1
00:00:00,540 --> 00:00:07,440
Okay. Welcome back. This is now the eighth lecture of the condensed matter course.

2
00:00:08,820 --> 00:00:14,940
One of the more wonderful things about quantum mechanics is that particles are waves and waves are particles.

3
00:00:15,390 --> 00:00:22,170
And that means the things that we learn when we study vibrational waves in solids can often be applied to other types of waves in solids,

4
00:00:22,170 --> 00:00:24,690
such as electron waves or even electromagnetic waves.

5
00:00:25,110 --> 00:00:31,650
So today we're going to be studying electron waves in solids, and this lecture is a little bit out of place because later on in the course,

6
00:00:31,890 --> 00:00:36,900
we're going to spend several days studying electron band structure, electron waves and solids in some amount of depth.

7
00:00:37,260 --> 00:00:43,229
But the reason I am inserting this lecture here is to make the point that we're really studying the same thing.

8
00:00:43,230 --> 00:00:49,530
Whether we're studying vibrational waves or we're studying electron electron waves, it's really very, very similar.

9
00:00:50,220 --> 00:00:55,890
We're also going to see that much of the calculation we do today is similar to what we did when we studied the covalent bond,

10
00:00:56,100 --> 00:01:05,250
which was a wave just between two atoms. So the picture we're going to look at today is a one dimensional, one dimensional, tight, binding chain.

11
00:01:09,850 --> 00:01:16,850
And it's going to be very analogous to the one dimensional vibrational change that we looked at in the last couple of lectures,

12
00:01:17,810 --> 00:01:26,240
also similar to the covalent bond. So we're going to imagine having a bunch of nuclei in a chain like this,

13
00:01:27,380 --> 00:01:32,210
and it will give them a lattice constant, a distance between identical nuclei.

14
00:01:32,570 --> 00:01:35,510
And we're going to add one electron to this chain of nuclei,

15
00:01:35,720 --> 00:01:42,110
and we're going to see what happens as we allow the electron to hop back and forth between the different nuclei.

16
00:01:42,830 --> 00:01:49,080
So of course, we have to start with the Hamiltonian as the usual piece squared over to term.

17
00:01:49,610 --> 00:02:04,100
Plus, it will have an interaction with all of the different nuclei of our minus our sub j where our sub j is position, position of nucleus.

18
00:02:07,280 --> 00:02:14,000
J And as I did when I when we studied the covalent bond, we're going to abbreviate these terms for convenience.

19
00:02:14,120 --> 00:02:23,330
This term will be called K for kinetic energy and these terms will be called V sub J for interaction with the nucleus.

20
00:02:25,040 --> 00:02:27,859
Now, similar to what we did when we started the covalent bond,

21
00:02:27,860 --> 00:02:36,020
it's useful to think first about an electron only interacting with a single nuclei, not with any of the other nuclei.

22
00:02:36,380 --> 00:02:44,630
So we'll write h oops or i k plus v some m as our Hamiltonian.

23
00:02:44,630 --> 00:02:49,530
And that means that the electron has this kinetic energy and it's interacting with the nucleus only.

24
00:02:51,200 --> 00:02:55,459
And we'll give it the atomic when it interacts with the nucleus.

25
00:02:55,460 --> 00:03:00,380
M only. So the eigen state of the Hamiltonian, which has it interacting with the nucleus,

26
00:03:00,590 --> 00:03:07,790
we'll call that the cat M and that will put the electron on the end of the nucleus, as if all of the other nuclei were not there at all.

27
00:03:08,030 --> 00:03:14,240
So a label this cat one if the electron is sitting here as if it's not interacting with any of the other nuclei,

28
00:03:14,450 --> 00:03:17,810
we'll call this one to this one, three and so forth.

29
00:03:18,880 --> 00:03:22,270
Okay. Happy so far. Okay.

30
00:03:23,470 --> 00:03:28,690
All right. Now, as we did when we started the covalent bond, we're going to make a bad assumption,

31
00:03:29,500 --> 00:03:37,900
that assumption, which is that these cats and an m are or the normal.

32
00:03:39,070 --> 00:03:45,460
This is not too bad if the nuclei are far apart from each other because an electron sitting over here,

33
00:03:45,490 --> 00:03:49,150
an electron sitting way over there, are pretty much orthogonal to each other.

34
00:03:49,480 --> 00:03:54,010
But when the nuclei get close together, then they're not orthogonal anymore.

35
00:03:54,370 --> 00:03:58,929
And the reason we make this approximation, even those of an approximation, is for simplicity.

36
00:03:58,930 --> 00:04:04,720
A lot of the calculation just gets a lot easier if we make assumption and we don't learn a whole lot more from doing more properly.

37
00:04:04,900 --> 00:04:10,960
It is not that much harder to do it properly. There's an exercise in the book that walks you through it and you can go through it if you want,

38
00:04:11,230 --> 00:04:15,520
but you'll get most of the interesting physics out of just this simplified approximation.

39
00:04:16,390 --> 00:04:23,050
Okay. So once we have made this that assumption, we can write down our trial wave function,

40
00:04:23,590 --> 00:04:37,540
trial wave function very similar to what we did with the covalent bond, which will have the form psi equals sum over n phi n n.

41
00:04:37,900 --> 00:04:40,450
A linear combination of atomic orbitals,

42
00:04:40,660 --> 00:05:01,120
which is a word we used equivalent to type binding linear combination of atomic orbitals, of atomic orbitals or LCA.

43
00:05:01,150 --> 00:05:11,530
Oh, we're making, we have a bunch of atomic orbitals electrons sitting on nuclear.

44
00:05:11,560 --> 00:05:15,640
So we're going to make a linear combination of them with coefficients vice some n.

45
00:05:15,910 --> 00:05:21,160
The reason people love this type of approximation is because you can make it more and more accurate

46
00:05:21,370 --> 00:05:27,579
by just adding more things to the right hand side with variational parameters phi in front of them.

47
00:05:27,580 --> 00:05:34,480
So for example, you could have an electron sitting on site n in excited state alpha.

48
00:05:34,690 --> 00:05:38,590
So this could be 1s2 as to P, so forth and so on.

49
00:05:38,740 --> 00:05:44,380
And we can just make our bases state bigger and bigger. Bases get bigger and bigger and bigger and give all of these coefficients.

50
00:05:44,530 --> 00:05:50,470
And as we make the bases set bigger and bigger and bigger and bigger, we get a more and more accurate approximation of the true wave function.

51
00:05:50,750 --> 00:05:57,490
Okay, this is sort of a variational approach so we can solve for the ground state by finding the best r coefficients.

52
00:05:57,730 --> 00:06:01,720
Then once we have the ground state, we can solve for the first excited state by finding the best coefficients,

53
00:06:01,900 --> 00:06:07,600
the lowest energy coefficients, subject to being orthogonal to the first thing that we just found and so forth and so on.

