push the Fermi
energy up, to reach--

663
00:28:01,639 --> 00:28:04,180
and of course, the work function
in the metal's not changing,

664
00:28:04,180 --> 00:28:05,990
and so the vacuum
energy also goes up.

665
00:28:05,990 --> 00:28:08,240
But since there can't be a
discontinuity in the vacuum

666
00:28:08,240 --> 00:28:11,700
energy, you see a little
bit of a rise of the vacuum

667
00:28:11,700 --> 00:28:13,290
energy in the semiconductor.

668
00:28:13,290 --> 00:28:16,430
And of course, the valence band
and conduction bands follow.

669
00:28:16,430 --> 00:28:18,540
And that results
in this little rise

670
00:28:18,540 --> 00:28:20,380
right here, right
next to the interface

671
00:28:20,380 --> 00:28:22,950
between the semiconductor
and the metal.

672
00:28:22,950 --> 00:28:24,530
Any questions so far?

673
00:28:24,530 --> 00:28:25,970
AUDIENCE: Is there
a discontinuity

674
00:28:25,970 --> 00:28:28,370
in the first derivative
of the vacuum level?

675
00:28:28,370 --> 00:28:29,330
Is there a kink in it?

676
00:28:29,330 --> 00:28:30,300
TONIO BUONASSISI: Ha.

677
00:28:30,300 --> 00:28:31,466
A very interesting question.

678
00:28:31,466 --> 00:28:34,650
So typically, we draw it
as having a discontinuity

679
00:28:34,650 --> 00:28:35,750
in the first derivative.

680
00:28:35,750 --> 00:28:37,820
We would draw a very sharp
interface right there.

681
00:28:37,820 --> 00:28:41,405
But in a practical sense,
I'll reserve judgment

682
00:28:41,405 --> 00:28:43,780
on that, until I think a little
bit more deeply about it.

683
00:28:43,780 --> 00:28:46,380
It's difficult for me to
believe that there would be,

684
00:28:46,380 --> 00:28:49,010
if you zoom in very, very
closely, very finely.

685
00:28:52,430 --> 00:28:56,050
The bending, let's think
about it from another way,

686
00:28:56,050 --> 00:28:57,460
from what we already know.

687
00:28:57,460 --> 00:29:00,060
This is essentially
the second integral

688
00:29:00,060 --> 00:29:03,450
of the charge distribution
across that interface.

689
00:29:03,450 --> 00:29:07,310
And if the charge distribution
is sharp enough, sure,

690
00:29:07,310 --> 00:29:10,870
I suppose you could
get some difference.

691
00:29:10,870 --> 00:29:15,210
But I think-- I'll have
to research judgement,

692
00:29:15,210 --> 00:29:17,400
but my gut is telling
me that there would not

693
00:29:17,400 --> 00:29:21,280
be at a very fine
microscopic level.

694
00:29:21,280 --> 00:29:23,970
In general, though, we draw
it looking much like this.

695
00:29:26,814 --> 00:29:28,710
OK?

696
00:29:28,710 --> 00:29:31,740
Any other questions about this,
about how we constructed it,

697
00:29:31,740 --> 00:29:34,720
using the Anderson method?

698
00:29:34,720 --> 00:29:35,220
No?

699
00:29:35,220 --> 00:29:36,090
OK.

700
00:29:36,090 --> 00:29:37,990
Why don't we continue?

701
00:29:37,990 --> 00:29:40,804
What we've done
right here is fill

702
00:29:40,804 --> 00:29:43,220
in those diagrams, the same
ones that you were looking at,

703
00:29:43,220 --> 00:29:45,410
with many more variables.

704
00:29:45,410 --> 00:29:48,180
The purpose of this
is to get to the point

705
00:29:48,180 --> 00:29:52,330
where we can describe
the built-in voltage.

706
00:29:52,330 --> 00:29:53,840
Remember, we
mentioned that, if we

707
00:29:53,840 --> 00:29:56,040
have some exponential
current voltage response,

708
00:29:56,040 --> 00:29:59,140
there should be some barrier
embedded inside of our system.

709
00:29:59,140 --> 00:30:01,320
And this right here
is that barrier.

710
00:30:01,320 --> 00:30:05,390
So these series of
variables are teaching us

711
00:30:05,390 --> 00:30:08,490
how to derive it,
how to discover

712
00:30:08,490 --> 00:30:10,990
how large that barrier is.

713
00:30:10,990 --> 00:30:12,700
And this other
component right here,

714
00:30:12,700 --> 00:30:14,866
which is the width of the
space charge region, which

715
00:30:14,866 --> 00:30:19,750
is the width of this, so-called,
curved region, if you will,

716
00:30:19,750 --> 00:30:22,950
is also going to
be of importance.

717
00:30:22,950 --> 00:30:26,250
So first off, let's
think about it this way.

718
00:30:26,250 --> 00:30:28,980
If we are looking at this
system from the electron's

719
00:30:28,980 --> 00:30:31,720
perspective, we have
an n-type semiconductor

720
00:30:31,720 --> 00:30:36,070
right here coming up
against the metal.

721
00:30:36,070 --> 00:30:38,620
If our work function
of the metal

722
00:30:38,620 --> 00:30:40,870
is larger than the
work function--

723
00:30:40,870 --> 00:30:44,170
or the so-called work
function of the semiconductor,

724
00:30:44,170 --> 00:30:46,282
which is really the
electron affinity

725
00:30:46,282 --> 00:30:48,240
plus the delta in energy
between the conduction

726
00:30:48,240 --> 00:30:49,660
band of the Fermi energy.

727
00:30:49,660 --> 00:30:52,300
But if the work function of
the metal is larger than that,

728
00:30:52,300 --> 00:30:55,430
we can see that the electrons
should very easily plop down

729
00:30:55,430 --> 00:30:56,650
into the metal.

730
00:30:56,650 --> 00:31:00,774
And there shouldn't be-- let's
see, in that particular case,

731
00:31:00,774 --> 00:31:02,190
work function of
the mental should

732
00:31:02,190 --> 00:31:05,360
result in an Ohmic contact
for the minority carriers,

733
00:31:05,360 --> 00:31:06,550
in that case.

734
00:31:06,550 --> 00:31:09,150
But in terms of
the barrier height,

735
00:31:09,150 --> 00:31:11,690
this barrier height right
here would be determined

736
00:31:11,690 --> 00:31:14,830
by that equation there.

737
00:31:14,830 --> 00:31:19,280
The contact potential,
the built-in bias voltage

738
00:31:19,280 --> 00:31:22,280
would be here, and the
barrier height there.

739
00:31:22,280 --> 00:31:24,550
And the width of the space
charge region, once again,

740
00:31:24,550 --> 00:31:26,450
is dependent on
the dopant density

741
00:31:26,450 --> 00:31:27,890
inside of the semiconductor.

742
00:31:27,890 --> 00:31:29,920
Here, we've defined it
as Nd, because we're

743
00:31:29,920 --> 00:31:33,050
assuming that this is an n-type
semiconductor with donor atoms.

744
00:31:33,050 --> 00:31:35,175
But of course, if this were
a p-type semiconductor,

745
00:31:35,175 --> 00:31:38,360
we'd just replace this with
Na to determine the width.

746
00:31:38,360 --> 00:31:40,500
And if the dopant
density is small,

747
00:31:40,500 --> 00:31:42,690
this means that the width
of the space charge region

748
00:31:42,690 --> 00:31:43,642
is going to be large.

749
00:31:43,642 --> 00:31:45,100
And that makes
sense, because there

750
00:31:45,100 --> 00:31:47,210
has to be an equal
amount of charge

751
00:31:47,210 --> 00:31:48,630
on either side of the junction.

752
00:31:48,630 --> 00:31:51,910
And if there is less charge
per unit volume here,

753
00:31:51,910 --> 00:31:54,310
you need a larger volume
to compensate the charge

754
00:31:54,310 --> 00:31:55,760
on the other side.

755
00:31:55,760 --> 00:31:59,910
So this is, again, a similar
expression to the one

756
00:31:59,910 --> 00:32:02,990
that we saw when we were
studying the PN junctions.

757
00:32:02,990 --> 00:32:05,860
Now, the important
thing to recognize here

758
00:32:05,860 --> 00:32:09,250
is that, if we have the bands
bending this way, for electrons

759
00:32:09,250 --> 00:32:12,190
in our system, there may
be a barrier to hop out.

760
00:32:12,190 --> 00:32:15,010
For holes in the
system, though, this

761
00:32:15,010 --> 00:32:18,479
could be a favorable
exit strategy,

762
00:32:18,479 --> 00:32:19,770
if the bands are bent that way.

763
00:32:19,770 --> 00:32:21,310
If the bends are
bent the other way,

764
00:32:21,310 --> 00:32:23,640
electrons can rather
easily roll down.

765
00:32:23,640 --> 00:32:27,070
But holes will have a barrier
to go out of the material.

766
00:32:27,070 --> 00:32:28,790
So the same type
of metal that might

767
00:32:28,790 --> 00:32:31,620
make an Ohmic contact
for an n-type material

768
00:32:31,620 --> 00:32:36,510
might make a Schottky contact
for a p-type material.

769
00:32:36,510 --> 00:32:38,480
Yeah, that's about it.

770
00:32:38,480 --> 00:32:40,740
So we have to be very careful
in terms of our matching

771
00:32:40,740 --> 00:32:43,560
or pairing between the
specific type of metal

772
00:32:43,560 --> 00:32:46,870
and the specific
semiconductor involved.

773
00:32:46,870 --> 00:32:52,400
Let me put that into
perspective here with something

774
00:32:52,400 --> 00:32:54,220
we've already seen.

775
00:32:54,220 --> 00:32:56,990
Again, the Ohmic contact is one
in which we have a linear IV

776
00:32:56,990 --> 00:32:57,950
characteristic.

777
00:32:57,950 --> 00:33:00,310
And the electron barrier
height has to be less than

778
00:33:00,310 --> 00:33:05,320
or equal to 0, to have the
electrons very easily pop down

779
00:33:05,320 --> 00:33:06,257
into the contact.

780
00:33:06,257 --> 00:33:07,840
Whereas, for a
Schottky contact, there

781
00:33:07,840 --> 00:33:09,090
should be some barrier height.

782
00:33:09,090 --> 00:33:10,190
It should be finite.

783
00:33:10,190 --> 00:33:13,320
And we should have a difficulty
of moving the electron

784
00:33:13,320 --> 00:33:18,100
from the semiconductor into the
metal, in that particular case.

785
00:33:18,100 --> 00:33:23,850
So the natural question is,
what metals should we use?

786
00:33:23,850 --> 00:33:26,020
Or what is the range of
metals that are available?

787
00:33:26,020 --> 00:33:28,540
If the barrier height is a
function of the work function

788
00:33:28,540 --> 00:33:30,590
of the metal, what is the
range of work functions

789
00:33:30,590 --> 00:33:32,260
that are available to us?

790
00:33:32,260 --> 00:33:35,430
On this green arrow right
here, which is Q times the work

791
00:33:35,430 --> 00:33:39,180
function of the metal, we have
a range of different metals,

792
00:33:39,180 --> 00:33:43,010
from aluminum to platinum,
and titanium, zinc, tungsten,

793
00:33:43,010 --> 00:33:45,219
molybdenum, copper,
nickel, gold, in between.

794
00:33:45,219 --> 00:33:47,010
And of course, there
are other metals, too,

795
00:33:47,010 --> 00:33:49,450
in the periodic table that
you can add to this chart.

796
00:33:49,450 --> 00:33:52,350
But this gives you
a sense of where

797
00:33:52,350 --> 00:33:54,270
the Fermi energy of
the metal would lie,

798
00:33:54,270 --> 00:33:56,280
relative to the
vacuum level, and how

799
00:33:56,280 --> 00:33:59,190
that would match up against
your arbitrary semiconductor.

800
00:33:59,190 --> 00:34:01,790
This one here happens to be
a silicon carbide material.

801
00:34:01,790 --> 00:34:04,290
But if you happen to be working
on a different semiconductor

802
00:34:04,290 --> 00:34:06,790
and place it right here, you
can see where your Fermi energy

803
00:34:06,790 --> 00:34:09,620
would lie, relative to the
work function of the metal,

804
00:34:09,620 --> 00:34:11,274
if you line up
the vacuum levels.