54
00:06:08,730 --> 00:06:15,150
So now what we have to do is we have to write down an equation for these coefficients.

55
00:06:15,420 --> 00:06:30,200
And the equation, again, something that you will solve for in your your homework will be of the form h and am phi m equals e phi n and you know,

56
00:06:30,540 --> 00:06:33,629
it looks like a shorter equation. It quacks like a Schrodinger equation.

57
00:06:33,630 --> 00:06:41,550
It probably is a shortening equation. It's not the real Schrodinger equation because the real equation has a full Hamiltonian sitting here, here.

58
00:06:41,550 --> 00:06:51,330
Instead we just have a matrix and h m and you can think of this as being the projection of the true Schroder

59
00:06:51,330 --> 00:06:58,680
equation onto the basis said made up of these cats and that we that we are working with does that make sense.

60
00:06:58,680 --> 00:07:03,720
It looks like a Schrödinger equation. It's going to act like a shorthand equation and you'll derive it also.

61
00:07:04,860 --> 00:07:15,450
Okay. So given our an equation over here, we have to now calculate these matrix elements that go into our ordering equation.

62
00:07:15,450 --> 00:07:21,240
So let's do that. It is useful to take our Hamiltonian and divide it up into pieces.

63
00:07:21,660 --> 00:07:31,300
So first we will remove the m the interaction with the nucleus from the interaction with all the other nuclei.

64
00:07:31,320 --> 00:07:34,380
So this will be j not equal to m the sum j.

65
00:07:35,640 --> 00:07:44,200
The reason we do this is because then we can write h on the cat m equals k plus we m on the cat.

66
00:07:44,220 --> 00:07:49,830
M plus the interaction with all the other nuclei.

67
00:07:49,860 --> 00:07:58,319
The sub j and cat. M. Okay. Now we defined the cat m to be the eigen state of k plus v m.

68
00:07:58,320 --> 00:08:03,600
So this thing here is just atomic on cat m.

69
00:08:04,950 --> 00:08:10,190
Good. Happy with that. So then we can take the inner product,

70
00:08:10,400 --> 00:08:29,330
close up the inner product with the head and over here and we get the atomic delta and m plus and some over j not equal to m v some j m.

71
00:08:29,900 --> 00:08:39,370
And this is the interesting term here. So this term here just tells us that no matter what side the electron is sitting on,

72
00:08:39,370 --> 00:08:43,360
which nucleus, the electron sitting sitting on it has energy, atomic.

73
00:08:43,600 --> 00:08:49,360
And this is all the interaction with all of the other nuclei. Now, there's a couple of things that can happen with this term.

74
00:08:49,390 --> 00:08:55,480
One possibility is that any equals M, in which case this is what this is telling you,

75
00:08:55,480 --> 00:09:04,630
is that there is some change in its energy sitting on Nucleus M due to its interaction with all of the other atoms, not M okay.

76
00:09:04,990 --> 00:09:15,670
So this will be will give this thing an energy. We'll call it V, not equals R and some big j, not equal to M.

77
00:09:17,170 --> 00:09:28,030
And I guess these can be ends like that. And this is so this is interacting with all the nuclei, not including M and its expectation of its energy.

78
00:09:28,030 --> 00:09:32,470
So it's just shifting its energy on that particular site. That's not particularly interesting.

79
00:09:32,770 --> 00:09:36,610
The more interesting thing is what happens is and not equal to M.

80
00:09:37,120 --> 00:09:43,420
So in this case you have a this term is maybe I should call this term something first.

81
00:09:43,440 --> 00:09:53,740
This is what we call direct before and I guess I call it V Cross when we talked about the covalent bond for the energy equal to M term,

82
00:09:54,190 --> 00:10:04,690
this is what we call hopping before. And the reason we call it hopping was because it will give it the value minus T, minus T.

83
00:10:05,680 --> 00:10:10,990
The reason we call it hopping is if you think in the time dependent Schrödinger equation type of way,

84
00:10:11,860 --> 00:10:17,440
you can take an electron sitting inside M and have it end up onsite rn.

85
00:10:17,710 --> 00:10:23,850
So this off diagonal term in the Hamiltonian allows an electron to move from one side to another.

86
00:10:23,860 --> 00:10:29,110
Hence we call it hopping. Now an approximation, which is actually a fairly good approximation.

87
00:10:29,410 --> 00:10:41,080
Is that n minus M greater than one hopping is zero, hopping equals zero or is approximately zero.

88
00:10:41,500 --> 00:10:46,750
And the reason for that is because it's very hard for an electron to hop very far in one step.

89
00:10:47,020 --> 00:10:48,520
If you think about it for a second,

90
00:10:48,790 --> 00:10:55,540
what we're really calculating when we take this matrix columns is some sort of when we write out the matrix element explicitly,

91
00:10:55,960 --> 00:11:01,820
it's something like this. Right.

92
00:11:01,970 --> 00:11:06,380
This is what matrix elements look like. A bra, a cat, and an interaction.

93
00:11:06,650 --> 00:11:14,450
Now, if I n if an m are far apart, these wave functions decay very quickly as you go away from the nucleus.

94
00:11:14,720 --> 00:11:18,530
So there there'll be no point in space where both this is large and this is large.

95
00:11:18,770 --> 00:11:20,720
If the two nuclei are very far apart.

96
00:11:21,080 --> 00:11:29,930
So that's why we can we can assume that this matrix element is going to be zero in less and an m are essentially neighbours.

97
00:11:30,050 --> 00:11:43,370
Okay, good. So what we have now is we have this a in the end we have that we can write and sum over j not equal to m v sub j.

98
00:11:43,550 --> 00:11:54,890
M equals the direct term v not if and equals m we'll call it minus t if n equals m plus or minus one.

99
00:11:55,190 --> 00:11:58,460
So we can have one side only and zero. Otherwise.

100
00:12:01,870 --> 00:12:05,650
Okay. Good people happy. Fairly happy with that.

101
00:12:07,510 --> 00:12:10,900
Ken? Yes. Someone had someone say. Someone say yes.

102
00:12:11,050 --> 00:12:15,820
Yes. Thank you. I would send you another chocolate, but I gave you one yesterday, so.

103
00:12:15,850 --> 00:12:23,919
Okay. Good. So we can take a Hamiltonian and rewrite it as h as a big matrix.

104
00:12:23,920 --> 00:12:27,670
H&M equals. What is it?

105
00:12:27,690 --> 00:12:36,960
It's atomic. Plus the interaction with all the other nuclei, not including the site itself.

106
00:12:37,170 --> 00:12:40,770
If you're sitting on one site interacting with all the other ones, not including yourself.

107
00:12:41,250 --> 00:12:49,830
Plus there's going to be another term which delta and and plus one plus delta and comma and minus one.

108
00:12:50,490 --> 00:12:56,580
So there's this additional term here which allows you to hop one step to the left or one step to the right.

109
00:12:58,580 --> 00:13:02,210
With an amplitude. A t. Okay.

110
00:13:03,300 --> 00:13:06,780
Good. So this is a great big matrix.

111
00:13:06,780 --> 00:13:17,610
If we have let's say we have NW and nuclei to begin with, then H is an end by and matrix and an end by an Hamiltonian matrix.

112
00:13:19,140 --> 00:13:22,230
And we need to find its eigenvalues. So how do we do that?

113
00:13:22,260 --> 00:13:26,460
That looks like a complicated problem if and is a is a pretty large number.

114
00:13:27,090 --> 00:13:33,300
Well, again, we can solve this very similar to what we did for the vibrational chains we use.

115
00:13:33,600 --> 00:13:37,440
And on that, which is an English word. And that's.

116
00:13:38,820 --> 00:13:46,650
And the answers we use is via is a plane way vanguards I k and a like this.

117
00:13:47,910 --> 00:13:56,070
Now a couple of comments about this. First of all, you may be expecting an the Iomega T from what we did with the vibrational chain.

118
00:13:56,460 --> 00:14:04,140
The reason there's no Iomega t is because we're solving the time independent shortening equation, not the time dependent shortening equation.

119
00:14:04,560 --> 00:14:08,970
If we were solving the time dependent shortening equation, there'd be an either the Iomega t as well.

120
00:14:09,270 --> 00:14:10,889
Okay. That's why it's not there. In this case.

121
00:14:10,890 --> 00:14:16,830
It's just simpler in quantum mechanics to work with time independent agencies than it is to work with time dependent wave functions.

122
00:14:17,340 --> 00:14:18,750
So that's the first thing.

123
00:14:18,930 --> 00:14:26,250
Second thing is that you probably if you're careful, you put now a square root of capital N downstairs so that this is a normalised wave function.

124
00:14:26,460 --> 00:14:30,990
The normalisation is not going to matter much for us, but strictly speaking, it should probably be there.

125
00:14:32,050 --> 00:14:36,490
And the third thing to note is that this this wave is the same.

126
00:14:36,520 --> 00:14:42,459
If you shift K to K plus to pi over A is something we discovered last time.

127
00:14:42,460 --> 00:14:45,430
The thing that is important is not the momentum but the crystal momentum.

128
00:14:45,610 --> 00:14:50,020
If you're shifting k by two pi over a, get back exactly the same the same wave.

129
00:14:50,440 --> 00:14:59,139
Okay, so let's take our and plug it into that Hamiltonian one step here we get epsilon,

130
00:14:59,140 --> 00:15:17,670
not even minus i k and a minus t into the minus i k and plus one a plus e to the minus i k and minus one a equals E in

131
00:15:17,680 --> 00:15:25,780
the minus i k and a So it's just plugging the on that into the shortening equation using that form of the Hamiltonian.

132
00:15:27,140 --> 00:15:32,080
Okay. So it's this is the oops. I didn't tell you what this I didn't tell you what it is.

133
00:15:32,090 --> 00:15:35,240
Sorry about that. This thing here I called in not.

134
00:15:39,710 --> 00:15:44,480
So this is the energy on site. This allows you to hot to let hop to the right.

135
00:15:44,720 --> 00:15:48,580
And this is the eigen energy on the other side. People happy.

136
00:15:49,970 --> 00:15:52,650
Okay, good. Thank you. Good.

137
00:15:52,660 --> 00:16:04,730
So then you just cancel out a bunch of factors, a bunch of exponential factors, and you get E is e, not minus two t cosine k.

138
00:16:06,310 --> 00:16:12,420
So let's actually plot that. You go?

139
00:16:13,690 --> 00:16:20,140
So here's pi over a year, minus pi over a year.

140
00:16:20,200 --> 00:16:23,530
This is energy on this axis. Energy.

141
00:16:23,830 --> 00:16:30,100
This is not. And then we'll have a nice cosine form that looks kind of like this looks kind of like this.

142
00:16:31,450 --> 00:16:35,260
And if you wanted to, you can continue it out periodically further.

143
00:16:35,770 --> 00:16:44,890
But what we're really interested in is the ways within the Bruins zone, one zone from here to here,

144
00:16:45,130 --> 00:16:49,450
because once we go outside of the Brown Zone, we're just reproducing the same waves over and over again.

145
00:16:49,720 --> 00:16:55,180
And if we want different waves, we have to be all the ways within the bronze zone in different from each other.

146
00:16:55,180 --> 00:16:59,170
But once we go outside of the bronze zone, we start repeating things that we've already considered.

147
00:17:00,340 --> 00:17:05,469
Okay. Just a bit of nomenclature, which is which is fairly useful.

148
00:17:05,470 --> 00:17:09,790
May I put it over here? An energy band. Energy.

149
00:17:12,010 --> 00:17:24,010
Banned means one case, one iron state at each.

150
00:17:26,170 --> 00:17:29,650
Each kick in the bronze arm.

151
00:17:36,640 --> 00:17:39,680
So here we've drawn an energy band.

152
00:17:39,700 --> 00:17:43,479
Now, you'll remember this looks a little bit like what we got when we.

153
00:17:43,480 --> 00:17:46,330
When we solved the monotonic harmonic chain.

154
00:17:46,330 --> 00:17:51,970
The vibrations of a single chain of where every atom is exactly the same and every spring is exactly the same.

155
00:17:52,420 --> 00:17:58,959
I last lecture when we when we solved the diatomic chain or the alternating chain, we found that there were two branches of excitations.

156
00:17:58,960 --> 00:18:04,330
There were two normal modes at each k vector K in this direction.

157
00:18:05,350 --> 00:18:09,940
If we had something similar in this picture where there are two eigen states, a low energy one and a high energy one,

158
00:18:11,080 --> 00:18:14,260
we would say that there are two energy bands aligned, one in a high energy band.

159
00:18:14,560 --> 00:18:22,540
We do not use the words acoustic and optical when we're talking about electrons for reasons that will become clear in a moment.

160
00:18:22,780 --> 00:18:31,360
But it is useful to compare this this dispersion to what we got for vibrations, vibrations,

161
00:18:32,410 --> 00:18:43,750
the atomic chain one atomic we had omega squared equals R I guess it was two kappa

162
00:18:43,750 --> 00:18:52,240
over M minus two kappa over m cosine k and I believe it was equivalent to that.

163
00:18:54,350 --> 00:18:58,960
You'll see that the two dispersions look awfully similar, but there is a notable difference.