805
00:34:11,274 --> 00:34:12,690
And this gives you
an idea already

806
00:34:12,690 --> 00:34:16,949
of how the bands
will bend, once you

807
00:34:16,949 --> 00:34:22,580
adjust the position of the metal
to match up the Fermi energy

808
00:34:22,580 --> 00:34:24,532
inside of your
semiconductor material.

809
00:34:24,532 --> 00:34:25,032
Yeah.

810
00:34:25,032 --> 00:34:28,364
AUDIENCE: So on the left is
an intrinsic semiconductor?

811
00:34:28,364 --> 00:34:29,320
Is it?

812
00:34:29,320 --> 00:34:30,494
TONIO BUONASSISI: Yeah, it's
pretty close to intrinsic.

813
00:34:30,494 --> 00:34:31,993
It's a little n-type,
it looks like.

814
00:34:31,993 --> 00:34:33,650
But it's pretty
close to intrinsic.

815
00:34:33,650 --> 00:34:38,066
This, in this case, just a
variant of silicon carbide.

816
00:34:38,066 --> 00:34:39,940
There are many polymorphs
of silicon carbide,

817
00:34:39,940 --> 00:34:44,896
meaning same composition,
same silicon carbon ratio,

818
00:34:44,896 --> 00:34:46,270
but different
crystal structures,

819
00:34:46,270 --> 00:34:49,190
depending on how it was grown.

820
00:34:49,190 --> 00:34:49,691
Yep?

821
00:34:49,691 --> 00:34:51,606
AUDIENCE: Do you know
where silver would fall?

822
00:34:51,606 --> 00:34:52,830
TONIO BUONASSISI: Silver?

823
00:34:52,830 --> 00:34:54,150
AUDIENCE: It's not
listed [INAUDIBLE].

824
00:34:54,150 --> 00:34:54,840
TONIO BUONASSISI:
No, it's listed.

825
00:34:54,840 --> 00:34:56,702
AUDIENCE: [INAUDIBLE]
a lot for [INAUDIBLE].

826
00:34:56,702 --> 00:34:57,660
TONIO BUONASSISI: Yeah.

827
00:34:57,660 --> 00:34:59,160
I'm going to have
to guess that it's

828
00:34:59,160 --> 00:35:01,640
going to fall around here,
just from a similarity

829
00:35:01,640 --> 00:35:04,110
of other elemental
species in that list.

830
00:35:04,110 --> 00:35:06,600
Do you happen to remember the
work function of silver, Joe?

831
00:35:06,600 --> 00:35:08,035
Been working with
it a lot lately.

832
00:35:08,035 --> 00:35:08,534
JOE: Yeah.

833
00:35:08,534 --> 00:35:09,075
I don't know.

834
00:35:09,075 --> 00:35:11,200
TONIO BUONASSISI: OK.

835
00:35:11,200 --> 00:35:12,800
Quick, interesting
aside about silver,

836
00:35:12,800 --> 00:35:14,170
since it was brought up.

837
00:35:14,170 --> 00:35:17,220
It is used very often in contact
metallization in solar cells.

838
00:35:17,220 --> 00:35:20,080
As a matter of fact,
a bit too much now.

839
00:35:20,080 --> 00:35:22,780
So much so that it's driving
up the price of silver

840
00:35:22,780 --> 00:35:24,342
on the market.

841
00:35:24,342 --> 00:35:26,800
We talked about this, I think,
in a few classes ago, right?

842
00:35:26,800 --> 00:35:28,684
10%, approximately,
of all silver

843
00:35:28,684 --> 00:35:31,100
worldwide is currently being
used in contact metallization

844
00:35:31,100 --> 00:35:32,670
in solar cells.

845
00:35:32,670 --> 00:35:34,720
That's a pretty high number.

846
00:35:34,720 --> 00:35:39,480
OK, so we have a variety
of metals to choose from,

847
00:35:39,480 --> 00:35:41,310
almost like a menu.

848
00:35:41,310 --> 00:35:44,080
And then the semiconductor
here, and we can see, OK,

849
00:35:44,080 --> 00:35:47,840
if we make contact-- and
let's say we pick nickel,

850
00:35:47,840 --> 00:35:50,220
in this case-- if the
Fermi energies line up,

851
00:35:50,220 --> 00:35:52,540
that means we shift
the metal down.

852
00:35:52,540 --> 00:35:55,030
That means that our vacuum
level is going to fall down here

853
00:35:55,030 --> 00:35:56,190
in the semiconductor.

854
00:35:56,190 --> 00:35:59,390
And for electrons, at least,
will be very easily contacting

855
00:35:59,390 --> 00:36:02,180
this silicon carbide material.

856
00:36:02,180 --> 00:36:07,010
So from a simple energy
band diagram point of view,

857
00:36:07,010 --> 00:36:09,824
we can begin to see which
contacts will create

858
00:36:09,824 --> 00:36:11,240
Ohmic contacts,
and which contacts

859
00:36:11,240 --> 00:36:14,470
will create Schottky contacts.

860
00:36:14,470 --> 00:36:17,950
That's in the ideal world.

861
00:36:17,950 --> 00:36:21,970
In reality-- well, let me
make sure that we grasp this

862
00:36:21,970 --> 00:36:22,960
before we move on.

863
00:36:26,780 --> 00:36:30,224
Any questions
concerning this so far?

864
00:36:30,224 --> 00:36:31,637
Yeah?

865
00:36:31,637 --> 00:36:33,220
AUDIENCE: How would
we be able to tell

866
00:36:33,220 --> 00:36:35,382
if it's Ohmic or
Schottky in this sense?

867
00:36:35,382 --> 00:36:36,340
TONIO BUONASSISI: Sure.

868
00:36:36,340 --> 00:36:36,700
Sure.

869
00:36:36,700 --> 00:36:37,050
AUDIENCE: [INAUDIBLE].

870
00:36:37,050 --> 00:36:37,570
TONIO BUONASSISI: Absolutely.

871
00:36:37,570 --> 00:36:40,420
So from this diagram right
here, let's walk through it.

872
00:36:40,420 --> 00:36:45,549
So if I, say, have a
semicondting-- let's first line

873
00:36:45,549 --> 00:36:47,590
up the vacuum levels, like
we've done right here.

874
00:36:47,590 --> 00:36:51,290
That's a good first
step on, what we call,

875
00:36:51,290 --> 00:36:53,010
the Anderson method
of identifying

876
00:36:53,010 --> 00:36:56,470
with the band diagram or band
structure will look like.

877
00:36:56,470 --> 00:36:59,240
We start with the
vacuum levels lined up.

878
00:36:59,240 --> 00:37:03,321
I'm going to write
to this vacuum.

879
00:37:03,321 --> 00:37:06,520
And I'm going to dray my
semiconductor on one side.

880
00:37:06,520 --> 00:37:11,920
So let's say I have conduction
band and bands bent like so.

881
00:37:11,920 --> 00:37:15,749
And I have my Fermi
energy like so, Ef.

882
00:37:15,749 --> 00:37:16,915
So this is my semiconductor.

883
00:37:21,370 --> 00:37:25,817
And then I have an
imaginary barrier,

884
00:37:25,817 --> 00:37:27,400
and I have my metal
on the other side.

885
00:37:30,660 --> 00:37:35,380
So now, let's say we have
a metal that has a work

886
00:37:35,380 --> 00:37:38,480
function somewhere up here.

887
00:37:38,480 --> 00:37:40,640
So this is the work
function of the metal.

888
00:37:40,640 --> 00:37:43,890
We'll call it the Fermi
energy inside of the metal.

889
00:37:43,890 --> 00:37:46,790
What happens when I put
these two materials together?

890
00:37:46,790 --> 00:37:49,835
Well, step one is to remove
that imaginary barrier

891
00:37:49,835 --> 00:37:51,014
in between them.

892
00:37:51,014 --> 00:37:52,430
And now, you begin
to look at this

893
00:37:52,430 --> 00:37:54,225
and say, well, if
they're in contact,

894
00:37:54,225 --> 00:37:55,940
in good electrical
contact, and there's

895
00:37:55,940 --> 00:37:57,830
no resistance
throughout the material,

896
00:37:57,830 --> 00:38:00,380
the chemical potential
should be identical

897
00:38:00,380 --> 00:38:03,210
all throughout the material,
from left to right.

898
00:38:03,210 --> 00:38:07,045
And so now, I merge into
phase two, which is to say,

899
00:38:07,045 --> 00:38:08,920
I've removed this barrier
here in the middle,

900
00:38:08,920 --> 00:38:12,400
and now my Fermi energy is going
to be equilibrated throughout.

901
00:38:12,400 --> 00:38:15,320
So my Ef is going to
be the same throughout.

902
00:38:15,320 --> 00:38:17,280
And so, far away from
the junction region here

903
00:38:17,280 --> 00:38:21,220
in the middle, I'm going to
draw my conduction band here

904
00:38:21,220 --> 00:38:25,550
and my vacuum level here.

905
00:38:25,550 --> 00:38:28,017
And now, far away
from the junction

906
00:38:28,017 --> 00:38:30,100
on the other side in the
metal, if my Fermi energy

907
00:38:30,100 --> 00:38:32,770
is here and my vacuum
energy is going

908
00:38:32,770 --> 00:38:35,710
to be somewhere
down around here.

909
00:38:40,020 --> 00:38:42,002
So semiconductor and metal.

910
00:38:42,002 --> 00:38:43,960
Now, what happens closer
to the interface right

911
00:38:43,960 --> 00:38:46,140
here-- this is the Fermi energy.

912
00:38:46,140 --> 00:38:47,570
This is the conduction band.

913
00:38:47,570 --> 00:38:49,130
This is the valence band.

914
00:38:49,130 --> 00:38:52,110
This is the vacuum level in
the semiconductor, the vacuum

915
00:38:52,110 --> 00:38:54,060
level in the middle.

916
00:38:54,060 --> 00:38:57,970
And like we just
described, the vacuum level

917
00:38:57,970 --> 00:38:59,430
has to be continuous.

918
00:38:59,430 --> 00:39:01,840
There has to be a
continuity throughout.

919
00:39:01,840 --> 00:39:04,950
And so what we see
here at the interface,

920
00:39:04,950 --> 00:39:07,620
since the density of charge
is very high in the metal,

921
00:39:07,620 --> 00:39:09,920
but lower in the
semiconductor, there's

922
00:39:09,920 --> 00:39:11,760
going to be band bending
at the interface.

923
00:39:11,760 --> 00:39:13,300
And there's going
to be a lot more

924
00:39:13,300 --> 00:39:15,450
bending on the bands
in the semiconductor

925
00:39:15,450 --> 00:39:18,170
than there will be in the metal,
as a result of the difference

926
00:39:18,170 --> 00:39:19,670
in charge densities.

927
00:39:19,670 --> 00:39:23,790
And so what you'll see is
something that looks like this.

928
00:39:23,790 --> 00:39:27,630
The vacuum level will drop as
it reaches that semiconductor

929
00:39:27,630 --> 00:39:32,040
metal interface, in order
to match either side.

930
00:39:32,040 --> 00:39:34,530
And the conduction
band and valence bands

931
00:39:34,530 --> 00:39:39,320
will track the vacuum level on
either side of that junction.

932
00:39:39,320 --> 00:39:42,470
So this is the junction
right here, in between.

933
00:39:42,470 --> 00:39:44,660
And now I look at this
situation right here,

934
00:39:44,660 --> 00:39:47,500
and I say, OK, if I have
an electron approaching

935
00:39:47,500 --> 00:39:50,360
this junction right
here, there is

936
00:39:50,360 --> 00:39:52,840
no barrier to going from the
semiconductor into the metal.

937
00:39:52,840 --> 00:39:55,570
As a matter of fact, there's
almost an energy gain

938
00:39:55,570 --> 00:39:58,410
to be had from going from the
semiconductor into the metal.

939
00:39:58,410 --> 00:40:00,520
But now, from the
hole's perspective,

940
00:40:00,520 --> 00:40:02,740
if I'm a hole right
here in the valence band

941
00:40:02,740 --> 00:40:04,650
and approaching this
junction, there's

942
00:40:04,650 --> 00:40:08,980
actually a repulsion away from
that interface, if I'm a hole.

943
00:40:08,980 --> 00:40:12,460
And so this particular junction
here allows electrons to pass,

944
00:40:12,460 --> 00:40:13,860
but repels holes.

945
00:40:13,860 --> 00:40:16,660
This is looking an
awful lot like a diode.

946
00:40:16,660 --> 00:40:20,160
So it's this
construction right here

947
00:40:20,160 --> 00:40:24,940
that allows you to tell--
asterisk, tell, asterisk,

948
00:40:24,940 --> 00:40:27,220
according to the
Schottky band diagram--

949
00:40:27,220 --> 00:40:29,460
whether or not
this should behave

950
00:40:29,460 --> 00:40:31,760
in an Ohmic fashion
or a Schottky

951
00:40:31,760 --> 00:40:35,690
fashion for the particular
karyotype that you're probing.

952
00:40:35,690 --> 00:40:38,890
And I say asterisks,
because, in real life,

953
00:40:38,890 --> 00:40:42,020
contacts are never that simple.

954
00:40:42,020 --> 00:40:46,744
This is an example right
here of the calculated-- make

955
00:40:46,744 --> 00:40:48,160
sure I get all my
variables right.

956
00:40:48,160 --> 00:40:48,220
Yeah.

957
00:40:48,220 --> 00:40:49,511
So work function in the middle.

958
00:40:49,511 --> 00:40:52,710
The calculated barrier height
for different metal species

959
00:40:52,710 --> 00:40:55,040
here on a certain type, a
certain polymorphous silicon

960
00:40:55,040 --> 00:40:55,880
carbide.

961
00:40:55,880 --> 00:41:00,260
And these are the major barrier
heights for different metals,

962
00:41:00,260 --> 00:41:04,000
titanium, nickel and
gold, at different faces

963
00:41:04,000 --> 00:41:05,240
of the silicon carbide.

964
00:41:05,240 --> 00:41:07,990
So if you have a
binary semiconductor,

965
00:41:07,990 --> 00:41:11,150
meaning a semiconductor
comprised of two elements,

966
00:41:11,150 --> 00:41:14,250
if you terminate at a surface
and you cut the plane just

967
00:41:14,250 --> 00:41:20,010
right, you could wind up
exposing a plane of silicon

968
00:41:20,010 --> 00:41:22,680
atoms or of carbon atoms,
if you have silicon carbide,

969
00:41:22,680 --> 00:41:25,150
let's say, depending
on where you cut.

970
00:41:25,150 --> 00:41:26,980
So if you cut right
here, you might expose

971
00:41:26,980 --> 00:41:28,140
a plane of silicon atoms.

972
00:41:28,140 --> 00:41:30,060
You cut one atomic
layer above, you

973
00:41:30,060 --> 00:41:31,920
might expose a row
of carbon atoms.

974
00:41:31,920 --> 00:41:33,920
Cut a row above that,
it's silicon again.

975
00:41:33,920 --> 00:41:36,420
And so depending on
what face you expose

976
00:41:36,420 --> 00:41:39,580
and depending on the
orientation of that face,

977
00:41:39,580 --> 00:41:43,830
your barrier height, relative to
the metal, could be different.

978
00:41:43,830 --> 00:41:45,040
So, hm.

979
00:41:45,040 --> 00:41:48,050
You glance at that, and you
begin thinking to yourself,

980
00:41:48,050 --> 00:41:50,640
OK, this is
macroscopic right here.

981
00:41:50,640 --> 00:41:52,910
This tells you, kind of
from a continuum point

982
00:41:52,910 --> 00:41:56,000
of view, what the interface
should behave like,

983
00:41:56,000 --> 00:41:57,800
what the context should be.

984
00:41:57,800 --> 00:42:00,340
But what I'm seeing
right here from the data

985
00:42:00,340 --> 00:42:02,840
is that the atomic
configuration at the interface

986
00:42:02,840 --> 00:42:04,280
really matters.

987
00:42:04,280 --> 00:42:06,710
So there must be something
going on in an atomic level,

988
00:42:06,710 --> 00:42:10,880
as well, that's determining
the barrier at that interface.

989
00:42:10,880 --> 00:42:12,300
In fact, there is.

990
00:42:12,300 --> 00:42:15,070
The dipole at the
interface between the metal

991
00:42:15,070 --> 00:42:17,770
and that last layer of
semiconductor, in other words,

992
00:42:17,770 --> 00:42:20,940
the distribution of charge
between those two sets of atoms

993
00:42:20,940 --> 00:42:23,610
that are meeting at an
interface determines, in part,

994
00:42:23,610 --> 00:42:24,930
the barrier height.

995
00:42:24,930 --> 00:42:27,340
We'll get to that in
a few slides, as well.

996
00:42:27,340 --> 00:42:32,800
So this is meant to shine light
on a much broader picture,

997
00:42:32,800 --> 00:42:34,260
which is to say
that there can be

998
00:42:34,260 --> 00:42:37,290
substantial deviations
from Schottky theory

999
00:42:37,290 --> 00:42:38,330
at the interface.

1000
00:42:38,330 --> 00:42:41,080
And some of the effects are
due to orientation-dependent

1001
00:42:41,080 --> 00:42:42,640
surface states.

1002
00:42:42,640 --> 00:42:46,355
It could be due to the
specific elemental nature

1003
00:42:46,355 --> 00:42:48,600
of the terminated
surface, as we just

1004
00:42:48,600 --> 00:42:52,500
discussed, whether you're
terminating at the silicon

1005
00:42:52,500 --> 00:42:54,200
plane or the carbon plane.

1006
00:42:54,200 --> 00:42:56,250
Orientation-dependence
means, gee,

1007
00:42:56,250 --> 00:42:58,430
if I have an
anisotropic crystal,

1008
00:42:58,430 --> 00:43:02,300
meaning a crystal that looks
a little bit different, if I

1009
00:43:02,300 --> 00:43:05,140
rotate it around in
different orientations,

1010
00:43:05,140 --> 00:43:08,160
if I make cuts in different
directions exposing

1011
00:43:08,160 --> 00:43:10,630
different planes, they're
going to be a different density

1012
00:43:10,630 --> 00:43:12,120
of atoms in that plane.

1013
00:43:12,120 --> 00:43:13,080
There's going to be
a different charge

1014
00:43:13,080 --> 00:43:13,960
distribution in that plane.

1015
00:43:13,960 --> 00:43:15,150
And hence, I would
expect there to be

1016
00:43:15,150 --> 00:43:17,540
some different interaction
between the semiconductor

1017
00:43:17,540 --> 00:43:20,490
and the metal at that interface.

1018
00:43:20,490 --> 00:43:24,410
And finally, interface
dipoles, this relates back up

1019
00:43:24,410 --> 00:43:26,090
to the first one in part.

1020
00:43:26,090 --> 00:43:28,440
This just goes to
say that, if I have

1021
00:43:28,440 --> 00:43:31,480
an ionic material or an element
there at that interface that

1022
00:43:31,480 --> 00:43:33,790
is grabbing charge,
I could result

1023
00:43:33,790 --> 00:43:38,680
in a small region
of negative charge

1024
00:43:38,680 --> 00:43:40,920
and positive charge
forming a dipole right

1025
00:43:40,920 --> 00:43:42,070
there at the interface.

1026
00:43:42,070 --> 00:43:45,430
And if I have a buildup
of charge, I have a field.

1027
00:43:45,430 --> 00:43:47,040
If I have a field,
I have a potential.

1028
00:43:47,040 --> 00:43:48,540
And if I have a
potential, I'm going

1029
00:43:48,540 --> 00:43:52,082
to be disrupting this
precise interface right here.

1030
00:43:52,082 --> 00:43:54,540
And we can, as well, vary the
density of interface states--

1031
00:43:54,540 --> 00:43:56,456
I got you, Ashley, I'll
be there in a second--

1032
00:43:56,456 --> 00:43:59,920
so we can, as well, vary
the density of defect states

1033
00:43:59,920 --> 00:44:02,830
here at this interface
that can trap charge.

1034
00:44:02,830 --> 00:44:06,040
And again, with the accumulation
of charge, comes a field.

1035
00:44:06,040 --> 00:44:09,650
With a field comes a
deviation of the potential.

1036
00:44:09,650 --> 00:44:13,030
So what we've done for
ourselves is, effectively,

1037
00:44:13,030 --> 00:44:17,910
using the Schottky model,
we've set up for ourselves

1038
00:44:17,910 --> 00:44:24,020
the energy band diagram of a
metal semiconductor contact

1039
00:44:24,020 --> 00:44:26,000
in its simplest case.

1040
00:44:26,000 --> 00:44:28,160
And what we're going to do
over the next few slides

1041
00:44:28,160 --> 00:44:30,340
is explore some of
the more complexity

1042
00:44:30,340 --> 00:44:33,770
behind what goes into a
semiconductor metal contact.

1043
00:44:33,770 --> 00:44:35,140
Ashley?

1044
00:44:35,140 --> 00:44:38,860
ASHLEY: So going back
that with the green line,

1045
00:44:38,860 --> 00:44:41,260
so these metals,
given some [INAUDIBLE]

1046
00:44:41,260 --> 00:44:45,335
up there on the left, would all
result in Schottky contacts?

1047
00:44:45,335 --> 00:44:46,610
Is that right?

1048
00:44:46,610 --> 00:44:48,090
Because all of
their work functions

1049
00:44:48,090 --> 00:44:51,902
are above the Fermi level?

1050
00:44:51,902 --> 00:44:52,860
TONIO BUONASSISI: Yeah.

1051
00:44:52,860 --> 00:44:54,529
ASHLEY: Assuming [INAUDIBLE].

1052
00:44:54,529 --> 00:44:56,070
TONIO BUONASSISI:
So yeah, it depends

1053
00:44:56,070 --> 00:45:02,540
on what is the minority carrier
in this particular junction

1054
00:45:02,540 --> 00:45:03,330
that's driving it.

1055
00:45:03,330 --> 00:45:06,080
Because this is intrinsic,
it's a little difficult to say,

1056
00:45:06,080 --> 00:45:06,820
actually.

1057
00:45:06,820 --> 00:45:10,210
But if it was clear
there was n- or p-type,

1058
00:45:10,210 --> 00:45:14,420
and the minority carrier
density at that interface

1059
00:45:14,420 --> 00:45:17,156
was determining the
current flow across it,

1060
00:45:17,156 --> 00:45:18,780
then we'd be able to
say with certainty

1061
00:45:18,780 --> 00:45:19,960
whether it was one
type or another.

1062
00:45:19,960 --> 00:45:20,435
ASHLEY: OK.

1063
00:45:20,435 --> 00:45:20,910
TONIO BUONASSISI: But, yeah.

1064
00:45:20,910 --> 00:45:22,166
ASHLEY: So if it were--

1065
00:45:22,166 --> 00:45:24,290
TONIO BUONASSISI: So in
this case right over here--

1066
00:45:24,290 --> 00:45:25,248
let's make it specific.

1067
00:45:25,248 --> 00:45:25,770
ASHLEY: Yes.

1068
00:45:25,770 --> 00:45:27,728
TONIO BUONASSISI: In this
case right over here,

1069
00:45:27,728 --> 00:45:30,890
let's make this an
n-type material.

1070
00:45:30,890 --> 00:45:34,500
So my Ef is now higher.

1071
00:45:34,500 --> 00:45:35,240
Ew.

1072
00:45:35,240 --> 00:45:36,760
OK.

1073
00:45:36,760 --> 00:45:38,660
That actually had
the opposite effect.

1074
00:45:38,660 --> 00:45:40,000
Let's exacerbate this a bit.

1075
00:45:40,000 --> 00:45:41,750
I'm actually going to
invert that and make

1076
00:45:41,750 --> 00:45:43,960
this a p-type material.

1077
00:45:43,960 --> 00:45:46,920
The reason is I want a big
delta in my Fermi energies

1078
00:45:46,920 --> 00:45:50,510
right here, if my vacuum
level is constant,

1079
00:45:50,510 --> 00:45:53,480
so to exacerbate that band
bending at the interface.

1080
00:45:53,480 --> 00:45:56,240
And so now, I'm going to get
a pretty drastic band bending.

1081
00:45:56,240 --> 00:45:58,650
It'll be even more extreme
than what I drew here.

1082
00:45:58,650 --> 00:46:05,254
But let's say it's fine.

1083
00:46:05,254 --> 00:46:09,860
So what we would have to do
is we would have to move this

1084
00:46:09,860 --> 00:46:13,549
down further, right?

1085
00:46:13,549 --> 00:46:14,090
Actually, no.

1086
00:46:14,090 --> 00:46:14,590
Sorry.

1087
00:46:14,590 --> 00:46:17,010
This wouldn't change, because
this is still the same.