164
00:18:59,260 --> 00:19:02,620
And the notable difference is that with the shorter equation,

165
00:19:02,620 --> 00:19:07,960
we got e on the left hand side and with the vibrational chain we had omega squared on the left hand side.

166
00:19:08,260 --> 00:19:14,050
The origin of this difference is that the shorter equation, if you think about the time dependent shortening equation,

167
00:19:14,200 --> 00:19:20,080
it has one time derivative, whereas Newton's equations F equals May, the acceleration is two time derivatives.

168
00:19:20,380 --> 00:19:26,290
So that's why we're getting two frequencies over the left hand side, but one only one energy on the right hand side.

169
00:19:26,290 --> 00:19:30,850
When we're thinking about Schrodinger, that actually makes a rather important distinction between the two.

170
00:19:31,240 --> 00:19:42,790
When we had the vibrational chain at low energy, we got linear modes coming in down to zero frequency at zero wave vector in this picture here.

171
00:19:45,940 --> 00:19:49,000
In the in the electrons.

172
00:19:49,030 --> 00:20:00,040
In the case of electrons, if you expand near zero wave vector, it's not minus two T and then low substituted for small wave vector k a squared

173
00:20:00,040 --> 00:20:07,870
over two like this we see that we get to see here what we get is constant,

174
00:20:08,920 --> 00:20:11,920
which I'm not interested in plus k.

175
00:20:11,920 --> 00:20:19,480
A. Squared times t. It is quadratic at low energy, as you would expect from a cosine.

176
00:20:19,870 --> 00:20:24,580
And that is why we do not use the word acoustic for these low energy modes for vibrations.

177
00:20:24,730 --> 00:20:31,629
Acoustic modes, by definition, sound modes have linear dispersion that the frequency should be proportional to wave vector.

178
00:20:31,630 --> 00:20:35,770
That's the definition of sound. So here is quadratic.

179
00:20:35,770 --> 00:20:42,640
So we don't call it acoustic and we don't call it optical either if there is higher energy brand branches.

180
00:20:42,970 --> 00:20:52,000
But this this picture of the energy being quadratic in wave vector should be fairly familiar to you from other contexts.

181
00:20:52,330 --> 00:20:59,440
For example, if you have a free electron free electrons, an electron in outer space flying around or something,

182
00:21:00,670 --> 00:21:07,480
it's energy is some constant plus h bar squared k squared over two m.

183
00:21:09,000 --> 00:21:15,120
K squared over to our right. It is also quadratic in in k.

184
00:21:15,120 --> 00:21:21,960
You might choose C to be zero if you wanted to, or you could choose it to be M.C. squared if you're thinking about it's mass energy as well.

185
00:21:23,950 --> 00:21:30,230
But the important thing is that it's quadratic and in a way vector the same as this electron over here.

186
00:21:30,240 --> 00:21:36,270
So it might be useful for us to sort of think in terms of free electrons for a second.

187
00:21:36,570 --> 00:21:39,900
And so what we do is we define an effective mass.

188
00:21:41,820 --> 00:21:53,640
Effective mass and the star such that h squared over two and a star equals A squared times t.

189
00:21:54,890 --> 00:22:03,410
With this definition. Then for our energy band, for an energy band, for electrons in our energy band, energy band,

190
00:22:04,040 --> 00:22:13,920
we still have energy equals a constant, same constant plus h bar squared k squared over two m star.

191
00:22:15,050 --> 00:22:20,150
It looks just like free electrons is still quadratic in k, at least a small k,

192
00:22:20,480 --> 00:22:25,910
but now it's an effective mass m star rather than the real physical mass of the electron.

193
00:22:25,910 --> 00:22:31,340
M okay, now this. You should really think about this for a second.

194
00:22:31,610 --> 00:22:39,170
The effective mass that we're getting here in this tight binding chain that we're solving has nothing to do with the actual mass of the electron.

195
00:22:39,530 --> 00:22:44,330
It actually has to do with the hopping between nuclei for greater hopping.

196
00:22:44,600 --> 00:22:48,620
The mass gets smaller for smaller hopping, the mass gets larger.

197
00:22:49,160 --> 00:22:52,180
I think. And so it has.

198
00:22:52,290 --> 00:22:57,730
It's totally unrelated to the to the actual physical mass of the electron.

199
00:22:57,940 --> 00:23:03,850
You will find sample systems, actual physical systems, metals or whatnot,

200
00:23:04,060 --> 00:23:13,450
where the mass is 100 times less than the physical mass of the electron or 100 or a thousand times more than the physical mass of the electron.

201
00:23:13,570 --> 00:23:20,770
Although it's not uncommon to have effective masses in real materials which are on the order of the actual mass of the physical electron.

202
00:23:21,370 --> 00:23:28,480
The other thing to keep in mind here is that the K we're talking about in the energy band is not momentum, but is crystal momentum.

203
00:23:30,270 --> 00:23:31,830
Crystal momentum or.

204
00:23:34,520 --> 00:23:47,120
Or Bar-Kays Christian momentum if you put the H brian which means only that we are only defining our momentum modulo the crystal wave vector,

205
00:23:47,180 --> 00:23:53,960
the reciprocal lattice wave vector to pi over a If I shifted everything by two pi over a, I get back the same physical wave.

206
00:23:54,170 --> 00:23:59,690
So it is slightly different from the free electron waves that were that we are familiar with.

207
00:24:01,070 --> 00:24:04,490
The last thing that I want to emphasise here are the Eigen states.

208
00:24:05,660 --> 00:24:11,390
I can states our waves, our plane waves are just waves.

209
00:24:12,440 --> 00:24:13,790
Now why is that interesting?

210
00:24:16,170 --> 00:24:25,200
It means that I can take an electron and if I put it in some state here, it's sort of a left, a right moving wave at right moving wave.

211
00:24:25,200 --> 00:24:34,320
And in that right moving wave extends clear across the system is in the IK an our form of the wave which I guess scrolled off the top of the board.

212
00:24:34,590 --> 00:24:40,049
But it's a plane way that goes clear across the system and that might be surprising for a second.

213
00:24:40,050 --> 00:24:43,560
Y Well, remember, when we when we studied Sommerfeld theory,

214
00:24:43,560 --> 00:24:49,740
there was this issue that the scattering length of electrons in SA seemed unreasonably long.

215
00:24:49,920 --> 00:24:53,880
You have all these nuclei with positive charges and they're all over the place.

216
00:24:53,970 --> 00:24:56,970
Every few angstroms you run into a nucleus with a positive charge,

217
00:24:57,120 --> 00:25:02,640
and yet the electron has a scattering light that can be 100 angstroms, 1000 angstroms or a million angstroms long.

218
00:25:03,120 --> 00:25:10,260
Now here we have a model where we have lots of nuclei lined up periodically, one after the other, after the other, after the other.