1088
00:46:17,010 --> 00:46:18,468
What we've done
here is essentially

1089
00:46:18,468 --> 00:46:20,570
shifted everything up further.

1090
00:46:20,570 --> 00:46:25,540
So to be more precise, we'd have
to move this up, move this up.

1091
00:46:25,540 --> 00:46:28,120
And we'd have to move
this up, as well.

1092
00:46:28,120 --> 00:46:29,590
Again, p-type material.

1093
00:46:29,590 --> 00:46:30,620
Excellent.

1094
00:46:30,620 --> 00:46:34,090
And now, again, just
going through the logic

1095
00:46:34,090 --> 00:46:39,590
of the process here, we have to
make the vacuum levels match.

1096
00:46:39,590 --> 00:46:43,670
And there's going to be
an equal amount of charge

1097
00:46:43,670 --> 00:46:45,670
on either side of that junction.

1098
00:46:45,670 --> 00:46:47,900
Now, when there's a
charge, there's a field.

1099
00:46:47,900 --> 00:46:49,649
When there's a field,
there's a potential.

1100
00:46:49,649 --> 00:46:53,670
And the potential
will be manifested

1101
00:46:53,670 --> 00:46:56,535
in the bending of those bands
approaching the interface.

1102
00:46:56,535 --> 00:46:58,660
The bands will bend a lot
more in the semiconductor

1103
00:46:58,660 --> 00:47:01,516
than they will in the metal,
because the free charge

1104
00:47:01,516 --> 00:47:02,890
density here in
the semiconductor

1105
00:47:02,890 --> 00:47:04,870
is a lot less than
it is in the metal.

1106
00:47:04,870 --> 00:47:08,710
And going back to this
equation right here,

1107
00:47:08,710 --> 00:47:12,660
the width of the
region of bent bands

1108
00:47:12,660 --> 00:47:16,040
is going to depend inversely
on the dopant density

1109
00:47:16,040 --> 00:47:18,500
or the amount of
charge in that region.

1110
00:47:18,500 --> 00:47:21,170
So this is to say
that we'll have

1111
00:47:21,170 --> 00:47:23,670
a much more extreme bending
here in the semiconductor

1112
00:47:23,670 --> 00:47:25,092
than we will in the metal.

1113
00:47:25,092 --> 00:47:27,462
Virtually imperceptible
there in the metal.

1114
00:47:27,462 --> 00:47:29,300
And we drop this down, as well.

1115
00:47:29,300 --> 00:47:31,630
And we drop this down, as well.

1116
00:47:31,630 --> 00:47:33,890
So some really
interesting things

1117
00:47:33,890 --> 00:47:38,420
have happened right here to the
position of the Fermi energy

1118
00:47:38,420 --> 00:47:40,280
in the band gap, right?

1119
00:47:40,280 --> 00:47:42,970
Over here, we have
a p-type material.

1120
00:47:42,970 --> 00:47:46,030
And over here, at
the surface layer,

1121
00:47:46,030 --> 00:47:48,443
it almost looks like this
material is going n-type.

1122
00:47:48,443 --> 00:47:49,039
ASHLEY: Right.

1123
00:47:49,039 --> 00:47:51,330
TONIO BUONASSISI: Let me
pause right here, because this

1124
00:47:51,330 --> 00:47:53,038
is just utterly
fascinating, and tell you

1125
00:47:53,038 --> 00:47:54,982
the story about indium nitride.

1126
00:47:54,982 --> 00:47:56,690
It's a common material
that many people--

1127
00:47:56,690 --> 00:47:59,420
I see a few knowing nods
here in the audience.

1128
00:47:59,420 --> 00:48:02,140
So indium nitride was a
semiconducting material

1129
00:48:02,140 --> 00:48:04,200
that people really didn't
have a good handle on

1130
00:48:04,200 --> 00:48:05,590
for a variety of reasons.

1131
00:48:05,590 --> 00:48:08,220
And there was a
renewed interest in it,

1132
00:48:08,220 --> 00:48:11,060
because people finally started
to understand why the band

1133
00:48:11,060 --> 00:48:12,790
gap was the way it was.

1134
00:48:12,790 --> 00:48:15,440
And many groups
started studying it.

1135
00:48:15,440 --> 00:48:18,170
And they also reporting
this n-type indium nitride

1136
00:48:18,170 --> 00:48:20,920
behavior, very highly
n-type material,

1137
00:48:20,920 --> 00:48:22,550
highly doped n-type material.

1138
00:48:22,550 --> 00:48:25,170
It turned out that there
was this band bending here

1139
00:48:25,170 --> 00:48:27,230
at the surface.

1140
00:48:27,230 --> 00:48:30,790
And they were measuring a
high electron concentration

1141
00:48:30,790 --> 00:48:33,540
at the surface, because
of the bending of bands.

1142
00:48:33,540 --> 00:48:36,390
It was in reality the intrinsic
material deep in the bulk

1143
00:48:36,390 --> 00:48:37,367
was p-type.

1144
00:48:37,367 --> 00:48:39,450
But they just couldn't see
it because their probes

1145
00:48:39,450 --> 00:48:42,610
were touching the
semiconductor at the surface.

1146
00:48:42,610 --> 00:48:47,220
It took a certain scientist
by the name of Becca Jones

1147
00:48:47,220 --> 00:48:50,050
in Berkeley, Lawrence Berkeley
Lab, I believe, at the time.

1148
00:48:50,050 --> 00:48:53,030
She immersed the sample
into a liquid solution

1149
00:48:53,030 --> 00:48:55,420
of an acid-- I believe it
was hydrofluoric acid--

1150
00:48:55,420 --> 00:48:58,010
and managed to relieve
the pinning of the bands

1151
00:48:58,010 --> 00:48:58,860
at the surface.

1152
00:48:58,860 --> 00:49:00,151
And so they returned to normal.

1153
00:49:00,151 --> 00:49:03,030
And she was able to probe
using a semiconductor liquid

1154
00:49:03,030 --> 00:49:06,940
junction, the true
conductivity type

1155
00:49:06,940 --> 00:49:08,670
of the bulk of the material.

1156
00:49:08,670 --> 00:49:12,675
Becca Jones then went on to
produce what is now the-- oh,

1157
00:49:12,675 --> 00:49:14,050
she was on the
team that produced

1158
00:49:14,050 --> 00:49:17,050
one of the highest
efficiency cells

1159
00:49:17,050 --> 00:49:19,863
on the efficiency
versus time plots.

1160
00:49:19,863 --> 00:49:23,410
So she's doing well for
herself there in the PV world.

1161
00:49:23,410 --> 00:49:24,115
Yeah.

1162
00:49:24,115 --> 00:49:25,450
AUDIENCE: [INAUDIBLE]?

1163
00:49:25,450 --> 00:49:25,780
TONIO BUONASSISI: Sorry.

1164
00:49:25,780 --> 00:49:26,120
Sorry.

1165
00:49:26,120 --> 00:49:27,260
So let me get back to that.

1166
00:49:27,260 --> 00:49:30,670
What would happen if you
were an electron right here

1167
00:49:30,670 --> 00:49:32,790
and you were approaching
this junction?

1168
00:49:32,790 --> 00:49:36,820
This is E. This is the
energy of electrons.

1169
00:49:36,820 --> 00:49:39,350
So the electron will
seek to minimize,

1170
00:49:39,350 --> 00:49:41,721
it will seek to encounter
a lower energy state.

1171
00:49:41,721 --> 00:49:43,970
So if you're approaching
this junction as an electron,

1172
00:49:43,970 --> 00:49:44,720
what would happen?

1173
00:49:47,034 --> 00:49:47,950
AUDIENCE: [INAUDIBLE].

1174
00:49:47,950 --> 00:49:48,250
TONIO BUONASSISI: Zoom.

1175
00:49:48,250 --> 00:49:48,790
Yep.

1176
00:49:48,790 --> 00:49:49,520
So it's go down.

1177
00:49:49,520 --> 00:49:51,794
And there wouldn't be any
barrier for it to do so.

1178
00:49:51,794 --> 00:49:52,460
AUDIENCE: Right.

1179
00:49:52,460 --> 00:49:52,800
TONIO BUONASSISI:
It would just go,

1180
00:49:52,800 --> 00:49:54,258
[MAKES WHOOSH SOUND]
straight down.

1181
00:49:54,258 --> 00:49:56,140
Now, let's inverse the
situation and imagine,

1182
00:49:56,140 --> 00:49:59,190
what if you're a
hole right here?

1183
00:49:59,190 --> 00:50:00,940
What now?

1184
00:50:00,940 --> 00:50:02,810
If you think of electrons
as bowling balls

1185
00:50:02,810 --> 00:50:04,550
that want to roll
down hills, holes

1186
00:50:04,550 --> 00:50:06,350
would be the opposite,
which would be like balloons

1187
00:50:06,350 --> 00:50:07,420
that want to rise, right?

1188
00:50:07,420 --> 00:50:07,770
AUDIENCE: Right.

1189
00:50:07,770 --> 00:50:09,960
TONIO BUONASSISI: And so this is
actually a barrier for a hole.

1190
00:50:09,960 --> 00:50:10,640
AUDIENCE: Right.

1191
00:50:10,640 --> 00:50:12,848
TONIO BUONASSISI: You need
a certain amount of energy

1192
00:50:12,848 --> 00:50:14,550
to get the hole across.

1193
00:50:14,550 --> 00:50:17,490
And so the hole naturally will
be repelled by this junction.

1194
00:50:17,490 --> 00:50:18,906
The hole will want
to go this way.

1195
00:50:18,906 --> 00:50:21,049
The electron will
want to go that way.

1196
00:50:21,049 --> 00:50:21,590
AUDIENCE: OK.

1197
00:50:21,590 --> 00:50:25,106
So this is Schottky for
electrons, but Ohmic-- sorry,

1198
00:50:25,106 --> 00:50:26,980
Schottky for holes and
Ohmic for [INAUDIBLE].

1199
00:50:26,980 --> 00:50:27,630
TONIO BUONASSISI:
Let's imagine we

1200
00:50:27,630 --> 00:50:30,451
have a device that is driven
by the minority carrier flux

1201
00:50:30,451 --> 00:50:31,700
here at this interface, right?

1202
00:50:31,700 --> 00:50:32,310
AUDIENCE: Right.

1203
00:50:32,310 --> 00:50:33,580
TONIO BUONASSISI: And so
what is the minority carrier

1204
00:50:33,580 --> 00:50:35,470
in this particular case?

1205
00:50:35,470 --> 00:50:38,610
AUDIENCE: So the left is p-type,
and so the minority carrier

1206
00:50:38,610 --> 00:50:39,550
would be electrons?

1207
00:50:39,550 --> 00:50:40,110
TONIO BUONASSISI: Yep.

1208
00:50:40,110 --> 00:50:40,750
Exactly.

1209
00:50:40,750 --> 00:50:44,793
So what would you expect
this contact to behave like?

1210
00:50:44,793 --> 00:50:48,982
AUDIENCE: So it'd be Schottky
for-- no, Ohmic for electrons.

1211
00:50:48,982 --> 00:50:50,690
TONIO BUONASSISI: So
this one right here,

1212
00:50:50,690 --> 00:50:52,940
the electrons would be able
to flow over quite easily.

1213
00:50:52,940 --> 00:50:56,490
So in the illuminated case
where the current would

1214
00:50:56,490 --> 00:50:58,800
be driven by the
minority carrier flux,

1215
00:50:58,800 --> 00:51:01,915
you would expect to see a
behavior without a barrier.

1216
00:51:01,915 --> 00:51:02,456
AUDIENCE: OK.

1217
00:51:02,456 --> 00:51:03,524
So Ohmic for--

1218
00:51:03,524 --> 00:51:04,440
TONIO BUONASSISI: Yep.

1219
00:51:04,440 --> 00:51:05,240
AUDIENCE: OK.

1220
00:51:05,240 --> 00:51:08,750
TONIO BUONASSISI: Now, it
can become a little bit more

1221
00:51:08,750 --> 00:51:11,742
complicated, but let's
leave that right there.

1222
00:51:11,742 --> 00:51:12,408
AUDIENCE: Right.

1223
00:51:12,408 --> 00:51:13,336
OK.

1224
00:51:13,336 --> 00:51:16,520
AUDIENCE: If you have a
Schottkys barrier like

1225
00:51:16,520 --> 00:51:19,495
that, it's just extracting
electrons or holes,

1226
00:51:19,495 --> 00:51:22,800
then is the relevant quantity
on the left a quasi-Fermi

1227
00:51:22,800 --> 00:51:24,680
energy for electrons or holes?

1228
00:51:24,680 --> 00:51:26,822
Or is it it's always the
overall Fermi energy?

1229
00:51:26,822 --> 00:51:27,780
TONIO BUONASSISI: Yeah.

1230
00:51:27,780 --> 00:51:30,380
So you raise an
interesting point,

1231
00:51:30,380 --> 00:51:33,030
which is the question
of quasi-Fermi energy.

1232
00:51:33,030 --> 00:51:37,110
So if you illuminate your
device or you forward-bias it,

1233
00:51:37,110 --> 00:51:37,920
what happens?

1234
00:51:37,920 --> 00:51:39,540
Now, you start
injecting carriers.

1235
00:51:39,540 --> 00:51:41,570
And as we discussed
during class the time,

1236
00:51:41,570 --> 00:51:44,020
the total carrier
density is going

1237
00:51:44,020 --> 00:51:47,600
to be equal to, let's call
it, the intrinsic carrier

1238
00:51:47,600 --> 00:51:51,070
concentration plus
the dopant density,

1239
00:51:51,070 --> 00:51:53,250
if this n would be
the density of donors,

1240
00:51:53,250 --> 00:51:58,080
plus my delta n, which is the
injected carrier concentration.

1241
00:51:58,080 --> 00:52:02,290
So in this particular case--
let's parse this through-- this

1242
00:52:02,290 --> 00:52:04,150
is going to be small,
so we can ignore that.

1243
00:52:04,150 --> 00:52:05,515
Well, actually, no.

1244
00:52:05,515 --> 00:52:07,420
For the electrons
in this case, this

1245
00:52:07,420 --> 00:52:09,530
will actually be quite big.

1246
00:52:09,530 --> 00:52:13,350
This is going to be
smaller, rather negligible.

1247
00:52:13,350 --> 00:52:18,250
And let's think this
through carefully here.

1248
00:52:18,250 --> 00:52:21,890
So under equilibrium conditions,
under equilibrium conditions--

1249
00:52:21,890 --> 00:52:25,110
let's describe this
as an n0, briefly.

1250
00:52:25,110 --> 00:52:26,870
So under equilibrium
conditions, my n0

1251
00:52:26,870 --> 00:52:29,620
is going to be rather
small because, if I

1252
00:52:29,620 --> 00:52:32,490
have a high dopant density
of p-type material,

1253
00:52:32,490 --> 00:52:35,740
that means that my
electron and my n0

1254
00:52:35,740 --> 00:52:39,610
is going to be equal to
ni squared divided by p0.

1255
00:52:39,610 --> 00:52:43,300
And if p0 is in the order
of, say, 10 to the 16,

1256
00:52:43,300 --> 00:52:45,240
ni would be nearer
to 10 to the 10.

1257
00:52:45,240 --> 00:52:49,000
We'd have something in the
order of 10 to the 4 for n0.

1258
00:52:49,000 --> 00:52:50,700
That would be a
rather small number.

1259
00:52:50,700 --> 00:52:52,780
Now, if I inject light
on it, I could have, say,

1260
00:52:52,780 --> 00:52:55,020
10 to the 14 electrons
being created.

1261
00:52:55,020 --> 00:52:57,090
So this drowns out
everything else.

1262
00:52:57,090 --> 00:53:00,180
So now, my electron
concentration

1263
00:53:00,180 --> 00:53:03,000
is going to be driven
by the free carrier

1264
00:53:03,000 --> 00:53:06,180
concentration, the photo-excited
carrier concentration.

1265
00:53:06,180 --> 00:53:08,580
Now, on the p, on
the other hand,

1266
00:53:08,580 --> 00:53:12,800
this is going to be
equal to p0 plus delta p.

1267
00:53:12,800 --> 00:53:15,710
So let me drag this over here.

1268
00:53:15,710 --> 00:53:18,370
The delta p is still on the
same order as the delta n,

1269
00:53:18,370 --> 00:53:20,930
let's say, 10 to the 13,
somewhere in that range.

1270
00:53:20,930 --> 00:53:26,130
But my p0 is equal to the dopant
density, acceptor density,

1271
00:53:26,130 --> 00:53:28,420
which could be on the
order of 10 to the 16.

1272
00:53:28,420 --> 00:53:32,620
So this will be somewhere in
the range of, say, 10 to the 16,

1273
00:53:32,620 --> 00:53:35,450
driven by the acceptor
concentration.

1274
00:53:35,450 --> 00:53:38,560
And so now, I have a
situation in which I have 10

1275
00:53:38,560 --> 00:53:42,740
to the 13 electrons and
maybe 10 to the 16 holes.

1276
00:53:42,740 --> 00:53:44,205
And if I multiply
the two together,

1277
00:53:44,205 --> 00:53:46,330
I'm not getting the intrinsic
carrier concentration

1278
00:53:46,330 --> 00:53:49,630
squared, because I'm not
under dark conditions.

1279
00:53:49,630 --> 00:53:51,310
I'm not under
equilibrium conditions.

1280
00:53:51,310 --> 00:53:53,070
I'm in a steady state condition.

1281
00:53:53,070 --> 00:53:53,900
I'm shining light.

1282
00:53:53,900 --> 00:53:54,982
I'm exciting carriers.

1283
00:53:54,982 --> 00:53:56,690
They're recombining,
but at steady state.

1284
00:53:56,690 --> 00:53:59,290
I'm shining light,
generating the carrier again.

1285
00:53:59,290 --> 00:54:02,180
And we reach this
quasi-equilibrium.

1286
00:54:02,180 --> 00:54:04,440
And so what we have,
at the end of the day,

1287
00:54:04,440 --> 00:54:06,690
is a separation of
the Fermi energy

1288
00:54:06,690 --> 00:54:08,130
between electrons and holes.

1289
00:54:08,130 --> 00:54:09,600
The Fermi energy
for holes is going

1290
00:54:09,600 --> 00:54:10,750
to be relatively constant.

1291
00:54:10,750 --> 00:54:12,541
But the electrons for
the minority carrier,

1292
00:54:12,541 --> 00:54:15,010
you'll have a rise of
that quasi-Fermi energy,

1293
00:54:15,010 --> 00:54:18,120
because now you're filling in
more of your electrons up here.

1294
00:54:18,120 --> 00:54:20,230
And so the Fermi
energy, as defined

1295
00:54:20,230 --> 00:54:25,860
as the energy state that has
a 50% occupancy probability,

1296
00:54:25,860 --> 00:54:28,730
is now at a different energy
for electrons or for holes.

1297
00:54:28,730 --> 00:54:29,510
Awesome.

1298
00:54:29,510 --> 00:54:31,280
It's really one of
those things that

1299
00:54:31,280 --> 00:54:34,592
takes a while to really
wrap your mind around.

1300
00:54:34,592 --> 00:54:36,970
And so now, we might
have a situation

1301
00:54:36,970 --> 00:54:42,430
which my electron quasi-Fermi
energy is somewhere up here.

1302
00:54:45,630 --> 00:54:47,200
Let's make it a
little bit-- I don't

1303
00:54:47,200 --> 00:54:50,250
want to get into what's called
a two-dimension electron gas.

1304
00:54:50,250 --> 00:54:52,510
I'll draw it
somewhere down here.

1305
00:54:52,510 --> 00:54:56,090
So now, I have
this distance here

1306
00:54:56,090 --> 00:54:58,310
is defined by the minority
current diffusion length,

1307
00:54:58,310 --> 00:55:00,924
typically, from the junction.

1308
00:55:00,924 --> 00:55:02,667
AUDIENCE: And as
the electrons drop

1309
00:55:02,667 --> 00:55:05,904
from that quasi-Fermi
energy to the Fermi energy

1310
00:55:05,904 --> 00:55:07,404
of the metal, what happens?

1311
00:55:07,404 --> 00:55:09,070
TONIO BUONASSISI: So
the important thing

1312
00:55:09,070 --> 00:55:13,640
to think about is current flow
can be reduced to electrons

1313
00:55:13,640 --> 00:55:15,960
within the conduction
band, discrete particles

1314
00:55:15,960 --> 00:55:18,380
that are seeking to
minimize their free energy.

1315
00:55:18,380 --> 00:55:20,680
The voltage or the
potential across

1316
00:55:20,680 --> 00:55:21,994
is dictated by the ensemble.

1317
00:55:21,994 --> 00:55:24,118
And that's where the Fermi
energies come into play.

1318
00:55:28,700 --> 00:55:31,990
So in that particular case,
it might even cause the Fermi

1319
00:55:31,990 --> 00:55:33,930
energy over here to rise a bit.

1320
00:55:33,930 --> 00:55:36,480
And then you have
a potential across.

1321
00:55:36,480 --> 00:55:38,830
If you had a wire connecting
this to an external load,

1322
00:55:38,830 --> 00:55:40,350
you'd actually be
able to power it.

1323
00:55:40,350 --> 00:55:41,850
The electrons would
be photo-excited

1324
00:55:41,850 --> 00:55:46,180
in here, would be driven
over into the metal,

1325
00:55:46,180 --> 00:55:47,744
because of the built-in field.

1326
00:55:47,744 --> 00:55:49,910
And then they'd be able to
power an external circuit

1327
00:55:49,910 --> 00:55:51,480
and come back into the back.

1328
00:55:51,480 --> 00:55:52,580
So you're beginning
to see something that

1329
00:55:52,580 --> 00:55:53,871
looks a lot like a PN junction.

1330
00:55:59,310 --> 00:56:04,260
And in this particular case,
the dark current for electrons

1331
00:56:04,260 --> 00:56:07,042
would be dictated by their flow
back into the semiconductor.

1332
00:56:10,140 --> 00:56:14,982
So yes, in this
particular case here--

1333
00:56:14,982 --> 00:56:16,940
let me correct myself--
in this particular case

1334
00:56:16,940 --> 00:56:18,540
here, for the
electron, you would

1335
00:56:18,540 --> 00:56:21,410
have a barrier for the
diffusion current, which

1336
00:56:21,410 --> 00:56:27,550
would be driven from the metal
into the base right here.

1337
00:56:27,550 --> 00:56:30,040
And so the diffusion
current of the electrons

1338
00:56:30,040 --> 00:56:31,780
would be facing
this barrier here.

1339
00:56:31,780 --> 00:56:35,540
And this would be
Schottky-like in nature.

1340
00:56:35,540 --> 00:56:37,530
If you illuminate it,
your illumination current

1341
00:56:37,530 --> 00:56:41,460
would-- so let me back
up one step right here.

1342
00:56:41,460 --> 00:56:47,730
So in the dark, a completely
dark sample right here,

1343
00:56:47,730 --> 00:56:49,157
that I-V characteristic
is driven

1344
00:56:49,157 --> 00:56:51,240
by the diffusion current
of electrons to the metal

1345
00:56:51,240 --> 00:56:53,470
and to the semiconductor.

1346
00:56:53,470 --> 00:56:56,370
And that is, in this
case, with the barrier.

1347
00:56:56,370 --> 00:56:57,870
And as you begin
forward-biasing it,

1348
00:56:57,870 --> 00:57:00,180
you would have more and
more carriers coming across.

1349
00:57:00,180 --> 00:57:03,370
And that would drive the
diffusion current forward.

1350
00:57:03,370 --> 00:57:06,284
The illumination current would
be driving the opposite way.

1351
00:57:06,284 --> 00:57:08,200
And so, if you illuminated
this entire system,

1352
00:57:08,200 --> 00:57:09,894
your blue curve
would shift down.

1353
00:57:15,400 --> 00:57:17,750
Let's say, if you're
photo-exciting carriers

1354
00:57:17,750 --> 00:57:20,150
within the base right
here, they can very easily

1355
00:57:20,150 --> 00:57:21,380
reach the metal.

1356
00:57:21,380 --> 00:57:25,110
But if you're putting contacts
and probing the total current

1357
00:57:25,110 --> 00:57:28,030
flowing through your
system, in the dark,

1358
00:57:28,030 --> 00:57:30,230
you would be measuring
the diffusion current

1359
00:57:30,230 --> 00:57:33,410
from the metal into
the semiconductor.

1360
00:57:33,410 --> 00:57:36,140
There's some positive
electrons over here.

1361
00:57:36,140 --> 00:57:39,140
And so the electrons will
be driven from the metal

1362
00:57:39,140 --> 00:57:40,360
into the semiconductor.