219
00:25:10,470 --> 00:25:15,720
And the electron hops from one to the next to the next next and makes a wave that goes clear across the system.

220
00:25:15,840 --> 00:25:23,550
No scattering at all. You would observe this as a perfectly good wave packet going entirely across the system without back scattering one bit.

221
00:25:23,880 --> 00:25:30,870
So that's our first surprising result, and we're going to come back to that result very frequently later on in the term.

222
00:25:31,650 --> 00:25:39,610
Is everyone fairly happy with that? It is, yes. Well, there's lots of other things that we haven't accounted for.

223
00:25:39,640 --> 00:25:43,840
For example, I can we haven't talked about it. It could scatter into other electrons.

224
00:25:43,870 --> 00:25:47,110
So that's actually in fact, that's something we're not going to discuss all term,

225
00:25:47,380 --> 00:25:53,800
because the reason you can ignore scattering against other electrons is actually extremely subtle and you probably won't even learn it next year.

226
00:25:53,980 --> 00:25:57,700
It wasn't understood properly until probably the 1980s or even later.

227
00:25:59,500 --> 00:26:03,460
And it was it wasn't understood at all. Not even one little bit until the 1950s.

228
00:26:03,790 --> 00:26:06,810
So that's a really tough question. But there's other things you can scatter.

229
00:26:06,820 --> 00:26:12,570
There's impurities. So crystals aren't perfect. There's other junk in it and there's vibrations.

230
00:26:12,580 --> 00:26:19,450
So you can have an electron moving along and they can hit a vibrational a phonon and scatter off of a phone on as well.

231
00:26:19,460 --> 00:26:24,250
So to find a temperature, there's electron phonon scattering is frequently the limiting process.

232
00:26:24,910 --> 00:26:30,310
Okay, good, good question. I give you another chocolate, but you already have one.

233
00:26:31,000 --> 00:26:38,920
Okay. So we're going to do an exercise we did before counting counting states.

234
00:26:40,000 --> 00:26:44,980
We did this for when we counted normal nodes in our vibrational chain.

235
00:26:46,240 --> 00:26:51,790
So. Well, we have n nuclei. Nuclei we have.

236
00:26:51,880 --> 00:26:57,610
So that means the length of our system is and times a and if the length of the system is in times a,

237
00:26:57,640 --> 00:27:04,450
the case we are allowed must be of the form two pi over l times p, where p p is an integer.

238
00:27:05,740 --> 00:27:09,310
An integer. You guys are familiar with that, that notation integer.

239
00:27:09,340 --> 00:27:24,249
Yeah. Okay, good. P is integer. So the number of different keys, different case is equal to the range of k's that are different,

240
00:27:24,250 --> 00:27:30,490
which is two pi over a the branch zone from here to here divided by the spacing between the adjacent case,

241
00:27:30,820 --> 00:27:36,850
which is two pi over L which is l over a which is n not surprising.

242
00:27:37,120 --> 00:27:42,850
The number of different k's were allowed in the brian zone is equal to the number of unit cells in the entire system.

243
00:27:43,820 --> 00:27:47,660
In this case, it's easy to understand that in a different way.

244
00:27:47,900 --> 00:27:51,140
We started with a bunch of cats.

245
00:27:51,380 --> 00:27:56,120
M They were end of them and enemies and orbitals.

246
00:27:56,480 --> 00:28:02,510
These were sitting on our nuclei in nucleus one, nucleus two, nucleus three and so forth.

247
00:28:02,960 --> 00:28:05,990
And then we diagnosed our Hamiltonian and we get iron states.

248
00:28:06,710 --> 00:28:10,970
K And there are unsurprisingly there are end of them and eigen states.

249
00:28:12,470 --> 00:28:16,459
We had an M and Big N by a big A matrix and we analysed it.

250
00:28:16,460 --> 00:28:22,250
We got eigenvalues and Eigen states. Okay, so you put in end state, you get out and states.

251
00:28:23,180 --> 00:28:31,579
One other piece of nomenclature which is fairly useful here, which is this energy scale here,

252
00:28:31,580 --> 00:28:37,070
which I guess for our last caller it off the top, it's actually 40.

253
00:28:39,860 --> 00:28:44,069
C where was it? Yeah.

254
00:28:44,070 --> 00:28:53,010
So it's a two times the cosine. The cosine two times the cosine is the, the range and cosine goes from plus 1 to -1.

255
00:28:53,010 --> 00:28:58,980
So the total range from Emacs is Emacs here as in in here.

256
00:28:59,910 --> 00:29:06,720
I'm in here. So. Emacs.

257
00:29:07,830 --> 00:29:13,560
Emacs minus im in is known as the bandwidth.

258
00:29:16,030 --> 00:29:20,530
With which in this case is 40.

259
00:29:22,010 --> 00:29:26,390
Okay. Now, what do we what what is the bandwidth depend on?

260
00:29:26,630 --> 00:29:31,880
It depends on how much hopping you have. The more hopping you have, the bigger the bandwidth.

261
00:29:32,180 --> 00:29:35,300
So let's actually draw a little picture here.

262
00:29:35,960 --> 00:29:40,790
What does the bandwidth actually depend on? What is what is t depend on?

263
00:29:41,390 --> 00:29:43,160
Well, the thing we had at our disposal,

264
00:29:43,160 --> 00:29:52,070
the thing we could change was the lattice constant a the distance between the nuclei as the distance between the nuclei goes up,

265
00:29:52,430 --> 00:29:56,390
the hopping naturally goes down. It's harder to hop over a longer distance.

266
00:29:56,720 --> 00:30:01,550
So t goes up, goes up, going this way.

267
00:30:02,980 --> 00:30:06,520
Good people happy with that. Okay. And then energy is going to be this way.

268
00:30:07,600 --> 00:30:17,440
And then we can draw a picture that kind of looks like this. We start with a bunch of atoms, all having eigen states with energy e atomic.

269
00:30:18,910 --> 00:30:22,899
When the atoms are very far apart, they all have the same, the same energy.

270
00:30:22,900 --> 00:30:29,140
The electron can sit here, it can sit here, it can sit on any one of the atoms and its energy is always e atomic.

271
00:30:29,320 --> 00:30:33,640
And there are and eigen states, and I guess it can.

272
00:30:35,390 --> 00:30:42,830
Sit on any one of the N atoms. Now, as we bring the atoms closer together, as we bring nuclei closer together.

273
00:30:43,580 --> 00:30:51,200
T is going to increase and we're going to get a band that goes from Imean to Emacs here.

274
00:30:51,380 --> 00:30:56,350
And there will be eigen states within this range between Imen and iMac.

275
00:30:56,360 --> 00:31:01,820
So you pick any energy between Imean in Emacs and there will be some eigen states there.

276
00:31:02,900 --> 00:31:10,010
So any energy between here and here, there will always be some k which has an state at that, at that energy.

277
00:31:10,380 --> 00:31:14,490
Now, this will actually look a little bit familiar. From.

278
00:31:16,840 --> 00:31:23,260
From when we studied the covalent bond. When we studied the covalent bond, we had two atoms with energy.

279
00:31:23,650 --> 00:31:28,660
When the electrons sat on one, nuclear energy was inside. And when we sat on the other nucleus, this energy was inside.

280
00:31:28,900 --> 00:31:32,890
Then we brought them together and we got an anti bonding orbital anti bond,

281
00:31:33,280 --> 00:31:41,350
and we got a bonding orbital bond down here, one lower and one higher than the original energies.

282
00:31:41,620 --> 00:31:48,310
And that's exactly what we have over here. We have a whole bunch of orbitals, all with the same energy.

283
00:31:48,430 --> 00:31:49,389
We bring them together,

284
00:31:49,390 --> 00:31:58,030
we allow the electron to hop back and forth and some of the energies go down and some of the energies go up and it spreads out into into a bath.

285
00:31:58,450 --> 00:32:01,510
Could that make sense? Yes, hopefully.

286
00:32:02,320 --> 00:32:10,060
So the spreading into a band is coming from the the hopping of the electron back and forth.

287
00:32:11,590 --> 00:32:20,090
So. Right. So let's imagine now that we have not just a single electron that we want to consider,

288
00:32:20,090 --> 00:32:28,000
but we have many electrons that we want to consider since we had any possible case dates.

289
00:32:30,310 --> 00:32:36,740
Case dates. But each case date, each case can have two spins.

290
00:32:37,100 --> 00:32:41,900
It can have spin up or spin down.

291
00:32:42,830 --> 00:32:53,180
That means there's two. And total electrons can fit can fit in band.

292
00:32:57,780 --> 00:33:01,230
So we fill up all the states. You can fill them up with either spin up or spin down.

293
00:33:01,410 --> 00:33:05,160
So we can have a total of two n electrons fitting in that band.

294
00:33:05,670 --> 00:33:15,210
So let's first consider a we are going to put this maybe here monovalent monovalent atom.

295
00:33:17,940 --> 00:33:21,300
Which has one electron per nucleus.

296
00:33:23,430 --> 00:33:38,280
So that will give us a total of nine electrons total and electrons total and that half fills the band, half filled band.

297
00:33:41,930 --> 00:33:49,070
Okay. So if we have an electron's total or one electron per atom, we will fill up the band.

298
00:33:50,200 --> 00:33:53,540
Halfway. Up to here.

299
00:33:53,840 --> 00:33:58,700
And this is filled now with both spin up and spin down electrons.

300
00:34:00,190 --> 00:34:06,810
Right now, this picture of a half filled band, half filled down.

301
00:34:11,380 --> 00:34:14,760
It has some interesting properties. The first thing.

302
00:34:15,240 --> 00:34:24,030
One, it has a Fermi surface has a Fermi surface surface.

303
00:34:26,040 --> 00:34:31,470
With that Fermi surface, Fermi surface is this point here where the filled states meet the empty states.

304
00:34:31,770 --> 00:34:34,860
And this point here where the fill states meet the empty state.

305
00:34:34,880 --> 00:34:35,610
So there are just two.

306
00:34:36,000 --> 00:34:44,110
Since it's one dimensional, it's actually just two Fermi points, two points where where you have the highest energy filled state.

307
00:34:44,110 --> 00:34:48,510
And that is an empty state nearby because since it has a Fermi surface,

308
00:34:49,620 --> 00:34:53,849
it is possible to make low energy excitations by taking some electron from

309
00:34:53,850 --> 00:34:58,230
just below the Fermi surface and exciting it to just above the Fermi surface.

310
00:34:58,530 --> 00:35:05,519
Okay. And that means that the heat capacity is going to be proportional to t like we calculated

311
00:35:05,520 --> 00:35:09,300
in the Sommerfeld theory because you can make as many low energy excitations.

312
00:35:09,690 --> 00:35:19,460
Well, you can make lots of low energy excitations with no problem. Maybe less obvious to is that this is a metal, it conducts electricity.

313
00:35:19,880 --> 00:35:22,970
So that is maybe a little bit harder to imagine.

314
00:35:23,150 --> 00:35:30,740
But let's think about it for a second. Here, what we have is we have over here we have right moving electrons.

315
00:35:30,740 --> 00:35:38,090
Maybe I'll label them the sense this one these guys over here have positive k h bar-kays to the right.

316
00:35:38,390 --> 00:35:45,350
So these are right movers. Right movers over here on this side of the diagram.

317
00:35:45,650 --> 00:35:50,660
And over here, we have left the movers. Left movers over on this side.

318
00:35:51,140 --> 00:35:53,630
So positive versus negative momentum.

319
00:35:55,430 --> 00:36:05,060
And if we if we apply an electric field with very little energy cost, we can take some of the electrons from over here and move them to over here.

320
00:36:05,570 --> 00:36:10,970
We can overpopulate the right hand side by a little bit and underpopulated the left hand side by a little bit.

321
00:36:11,060 --> 00:36:13,400
It costs you only very little energy to do so.

322
00:36:13,580 --> 00:36:23,900
Just shift the Fermi surface just a little bit and then you have a net current so metal can shift Fermi surface.

323
00:36:26,090 --> 00:36:32,510
For me surface. And get current.

324
00:36:38,010 --> 00:36:43,260
In fact, the way the way it actually happens is that you should really think about it.

325
00:36:43,260 --> 00:36:49,409
You apply an electric field and each electron state accelerates a little bit in one direction.

326
00:36:49,410 --> 00:36:58,590
So each electron changes its momentum a little bit until these guys get overpopulated, these guys get underpopulated, and then you have a net current.

327
00:36:59,820 --> 00:37:07,740
So this is indeed a metal. And indeed it's very, very frequently the case that monovalent atoms are metals.

328
00:37:08,340 --> 00:37:19,140
Often, I mean, a simple picture should always be true, but often monovalent monovalent materials.

329
00:37:22,270 --> 00:37:29,320
Our metals. Now further.

330
00:37:30,610 --> 00:37:39,590
Let us come back over here to the case, to this picture, and imagine what happens when we have fill the band.

331
00:37:40,110 --> 00:37:47,360
So we're going to take this band. We're going to half fill it. Now, we filled these dates here and we've left these dates up here empty.

332
00:37:47,510 --> 00:37:55,370
So this is a Hatfield band. Now, this is very similar to the covalent bond,

333
00:37:55,670 --> 00:38:05,450
that there is a lower net energy when the electrons can localise between the two nuclei because they're filling only the bonding orbital.