1363
00:57:40,360 --> 00:57:42,210
And they would be
facing this barrier.

1364
00:57:42,210 --> 00:57:44,600
And hence, you would see
a Schottky-like behavior.

1365
00:57:44,600 --> 00:57:45,141
AUDIENCE: OK.

1366
00:57:45,141 --> 00:57:48,438
So in the dark, it would be
Schottky-like for electrons.

1367
00:57:48,438 --> 00:57:49,840
And in the [INAUDIBLE]--

1368
00:57:49,840 --> 00:57:51,340
TONIO BUONASSISI:
So you think of it

1369
00:57:51,340 --> 00:57:52,782
in terms of total current.

1370
00:57:52,782 --> 00:57:53,990
Yes, you're absolutely right.

1371
00:57:53,990 --> 00:57:54,410
AUDIENCE: Right.

1372
00:57:54,410 --> 00:57:56,910
TONIO BUONASSISI: And the holes
would be equal and opposite.

1373
00:57:56,910 --> 00:57:58,470
Then, when you
illuminate it, current

1374
00:57:58,470 --> 00:58:01,710
would start flowing
the other direction.

1375
00:58:01,710 --> 00:58:04,790
So in the dark, the
current is flowing, say,

1376
00:58:04,790 --> 00:58:07,295
in the positive direction.

1377
00:58:07,295 --> 00:58:08,920
And let's define the
positive direction

1378
00:58:08,920 --> 00:58:12,420
as electrons flowing
from here to here,

1379
00:58:12,420 --> 00:58:16,540
which means that positive flow
is going from left to right.

1380
00:58:16,540 --> 00:58:17,790
So that makes sense.

1381
00:58:17,790 --> 00:58:20,540
And when we illuminate it,
now, instead of the electrons

1382
00:58:20,540 --> 00:58:22,490
flowing from the
right into the left,

1383
00:58:22,490 --> 00:58:25,930
we have the illumination current
driving electrons from the left

1384
00:58:25,930 --> 00:58:27,210
and to the right.

1385
00:58:27,210 --> 00:58:29,220
So the direction of
current is reversed,

1386
00:58:29,220 --> 00:58:30,750
when we illuminate this.

1387
00:58:30,750 --> 00:58:32,330
And that's the same
thing as saying,

1388
00:58:32,330 --> 00:58:35,950
I'm going to shift this
down under illumination

1389
00:58:35,950 --> 00:58:40,170
and wind up with something that
behaves very similar to a PN

1390
00:58:40,170 --> 00:58:42,296
junction.

1391
00:58:42,296 --> 00:58:42,920
AUDIENCE: Wait.

1392
00:58:42,920 --> 00:58:47,338
I thought we said that, under
illumination conditions,

1393
00:58:47,338 --> 00:58:49,306
this situation is
Ohmic for electrons.

1394
00:58:49,306 --> 00:58:50,290
Or is that not true?

1395
00:58:50,290 --> 00:58:51,290
TONIO BUONASSISI: Sorry.

1396
00:58:51,290 --> 00:58:53,350
The electrons have no
barrier to travel over here

1397
00:58:53,350 --> 00:58:54,385
for the illumination current.

1398
00:58:54,385 --> 00:58:54,760
AUDIENCE: Right.

1399
00:58:54,760 --> 00:58:54,980
Oh, so this--

1400
00:58:54,980 --> 00:58:55,940
TONIO BUONASSISI: So the
illumination current's

1401
00:58:55,940 --> 00:58:57,020
going to be constant throughout.

1402
00:58:57,020 --> 00:58:58,353
AUDIENCE: So it's-- [INAUDIBLE].

1403
00:58:58,353 --> 00:59:01,250
TONIO BUONASSISI: But for
the diffusion current--

1404
00:59:01,250 --> 00:59:04,004
the net current behavior, the
net I-V characteristic of this

1405
00:59:04,004 --> 00:59:04,920
will be Schottky-like.

1406
00:59:04,920 --> 00:59:05,909
AUDIENCE: OK.

1407
00:59:05,909 --> 00:59:07,700
TONIO BUONASSISI: But
the diffusion current

1408
00:59:07,700 --> 00:59:11,080
is the one that's driving
the current in that case.

1409
00:59:11,080 --> 00:59:14,520
And the illumination current
is being driven by the minority

1410
00:59:14,520 --> 00:59:18,210
carrier flow right here at the
edge of the metal semiconductor

1411
00:59:18,210 --> 00:59:19,364
contact.

1412
00:59:19,364 --> 00:59:19,906
AUDIENCE: OK.

1413
00:59:19,906 --> 00:59:21,822
TONIO BUONASSISI: Joe,
why don't we flag this?

1414
00:59:21,822 --> 00:59:24,250
I sense this is going to
be an item for discussion

1415
00:59:24,250 --> 00:59:27,167
during the recitation session.

1416
00:59:27,167 --> 00:59:28,960
JOE: Sorry. [INAUDIBLE]?

1417
00:59:28,960 --> 00:59:31,440
TONIO BUONASSISI: Just the
metal semiconductor interface,

1418
00:59:31,440 --> 00:59:33,540
I think, we're going to
have to go through it

1419
00:59:33,540 --> 00:59:34,440
from scratch again.

1420
00:59:34,440 --> 00:59:36,790
This is very, very
similar to a PN junction.

1421
00:59:36,790 --> 00:59:37,380
AUDIENCE: Yep.

1422
00:59:37,380 --> 00:59:37,860
TONIO BUONASSISI:
So I think where

1423
00:59:37,860 --> 00:59:40,290
folks are getting tripped
up is thinking about--

1424
00:59:40,290 --> 00:59:42,050
and even I sometimes
get tripped up--

1425
00:59:42,050 --> 00:59:44,540
is thinking about what happens
to the illuminated current

1426
00:59:44,540 --> 00:59:47,980
and what happens to the drift
current and the diffusion

1427
00:59:47,980 --> 00:59:48,803
current.

1428
00:59:48,803 --> 00:59:49,790
AUDIENCE: Right.

1429
00:59:49,790 --> 00:59:50,780
TONIO BUONASSISI: So
it's all very similar.

1430
00:59:50,780 --> 00:59:51,770
It's all stuff
we've seen before.

1431
00:59:51,770 --> 00:59:53,630
We just have to think
through it step by step

1432
00:59:53,630 --> 00:59:55,796
and make sure we don't get
tripped up along the way.

1433
00:59:55,796 --> 00:59:56,910
AUDIENCE: Yep.

1434
00:59:56,910 --> 00:59:57,785
TONIO BUONASSISI: OK.

1435
00:59:59,810 --> 01:00:02,590
So let's talk about some
of the non-idealities

1436
01:00:02,590 --> 01:00:06,390
because this is where
the fun stuff lies.

1437
01:00:06,390 --> 01:00:10,670
When we have a metal
on a semiconductor,

1438
01:00:10,670 --> 01:00:13,410
a lot goes into it.

1439
01:00:13,410 --> 01:00:16,160
For example, if we have
interface states or surface

1440
01:00:16,160 --> 01:00:20,040
states, we can form band
bending there at that interface.

1441
01:00:20,040 --> 01:00:21,500
Let me demonstrate how.

1442
01:00:21,500 --> 01:00:24,150
This is an article that everyone
should have read already.

1443
01:00:24,150 --> 01:00:26,270
It was part of the
assigned reading today.

1444
01:00:26,270 --> 01:00:28,994
And if my past
experience is a guide,

1445
01:00:28,994 --> 01:00:30,410
there is a fixed
percentage of you

1446
01:00:30,410 --> 01:00:31,951
who haven't seen it
yet, so I'm going

1447
01:00:31,951 --> 01:00:34,750
to pass around the article
that you should have read.

1448
01:00:34,750 --> 01:00:38,710
So this is something-- feel
free to skim through it,

1449
01:00:38,710 --> 01:00:42,390
gain an appreciation for,
really, the spectacular nature

1450
01:00:42,390 --> 01:00:43,230
of that article.

1451
01:00:43,230 --> 01:00:44,896
And then pass it
around your colleagues.

1452
01:00:44,896 --> 01:00:48,350
Make sure everybody
gets to see them.

1453
01:00:48,350 --> 01:00:51,850
On the left-hand side, we have
an example of a semiconductor,

1454
01:00:51,850 --> 01:00:53,420
let's say, surface.

1455
01:00:53,420 --> 01:00:56,240
The surface is represented
by this really large band gap

1456
01:00:56,240 --> 01:00:58,720
material right there, that
box, that rectilinear box,

1457
01:00:58,720 --> 01:01:00,550
very thin and skinny.

1458
01:01:00,550 --> 01:01:04,560
It's thin because the
surface layer is very thin.

1459
01:01:04,560 --> 01:01:07,102
It's maybe a few nanometers,
typically, in a semiconductor.

1460
01:01:07,102 --> 01:01:08,560
And it's very large
and band gapped

1461
01:01:08,560 --> 01:01:11,800
because we formed a
semiconductor and oxide,

1462
01:01:11,800 --> 01:01:14,890
let's say, a silicon oxide, a
silica layer at that surface.

1463
01:01:14,890 --> 01:01:17,190
So if we have pure silicon
and we expose it to air,

1464
01:01:17,190 --> 01:01:21,070
the silicon will react with air
and form a very thin SiO2 layer

1465
01:01:21,070 --> 01:01:22,447
or silicon oxide layer.

1466
01:01:22,447 --> 01:01:23,780
And that's what this represents.

1467
01:01:23,780 --> 01:01:26,790
This is a very large band gap
material, and it's very thin.

1468
01:01:26,790 --> 01:01:28,800
And so if there are
no interface states,

1469
01:01:28,800 --> 01:01:31,530
if we have an ideal material,
we have our semiconductor

1470
01:01:31,530 --> 01:01:34,390
over here and our
surface layer there.

1471
01:01:34,390 --> 01:01:36,180
And at that interface, boom.

1472
01:01:36,180 --> 01:01:40,290
You just have a discontinuity
in the conduction bands.

1473
01:01:40,290 --> 01:01:43,350
And of course, the vacuum
levels are matching up.

1474
01:01:43,350 --> 01:01:47,580
But in the case where
you have surface states

1475
01:01:47,580 --> 01:01:50,050
at that interface, what happens?

1476
01:01:50,050 --> 01:01:51,490
Well now, those
surface states can

1477
01:01:51,490 --> 01:01:56,480
be filled by electrons in
the neighboring region.

1478
01:01:56,480 --> 01:01:59,630
And as a result, you'll
have a natural band

1479
01:01:59,630 --> 01:02:01,360
bending at that interface.

1480
01:02:01,360 --> 01:02:03,267
So as you begin filling
in those states,

1481
01:02:03,267 --> 01:02:05,100
another way to think
about it is that you're

1482
01:02:05,100 --> 01:02:07,440
pulling the Fermi
energy toward mid-gap,

1483
01:02:07,440 --> 01:02:09,990
because now, as the
electrons are depleted

1484
01:02:09,990 --> 01:02:13,770
from that near-surface region,
as they fall into those surface

1485
01:02:13,770 --> 01:02:16,979
states, the Fermi energy is
going to shift towards mid-gap.

1486
01:02:16,979 --> 01:02:19,270
And if you indeed look at
the semiconductor, right here

1487
01:02:19,270 --> 01:02:21,620
at the surface, you'll see
the Fermi energy is almost

1488
01:02:21,620 --> 01:02:23,450
at the mid-gap.

1489
01:02:23,450 --> 01:02:26,060
And so this is a very
curious phenomenon that

1490
01:02:26,060 --> 01:02:27,740
occurs in many semiconductors.

1491
01:02:27,740 --> 01:02:29,840
There's a natural band
bend toward the surface,

1492
01:02:29,840 --> 01:02:32,010
depending on the
surface chemistry.

1493
01:02:32,010 --> 01:02:36,424
If you have a semi-conducting
oxide forming at the surface,

1494
01:02:36,424 --> 01:02:37,840
or if you happen
to be exposing it

1495
01:02:37,840 --> 01:02:41,080
to a sulfur based gas
right after you grow it,

1496
01:02:41,080 --> 01:02:44,170
and you form some
sulfide at the surface,

1497
01:02:44,170 --> 01:02:46,827
the surface chemistry is
very important for dictating

1498
01:02:46,827 --> 01:02:48,160
the band bending at the surface.

1499
01:02:48,160 --> 01:02:49,429
Why is that important?