334
00:38:05,660 --> 00:38:11,990
And as long as you don't have to fill the anti bonding orbital. So here when you start letting the electrons hop back and forth,

335
00:38:12,320 --> 00:38:18,560
they can lower their energy compared to this energy that they started with, this atomic energy they started with.

336
00:38:18,770 --> 00:38:25,760
So there's a attractive force between the nuclei trying to get them closer together, trying to make the hopping much greater.

337
00:38:25,970 --> 00:38:31,990
So that is what's forming the metallic bond between forms.

338
00:38:32,780 --> 00:38:46,189
Metallic bond. Metallic bond between the nuclei is very, very similar to the covalent bond by letting the the electrons d localise,

339
00:38:46,190 --> 00:38:54,560
by letting them reduce their kinetic energy, by spreading out their wave function, it forms a bonding force that holds the nuclei together.

340
00:38:55,460 --> 00:39:01,280
Now, in this very simple picture, we have the same problem that we had when we started the covalent bond.

341
00:39:01,580 --> 00:39:06,139
That by this picture, you would expect that the nuclei would be happier and happier and happier as they

342
00:39:06,140 --> 00:39:10,160
got closer and closer together and everything would go down to zero distance.

343
00:39:10,400 --> 00:39:14,720
And that's not true, because some of our approximations start to break down.

344
00:39:15,020 --> 00:39:21,499
One of our approximations that this starts to break down is that we assumed orthogonal orbitals that simplifying assumption,

345
00:39:21,500 --> 00:39:25,280
that that assumption that we started with another. So that's going to break down.

346
00:39:25,280 --> 00:39:29,030
Another thing that breaks down is we forgot about the nuclear nuclear interaction,

347
00:39:29,240 --> 00:39:35,510
which is more or less similar to the covalent bond case, is more or less cancelled by the direct interaction.

348
00:39:35,690 --> 00:39:41,590
But when the nuclei start to get very close to each other, that cancellation is no good anymore and the nuclei start to repel.

349
00:39:41,600 --> 00:39:47,360
So they're not going to get infinitely close anymore. But roughly, this is what causes the bonding in metals.

350
00:39:47,360 --> 00:39:54,050
It's allowing the the electrons to spread out over many, over many atoms and lower their kinetic energy.

351
00:39:54,440 --> 00:39:57,950
Okay. Happy. Okay, good.

352
00:39:58,370 --> 00:40:03,710
So. We can now consider a more general case of dive in one's atoms.

353
00:40:04,460 --> 00:40:07,790
Die valence atoms? I guess not more general.

354
00:40:07,790 --> 00:40:11,300
This is just a different case. Diving into atoms like maybe helium.

355
00:40:11,750 --> 00:40:21,460
Helium has two electrons. In this case we have two and electrons and two and eigen states we can fill.

356
00:40:23,450 --> 00:40:27,580
So we get an entirely filled down. Filled then.

357
00:40:30,110 --> 00:40:32,480
Well, okay, what happens then. Let's go back to this picture.

358
00:40:34,070 --> 00:40:41,600
So now with the DI valent material, we completely fill this band and we completely fill like this.

359
00:40:41,750 --> 00:40:47,690
Everything is filled. In this case, the fill band has is very boring.

360
00:40:47,690 --> 00:40:51,560
The fill band is inert, filled and maybe inert.

361
00:40:54,080 --> 00:41:01,500
Why is it a nerd? Well. When you think about can absorb, can, is it possible for the filled band to absorb any energy?

362
00:41:01,770 --> 00:41:05,060
And the answer is no. It can't absorb any energy because all the states are already filled.

363
00:41:05,070 --> 00:41:10,380
You can't transfer an electron back and forth from one state to another in order to give it energy or take it away.

364
00:41:10,590 --> 00:41:15,810
All the states are filled is a unique state. There is nothing that can be moved around, cannot carry current.

365
00:41:16,020 --> 00:41:20,310
It can't carry current because you can't change the number of left movers versus right movers.

366
00:41:20,550 --> 00:41:24,180
They're all filled. So it has no heat capacity.

367
00:41:24,540 --> 00:41:35,160
No heat capacity carries no current carries, no current.

368
00:41:38,040 --> 00:41:50,879
And it also has no metallic bonding. And this is, again, very similar to what we had in the case of the other heat,

369
00:41:50,880 --> 00:41:56,220
when we consider the possible bonding between this helium atom, between two helium atoms.

370
00:41:56,220 --> 00:42:00,660
If you have to fill the bonding orbital and the anti bonding orbital, you don't gain any energy.

371
00:42:01,050 --> 00:42:08,280
So here, if we have a fill ban, you had to you fill the lower energy states, but you had to fill the higher energy states too.

372
00:42:08,490 --> 00:42:15,000
So it does not gain you anything to have the the hopping go up anymore.

373
00:42:16,110 --> 00:42:19,920
Okay. So how are we doing?

374
00:42:21,060 --> 00:42:29,760
Okay. Good. So as in the case of the vibrational change that we studied earlier this week,

375
00:42:30,000 --> 00:42:35,940
you can also have a situation where not every atom or not every orbital you're considering is the same.

376
00:42:36,240 --> 00:42:45,000
So let's now consider a case where there are two orbitals, two different orbitals per unit cell,

377
00:42:46,020 --> 00:42:51,720
and this can kill can can occur in two different ways or several different ways.

378
00:42:51,750 --> 00:42:57,510
One possible way is that you have two different atoms in the unit cell and they have different types of orbitals in them,

379
00:42:57,510 --> 00:43:00,360
sort of a sodium in a chlorine atom or something like that in a new cell.

380
00:43:00,630 --> 00:43:06,360
Another possible case is you have one atom but two different orbitals on that atom,

381
00:43:06,360 --> 00:43:10,890
like an orbital and a p orbital or a1s orbital and a two orbital that you want to consider.

382
00:43:11,370 --> 00:43:22,680
So without actually solving this problem in in detail, I can show you what the, what the spectrum looks like, what the dispersion looks like.

383
00:43:23,040 --> 00:43:31,650
It's very similar to the case of the vibrational chain. So here's K, here's pi over a year, minus pi over A, here's E.

384
00:43:34,650 --> 00:43:42,270
And you get a low energy band. Analogous to the acoustic mode that we got in the vibrational chain.

385
00:43:42,270 --> 00:43:49,440
And you get a high energy band analogous to the optical mode that we had in the vibrational technology that is supposed to look symmetric.

386
00:43:49,740 --> 00:43:57,690
Apologise about that. So now we have two energy bands, a lower energy band and a higher energy band.

387
00:43:57,930 --> 00:43:59,100
We can also, if we want.

388
00:43:59,100 --> 00:44:12,060
This is the reduced zone scheme, reduced zone scheme, and we can also draw it in the extended zone scheme which which I'll do over here actually.

389
00:44:16,720 --> 00:44:25,100
Yeah. Extended zone scheme. Zone scheme.

390
00:44:26,270 --> 00:44:29,870
We could draw the same thing, which would look kind of like this.

391
00:44:30,770 --> 00:44:35,120
K. E. Here's pi over a.

392
00:44:38,190 --> 00:44:45,540
Here is minus pi over a year to pi over a year minus two pi over a and the

393
00:44:45,540 --> 00:44:54,599
idea of extended zone scheme is to spread out the bands into two boron zones.

394
00:44:54,600 --> 00:45:02,750
So this is the first zone, first busy here and the second is out here and out here.

395
00:45:02,760 --> 00:45:07,410
Second to busy and here. Let's see.

396
00:45:08,170 --> 00:45:15,479
Okay. So all I did was I took. This piece and move over here by two pi over in this piece and moved over here

397
00:45:15,480 --> 00:45:20,520
by two pi over a such at each possible k there's only one one excitation.

398
00:45:22,390 --> 00:45:27,040
And in either case, this is supposed to line up right at the at the brand's own boundaries.

399
00:45:27,040 --> 00:45:30,970
Gap comes in here. This gap comes in here at the bronze zone boundary.

400
00:45:31,450 --> 00:45:35,050
Notice that there's a gap right at the zone boundary.

401
00:45:35,290 --> 00:45:38,800
And it's not coincidental that if you sort of put this together and squint your eyes,

402
00:45:39,100 --> 00:45:43,900
this looks like one overall parabola that a free electron might have,

403
00:45:44,080 --> 00:45:51,160
but you've opened up gaps at the zone boundary, and we'll explain why that is later on in the term.

404
00:45:51,400 --> 00:45:58,000
Now, one thing that you might consider here is suppose suppose the unit sell.

405
00:45:59,860 --> 00:46:05,890
Unit sell has three electrons, has three electrons in it.

406
00:46:06,670 --> 00:46:15,940
Well, then what happens? Then you fill up the entire lower band with both spin up and spin down electrons,

407
00:46:16,240 --> 00:46:23,630
and you have half of the upper band filled to here and half of the upper band here is filled.

408
00:46:24,580 --> 00:46:27,940
Half of the upper bands may fill the lower band and he filled half of the upper band.

409
00:46:28,360 --> 00:46:31,750
Well, you'll notice that we have a entirely filled lower band.

410
00:46:32,050 --> 00:46:35,380
The lower band is inert. Lower band is inert.

411
00:46:38,920 --> 00:46:46,660
And we only need to worry if we're thinking about transport properties or heat properties that, you know, specific heat capacity, something like that.

412
00:46:46,870 --> 00:46:50,290
We only have to worry about the upper band because the lower band is completely inert,

413
00:46:50,470 --> 00:46:55,390
at least unless you give it an enormous energy to excite things out of the lower bound and up to the upper band.

414
00:46:55,720 --> 00:46:58,450
Very high energy because these gaps are typically very large.

415
00:47:00,040 --> 00:47:05,440
You can just ignore the lower band altogether and only worry about the electrons in the upper band.

416
00:47:05,740 --> 00:47:11,800
But this is actually this is actually the answer to one of those questions, one of these puzzles that we wrote down earlier in the term.

417
00:47:12,920 --> 00:47:19,740
And the puzzle from earlier in the term was why is it that you sometimes don't have to worry about core electrons?

418
00:47:19,780 --> 00:47:26,030
If electrons are in core orbitals, why is it you can just throw them away when you're counting electrons for sodium,

419
00:47:26,030 --> 00:47:33,110
you only count one electron per unit per per atom and not 11 electrons per atom when you're trying to calculate the metallic density.

420
00:47:33,350 --> 00:47:39,499
And the reason is because basically those other electrons are completely filling bands and the bands

421
00:47:39,500 --> 00:47:46,370
become inert and just aren't part of the interesting physics like heat capacity and and conduction.

422
00:47:48,290 --> 00:47:54,260
Couple more comments here. So what we've had is that have these important principles.

423
00:47:54,560 --> 00:47:58,080
Is it half filled band? Band is a metal.

424
00:47:58,100 --> 00:48:02,110
Usually a metal. Usually. Medal.

425
00:48:04,430 --> 00:48:11,140
And this is for one electron for unitel or actually any odd number of electrons present itself.

426
00:48:11,800 --> 00:48:16,000
So maybe I'll say odd, odd number of electrons per unit cell,

427
00:48:17,500 --> 00:48:23,560
whereas an even number of electrons buta cell, even number of electrons per unit cell per unit cell.

428
00:48:26,050 --> 00:48:40,600
You might think that this is might or might be be an insulator because you would completely fill bands and then you have just an inert situation.

429
00:48:40,840 --> 00:48:48,190
However, it turns out that there are many, many cases where you have an even number of electrons per unit cell, and yet it's a metal.

430
00:48:48,700 --> 00:48:53,530
And the reason that that can occur is if you have a band structure, it looks as follows.

431
00:48:54,820 --> 00:48:58,790
They have a more complicated band structure that looks like this.

432
00:48:58,840 --> 00:49:03,340
So here's a K, here's pi over a, here's minus pi over a.

433
00:49:04,090 --> 00:49:09,610
And here's the lower band like this. And then imagine that we have an upper band.

434
00:49:10,210 --> 00:49:15,480
It does this. Okay. So it's it's dispersion and more complicated models.

435
00:49:15,480 --> 00:49:22,470
You can have a dispersion that looks like this. In particular, the upper band slips below the top of the lower and the lower band.

436
00:49:22,710 --> 00:49:30,810
So if I have two electrons per unit cell, in fact you could fill the entire lower band.

437
00:49:31,020 --> 00:49:36,270
But it's lower energy. To partially fill this band and partially fill.

438
00:49:37,710 --> 00:49:41,340
This ban. And that would that would make up your two electrons per unit.

439
00:49:41,340 --> 00:49:45,150
So instead of entirely filling the lower band and leaving these states empty.

440
00:49:45,360 --> 00:49:49,590
In that case, you have to partially fill bands instead of one completely filled bands.

441
00:49:49,800 --> 00:49:55,680
And since you have to partially fill bands, you then have low energy excitations here and low energy excitations here.

442
00:49:55,950 --> 00:49:59,790
And so this thing actually becomes a metal. So there are lots of cases of that.

443
00:49:59,820 --> 00:50:05,010
Things like calcium has calcium is a metal, even though it has two electrons per unit itself.

444
00:50:05,160 --> 00:50:09,450
And it's exactly this physics that's occurring. The two bands are actually crossing an energy.

445
00:50:09,630 --> 00:50:13,200
So instead of completely filling one band, you have to partially fill bands.

446
00:50:14,370 --> 00:50:17,760
Okay, I think maybe I better stop there and I see you, I guess tomorrow.