1500
01:02:49,429 --> 01:02:51,720
That's important because,
when you go ahead and deposit

1501
01:02:51,720 --> 01:02:53,770
a metal on top of
it, now, you not only

1502
01:02:53,770 --> 01:02:56,730
have to contend with what
the semiconductor is doing,

1503
01:02:56,730 --> 01:03:00,700
but what that semiconductor
surface layer is doing.

1504
01:03:00,700 --> 01:03:03,970
Now, the situation is
really becoming complex.

1505
01:03:03,970 --> 01:03:07,900
So this is meant to
represent right here what

1506
01:03:07,900 --> 01:03:10,760
happens when you have
a metal-- in this case,

1507
01:03:10,760 --> 01:03:14,550
a semi-conducting oxide
semiconductor interface,

1508
01:03:14,550 --> 01:03:18,870
also known as a MOS, metal oxide
semiconductor-- what happens?

1509
01:03:18,870 --> 01:03:22,680
Well, in an ideal case,
you would have the bands

1510
01:03:22,680 --> 01:03:24,850
bending a certain way.

1511
01:03:24,850 --> 01:03:28,000
And then, with the metal
oxide layer in between,

1512
01:03:28,000 --> 01:03:29,960
you can have a definite shift.

1513
01:03:29,960 --> 01:03:32,550
So what this is meant
to represent right here

1514
01:03:32,550 --> 01:03:35,120
is a metal with a
work function that

1515
01:03:35,120 --> 01:03:37,520
is larger than the
electron affinity,

1516
01:03:37,520 --> 01:03:39,730
and a work function that is
smaller than the electron

1517
01:03:39,730 --> 01:03:40,620
affinity.

1518
01:03:40,620 --> 01:03:42,960
And so, given what
we've seen before,

1519
01:03:42,960 --> 01:03:47,110
given the Schottky method
of producing the band

1520
01:03:47,110 --> 01:03:50,790
diagrams here, we would
expect that, in this case,

1521
01:03:50,790 --> 01:03:52,920
we would see bands
bending, say, down.

1522
01:03:52,920 --> 01:03:54,450
And in that case-- let's see.

1523
01:03:54,450 --> 01:03:54,590
No.

1524
01:03:54,590 --> 01:03:55,920
In this case, we'd
see bands bending down

1525
01:03:55,920 --> 01:03:56,690
in the semiconductor.

1526
01:03:56,690 --> 01:03:58,398
And in this case, the
bands would bend up

1527
01:03:58,398 --> 01:04:00,420
in the semiconductor as
we reach that interface.

1528
01:04:00,420 --> 01:04:05,080
But because the Fermi energy is
pinned by these surface states

1529
01:04:05,080 --> 01:04:11,460
here, we have very little
motion of those bands.

1530
01:04:11,460 --> 01:04:14,270
And so we wind up in a
case where we pin the Fermi

1531
01:04:14,270 --> 01:04:17,100
energy at the surface
and, pretty much,

1532
01:04:17,100 --> 01:04:19,710
no matter what metal we
deposit on top of our sample,

1533
01:04:19,710 --> 01:04:21,240
we can wind up
with a certain type

1534
01:04:21,240 --> 01:04:23,150
of behavior of
the semiconductor,

1535
01:04:23,150 --> 01:04:25,360
a certain current
voltage response.

1536
01:04:25,360 --> 01:04:28,381
And this can drive
someone absolutely insane

1537
01:04:28,381 --> 01:04:29,880
when you're in the
laboratory trying

1538
01:04:29,880 --> 01:04:31,910
to get a repeatable
contact and everything

1539
01:04:31,910 --> 01:04:33,450
is changing all the time.

1540
01:04:33,450 --> 01:04:35,190
And you walk through
the math, and you

1541
01:04:35,190 --> 01:04:38,420
know that you should be getting,
let's say, an Ohmic contact.

1542
01:04:38,420 --> 01:04:41,970
You walk through the math,
and you derive the energy band

1543
01:04:41,970 --> 01:04:43,020
diagram of the interface.

1544
01:04:43,020 --> 01:04:44,769
And you go to the lab,
and over, and over,

1545
01:04:44,769 --> 01:04:46,920
and over again, you
obtain a Schottky contact.

1546
01:04:46,920 --> 01:04:48,920
And it just drives you insane.

1547
01:04:48,920 --> 01:04:55,160
The trick is to always use
repeatable surface preparation.

1548
01:04:55,160 --> 01:04:56,900
Always prepare
your surfaces using

1549
01:04:56,900 --> 01:05:01,630
the same chemistry, the
same way, every single time.

1550
01:05:01,630 --> 01:05:05,130
And if you do that, most
likely, in many cases,

1551
01:05:05,130 --> 01:05:07,630
you can get rid of this
native surface oxide

1552
01:05:07,630 --> 01:05:09,080
on your semiconducting sample.

1553
01:05:09,080 --> 01:05:11,400
But if you take your sample
out of the growth chamber

1554
01:05:11,400 --> 01:05:13,099
and move it around
in air and then

1555
01:05:13,099 --> 01:05:14,640
put it into the
metallization chamber

1556
01:05:14,640 --> 01:05:16,340
where you evaporate
your contacts,

1557
01:05:16,340 --> 01:05:19,110
you're exposing your
sample to the air.

1558
01:05:19,110 --> 01:05:21,040
You'll form some
surface layer on it,

1559
01:05:21,040 --> 01:05:23,590
and your contacts
will be different.

1560
01:05:23,590 --> 01:05:25,610
There was a colleague
of mine, who's now

1561
01:05:25,610 --> 01:05:27,075
a professor at a university.

1562
01:05:27,075 --> 01:05:31,090
And during his PhD, I
think he spent about a year

1563
01:05:31,090 --> 01:05:34,550
on a problem like this,
trying to make good contacts.

1564
01:05:34,550 --> 01:05:37,460
And unfortunately,
contacts, it's

1565
01:05:37,460 --> 01:05:41,990
such a basic thing that
it's one of the first skills

1566
01:05:41,990 --> 01:05:45,024
to go when the field moves on.

1567
01:05:45,024 --> 01:05:46,440
So for example,
people really knew

1568
01:05:46,440 --> 01:05:50,230
how to make good contacts
at MIT 30 years ago,

1569
01:05:50,230 --> 01:05:53,240
when the IC industry was
the hot topic of the day.

1570
01:05:53,240 --> 01:05:55,860
Then, as that phased
out, the knowledge

1571
01:05:55,860 --> 01:05:59,920
of how to make good contact
onto silicon disappeared.

1572
01:05:59,920 --> 01:06:01,700
There was maybe one
or two professors

1573
01:06:01,700 --> 01:06:03,980
who really had that knowledge
embedded in their minds.

1574
01:06:03,980 --> 01:06:06,360
And so, when a new
PhD student came along

1575
01:06:06,360 --> 01:06:08,790
at Harvard University
trying to make

1576
01:06:08,790 --> 01:06:10,640
Ohmic contact into
silicon, there

1577
01:06:10,640 --> 01:06:12,615
wasn't anybody he could
go to and talk to.

1578
01:06:12,615 --> 01:06:14,670
He didn't know where
to start or who

1579
01:06:14,670 --> 01:06:17,970
to talk to, because this
is kind of base knowledge.

1580
01:06:17,970 --> 01:06:20,860
It's really fundamental,
instrumental knowledge

1581
01:06:20,860 --> 01:06:22,960
that is going to be
infinitely useful.

1582
01:06:22,960 --> 01:06:26,330
But it quickly gets lost
when a field moves on,

1583
01:06:26,330 --> 01:06:28,650
when the next material
system becomes hot,

1584
01:06:28,650 --> 01:06:33,030
or the next topical area is the
key research area of the day.

1585
01:06:33,030 --> 01:06:36,110
If you know how to make good
contacts onto a semiconductor,

1586
01:06:36,110 --> 01:06:38,070
I guarantee you, you
will find someplace

1587
01:06:38,070 --> 01:06:39,662
in industry or
someplace in academia

1588
01:06:39,662 --> 01:06:41,620
where you'll be well-sought
out, and that skill

1589
01:06:41,620 --> 01:06:43,452
will be put to good use.

1590
01:06:43,452 --> 01:06:45,410
Let's talk about this
space-charge region width

1591
01:06:45,410 --> 01:06:48,510
very briefly, because
we're running out of time.

1592
01:06:48,510 --> 01:06:50,230
We've talked about
this barrier as being

1593
01:06:50,230 --> 01:06:52,450
this insurmountable
barrier, let's say,

1594
01:06:52,450 --> 01:06:55,300
for an electron to travel
over, in this case,

1595
01:06:55,300 --> 01:06:56,660
to get to the metal.

1596
01:06:56,660 --> 01:07:00,260
But what happens if we
increase the dopant density

1597
01:07:00,260 --> 01:07:01,819
within the semiconductor?

1598
01:07:01,819 --> 01:07:04,110
If we increase the dopant
density in the semiconductor,

1599
01:07:04,110 --> 01:07:08,330
then the amount of charge per
unit volume at that interface

1600
01:07:08,330 --> 01:07:11,840
or per unit thickness at
the interface increases.

1601
01:07:11,840 --> 01:07:14,400
So to compensate some buildup
of charge in the metal,

1602
01:07:14,400 --> 01:07:17,640
we would need a thinner
region of the semiconductor,

1603
01:07:17,640 --> 01:07:20,470
because it's more heavily doped,
because there are more dopants

1604
01:07:20,470 --> 01:07:22,136
per cubic centimeter.

1605
01:07:22,136 --> 01:07:24,635
And hence, there are more free
carriers per cubic centimeter

1606
01:07:24,635 --> 01:07:26,060
as well.

1607
01:07:26,060 --> 01:07:28,600
And as a result, the width
of this space-charge region

1608
01:07:28,600 --> 01:07:30,521
will decrease.

1609
01:07:30,521 --> 01:07:32,770
And if the width of the
space-charge region decreases,

1610
01:07:32,770 --> 01:07:33,730
what happens?

1611
01:07:33,730 --> 01:07:36,750
This barrier height
still stays the same.

1612
01:07:36,750 --> 01:07:39,520
The barrier height doesn't
change, but the width changes.

1613
01:07:39,520 --> 01:07:41,320
And what do we know
from quantum mechanics,

1614
01:07:41,320 --> 01:07:44,390
if our barrier is very,
very thin or very narrow?

1615
01:07:44,390 --> 01:07:45,110
What happens?

1616
01:07:45,110 --> 01:07:45,752
Electrons will?

1617
01:07:45,752 --> 01:07:46,460
AUDIENCE: Tunnel.

1618
01:07:46,460 --> 01:07:47,501
TONIO BUONASSISI: Tunnel.

1619
01:07:47,501 --> 01:07:49,390
And that's exactly what
happens right here.

1620
01:07:49,390 --> 01:07:53,310
So if we have a very
narrow space-charge region,

1621
01:07:53,310 --> 01:07:56,850
we'll have a tunneling
junction, field emission effect,

1622
01:07:56,850 --> 01:07:58,090
it's also known as.

1623
01:07:58,090 --> 01:08:01,150
And irrespective of
the barrier height,

1624
01:08:01,150 --> 01:08:04,330
you will get electrons
moving straight across.

1625
01:08:04,330 --> 01:08:09,247
Now, if we have a wide barrier--
represented there at the top--

1626
01:08:09,247 --> 01:08:11,080
there is an energy
barrier for the electrons

1627
01:08:11,080 --> 01:08:15,210
to overcome before they reach
the metal on the other side.

1628
01:08:15,210 --> 01:08:17,779
And so there, we have
a thermionic emission,

1629
01:08:17,779 --> 01:08:21,729
which means that you need
to have an excited electron,

1630
01:08:21,729 --> 01:08:24,250
thermally excited-- that's
where the name therm comes

1631
01:08:24,250 --> 01:08:26,359
from-- so thermally
excited electron,

1632
01:08:26,359 --> 01:08:28,130
before it manages to hop across.

1633
01:08:28,130 --> 01:08:29,939
And because there's
and exponentially

1634
01:08:29,939 --> 01:08:32,064
decaying number of electrons
with higher and higher

1635
01:08:32,064 --> 01:08:34,830
energies, we'll get a very
small number of electrons

1636
01:08:34,830 --> 01:08:36,300
that can actually make it.

1637
01:08:36,300 --> 01:08:40,490
So the thermionic emission
case will, in effect,

1638
01:08:40,490 --> 01:08:43,470
manifest itself as a
higher series resistance

1639
01:08:43,470 --> 01:08:45,090
than the field emission case.

1640
01:08:45,090 --> 01:08:48,870
You can have the same metal
contacting a semiconductor.

1641
01:08:48,870 --> 01:08:52,870
But if the dopant
density is changing,

1642
01:08:52,870 --> 01:08:55,180
you will make a
better Ohmic contact.

1643
01:08:55,180 --> 01:08:57,470
And of course, when
you're actually

1644
01:08:57,470 --> 01:09:01,140
trying to measure the contact
resistance here or measure

1645
01:09:01,140 --> 01:09:03,649
the barrier height,
you really have

1646
01:09:03,649 --> 01:09:06,040
to know this interface
really, really well.

1647
01:09:06,040 --> 01:09:07,990
You can measure experimentally.

1648
01:09:07,990 --> 01:09:09,890
But to determine
from theory, you

1649
01:09:09,890 --> 01:09:12,090
get a good start by
using the Schottky

1650
01:09:12,090 --> 01:09:13,459
models we described right here.

1651
01:09:13,459 --> 01:09:16,000
But to really nail it, you need
to perform density functional

1652
01:09:16,000 --> 01:09:17,833
theory and determine
the charge distribution

1653
01:09:17,833 --> 01:09:19,709
across that interface.

1654
01:09:19,709 --> 01:09:22,729
It's really the best
way to go about it.

1655
01:09:22,729 --> 01:09:24,779
So in terms of
experiment, thankfully,

1656
01:09:24,779 --> 01:09:27,220
there are big tables
that we can look up.

1657
01:09:27,220 --> 01:09:30,140
If we have a particular
semiconducting material that's

1658
01:09:30,140 --> 01:09:33,410
well-studied, we can look
up the contact materials

1659
01:09:33,410 --> 01:09:35,930
that would make the best
contacts, and what techniques

1660
01:09:35,930 --> 01:09:39,040
we can use to deposit them, what
temperature we should anele at

1661
01:09:39,040 --> 01:09:40,359
to form that alloy.

1662
01:09:40,359 --> 01:09:42,229
These are very
practical look up tables

1663
01:09:42,229 --> 01:09:45,670
that exist in reference
books, trusted websites-- not

1664
01:09:45,670 --> 01:09:48,635
any website, but the trusted
ones-- review articles.

1665
01:09:48,635 --> 01:09:51,010
But let's say you're working
on a semiconducting material

1666
01:09:51,010 --> 01:09:52,529
that's not on this list.

1667
01:09:52,529 --> 01:09:55,220
Let's say you're working on
cuprous oxide, for instance,

1668
01:09:55,220 --> 01:09:56,940
and you have to
start from scratch.

1669
01:09:56,940 --> 01:09:58,610
The best way to go
about it would simply

1670
01:09:58,610 --> 01:10:01,580
be to measure the contact
resistance using the techniques

1671
01:10:01,580 --> 01:10:04,360
that we described two lectures
ago where we described the TLM

1672
01:10:04,360 --> 01:10:13,240
method or the-- sorry, not
much sleep-- the Transfer Line

1673
01:10:13,240 --> 01:10:13,940
Method.

1674
01:10:13,940 --> 01:10:16,897
So the important
thing here is to know

1675
01:10:16,897 --> 01:10:19,230
that you do have a few tools
at your disposal to measure

1676
01:10:19,230 --> 01:10:21,330
barrier heights experimentally.

1677
01:10:21,330 --> 01:10:26,290
And at the end of the day, you
can obtain precise numbers.

1678
01:10:26,290 --> 01:10:29,420
Just going quickly
into heterojunctions.

1679
01:10:29,420 --> 01:10:32,055
This is similar to
the Schottky junctions

1680
01:10:32,055 --> 01:10:34,430
we've been talking about so
far, Schottky, Ohmic contacts

1681
01:10:34,430 --> 01:10:36,750
between metals and
semiconductors.

1682
01:10:36,750 --> 01:10:39,100
This, very, very
briefly, is what

1683
01:10:39,100 --> 01:10:42,810
happens when a semiconductor
meets a semiconductor.

1684
01:10:42,810 --> 01:10:45,720
Not always can you form
a homo-type junction,

1685
01:10:45,720 --> 01:10:49,160
meaning not always can you
have n- and p-type dopants

1686
01:10:49,160 --> 01:10:51,290
in a semiconductor like silicon.

1687
01:10:51,290 --> 01:10:54,160
Sometimes, it just won't
dope both n- and p-type.

1688
01:10:54,160 --> 01:10:57,180
And so you have to make contact
with another semiconducting

1689
01:10:57,180 --> 01:10:57,920
material.

1690
01:10:57,920 --> 01:11:00,090
And for that, we call
it a heterojunction.

1691
01:11:00,090 --> 01:11:02,649
And there are three general
types of heterojunctions.

1692
01:11:02,649 --> 01:11:04,190
One in which you
have a wide band gap

1693
01:11:04,190 --> 01:11:07,390
material on the left, a narrow
band gap material on the right.

1694
01:11:07,390 --> 01:11:09,040
The two are in contact
with each other.

1695
01:11:09,040 --> 01:11:11,630
And the vacuum levels
line up into such a way

1696
01:11:11,630 --> 01:11:15,060
as to wind up with
this band alignment.

1697
01:11:15,060 --> 01:11:17,060
Now, let's say I excite
an electron hole pair

1698
01:11:17,060 --> 01:11:18,520
in this wide band gap material.

1699
01:11:18,520 --> 01:11:20,550
The hole will be
able to move across.

1700
01:11:20,550 --> 01:11:22,449
The electron will be
able to move across.

1701
01:11:22,449 --> 01:11:24,990
So both the electron and hole
have moved across the junction,

1702
01:11:24,990 --> 01:11:26,960
and we have not
separated charge.

1703
01:11:26,960 --> 01:11:30,400
So type one, very bad from the
photovoltaic point of view.

1704
01:11:30,400 --> 01:11:33,150
Type two, this is more
similar to the PN junction,

1705
01:11:33,150 --> 01:11:34,990
what we've seen so far.

1706
01:11:34,990 --> 01:11:36,880
And in this particular
case, if you

1707
01:11:36,880 --> 01:11:38,850
excite an electron-hole
pair, the electron

1708
01:11:38,850 --> 01:11:40,630
will move across
because of the field,

1709
01:11:40,630 --> 01:11:42,456
but the hole will be repelled.

1710
01:11:42,456 --> 01:11:44,580
And so the type two junction
is actually preferred,

1711
01:11:44,580 --> 01:11:47,150
from a photovoltaic
point of view.

1712
01:11:47,150 --> 01:11:50,180
Type three junction will
still separate charge.

1713
01:11:50,180 --> 01:11:52,312
The electron will move
from the left-hand side

1714
01:11:52,312 --> 01:11:53,270
to the right-hand side.

1715
01:11:53,270 --> 01:11:54,750
And the hole will be repelled.

1716
01:11:54,750 --> 01:11:57,480
But now that the electron has
moved to the right-hand side,

1717
01:11:57,480 --> 01:12:00,450
it actually has a lower energy
than on the right-hand side.

1718
01:12:00,450 --> 01:12:02,050
So there is no driving force.

1719
01:12:02,050 --> 01:12:04,050
There's no voltage built
up for the electron

1720
01:12:04,050 --> 01:12:06,280
to go across, outside
of an external circuit

1721
01:12:06,280 --> 01:12:08,070
and power and external load.

1722
01:12:08,070 --> 01:12:09,940
And so, from a
photovoltaic point of view,

1723
01:12:09,940 --> 01:12:12,330
the type three, unless
under very high illumination

1724
01:12:12,330 --> 01:12:16,220
conditions, is generally
considered not desirable,

1725
01:12:16,220 --> 01:12:16,860
as well.

1726
01:12:16,860 --> 01:12:20,510
So the type two junction is
the preferred junction type

1727
01:12:20,510 --> 01:12:22,670
for photovoltaic devices.

1728
01:12:22,670 --> 01:12:26,000
And yeah, that's pretty much it.

1729
01:12:26,000 --> 01:12:28,270
So you can go through the
exercise shown right here

1730
01:12:28,270 --> 01:12:31,790
on your own and determine,
for this particular system

1731
01:12:31,790 --> 01:12:34,250
right here, what will
happen at that junction.

1732
01:12:34,250 --> 01:12:35,960
How will those bands align?

1733
01:12:35,960 --> 01:12:39,150
You have the cheat
sheet shown right after.

1734
01:12:39,150 --> 01:12:41,540
But an open question
for you is, using

1735
01:12:41,540 --> 01:12:44,600
what you know from today where
we first lined up the vacuum

1736
01:12:44,600 --> 01:12:48,080
levels, drew out the band
diagrams independently

1737
01:12:48,080 --> 01:12:49,850
for material one
and material two,

1738
01:12:49,850 --> 01:12:52,270
just like we've done there,
now we eliminate that barrier

1739
01:12:52,270 --> 01:12:52,936
in between them.

1740
01:12:52,936 --> 01:12:55,890
We put them together, and
we match the Fermi energies.

1741
01:12:55,890 --> 01:12:59,540
In other words, this Fermi
energy, Ef 1 has to equal Ef 2.

1742
01:12:59,540 --> 01:13:02,180
So there has to be this
shift going on like that.

1743
01:13:02,180 --> 01:13:03,730
What happens then?

1744
01:13:03,730 --> 01:13:06,210
How will the bands align?

1745
01:13:06,210 --> 01:13:07,380
Follow the vacuum level.

1746
01:13:07,380 --> 01:13:09,860
That's my best advice
I can give you.

1747
01:13:09,860 --> 01:13:11,890
Let the conduction
band and valence

1748
01:13:11,890 --> 01:13:15,890
bands follow the bending
of the vacuum level.

1749
01:13:15,890 --> 01:13:17,400
And then, from
there, you should be

1750
01:13:17,400 --> 01:13:20,720
able to determine the
final structure that'll

1751
01:13:20,720 --> 01:13:22,140
look something like that.

1752
01:13:22,140 --> 01:13:24,400
But that's a
take-home assignment,

1753
01:13:24,400 --> 01:13:28,380
just to see if the concepts
really jelled today.

1754
01:13:28,380 --> 01:13:30,440
So what did we learn how to do?

1755
01:13:30,440 --> 01:13:34,610
Or at least, what do we have
a greater appreciation for?

1756
01:13:34,610 --> 01:13:36,340
We have a greater
appreciation for all

1757
01:13:36,340 --> 01:13:38,970
of these different
components, how

1758
01:13:38,970 --> 01:13:41,290
to make different
types of contacts

1759
01:13:41,290 --> 01:13:44,300
on a material, what they're
good for, and ultimately, how

1760
01:13:44,300 --> 01:13:48,630
this band alignment that we've
learned in our PN junction

1761
01:13:48,630 --> 01:13:52,040
expropriations can be applied
as well to metal contacts,

1762
01:13:52,040 --> 01:13:53,910
and ultimately,
how this can also

1763
01:13:53,910 --> 01:13:56,550
be applied to semiconductor,
semiconductor contacts where

1764
01:13:56,550 --> 01:13:58,674
we might have two different
types of semiconductors

1765
01:13:58,674 --> 01:14:00,770
coming together and
forming a junction.

1766
01:14:00,770 --> 01:14:03,200
How do we predict what
the energy band diagram

1767
01:14:03,200 --> 01:14:06,700
is going to look like, once
we have our final junction

1768
01:14:06,700 --> 01:14:07,520
right here?

1769
01:14:07,520 --> 01:14:10,110
And what do we expect that
to mean, in terms of charge

1770
01:14:10,110 --> 01:14:12,140
transport across that layer?

1771
01:14:12,140 --> 01:14:14,390
So those are the building
blocks, the basic building

1772
01:14:14,390 --> 01:14:18,840
blocks that you will need
to be able to predict how

1773
01:14:18,840 --> 01:14:21,230
two materials will behave
when they encounter each other

1774
01:14:21,230 --> 01:14:24,010
and so that you can
make solar cell devices.

1775
01:14:24,010 --> 01:14:27,180
And I think, with that,
we can pretty much wrap up

1776
01:14:27,180 --> 01:14:29,650
our fundamentals
portion of the class.

1777
01:14:29,650 --> 01:14:30,530
Woo-hoo!

1778
01:14:30,530 --> 01:14:33,390
We get to talk about
technologies next.