Determining n and : The Hall Effect

V
x
, E
x
- - - - - - - - -
+ + + + + + + + + + +
E
y
I, J
x
B
z
F
r
=qE
r
+qv
r
B
r
In steady state,
F
y
= ev
D
B
z
E
Y
=v
D
B
Z
=E
H
, the Hall Field
F
y
= eE
y
Since v
D
=-J
x
/en,
1
E
H
= J
x
B
Z
=R
H
J
X
B
Z
ne
1
R
H
=
=ne
ne
Experimental Hall Results on Metals
Valence=1 metals look like
free-electron Drude metals
Valence=2 and 3,
magnitude and sign
suggest problems
Metal Valence -1/R
H
nec
Li
Na
K
Rb
Cs
Cu
Ag
Au
Be
Mg
In
Al
1
1
1
1
1
1
1
1
2
2
3
3
1.2
1.1
1.0
0.9
1.5
1.3
1.5
0.2
-0.4
-0.3
0.3
0.8
Hall coefficients of selected elements in moderate to high fields*
* These are roughly the limiting values assumed by R
n
as the field becomes
very large (of order 10
4
G), and the temperature very low, in carefully
prepared specimens. The data are quoted in the form n
0
:n. Where n
0
is the
density for which the Drude form (1.21) agrees with the measured
R
u
:n
0
= -1/R
H
ec. Evidently the alkali metals obey the Drude result
reasonably well, the noble metals (Cu. Ag, Au) less well, and the remaining
entries, not at all.
Table by MIT OpenCourseWare.
Response of free e- to AC Electric
Fields
Microscopic picture
E =E e
it
Z O
e
-
B=0 in conductor,
dp(t) p(t)
it
r r r r
= eE
0
e
and
F(E) >>F(B)
dt
try p(t) = p
0
e
it
ip
0
=
p
0
eE
0

eE
>>1/, p out of phase with E
p =
0
eE
0
0
1
i
p
0
=
i
, p 0

<<1/, p in phase with E

p
0
= eE
0

What if >>1?
When will J = E break down? It depends on electrons undergoing many
collisions, on the average a collision time apart. As long as there are many
collisions per cycle of the AC field (<<1), the AC will be the DC .
x
x x
x
x x
x
x
x
x
x
x
x
x
x x x x
x
x
x
But consider the other limit: >>1.
Now there will be many cycles of the field between collisions. In this limit, the
electrons will behave more like electrons in vacuum, and the relation between J
and E will be different x x
Complex Representation of Waves
sin(kx-t), cos(kx-t), and e
-i(kx-t)
are all waves
e
-i(kx-t)
is the complex one and is the most general
imaginary
A
iAsin
Acos
real

e
i
=cos+isin
Response of e- to AC Electric Fields
Momentum represented in the complex plane
real
imaginary
p
p ( <<1/ )
p

(

>
>
1
/

)

Instead of a complex momentum, we can go back to macroscopic
and create a complex J and
i
nep
0
ne
2
J (t) =J
0
e
J
0
= nev =
m
=
1
E
0
m( i)

0
, =
ne
2

1i
0
m
Response of e- to AC Electric Fields
Low frequency (<<1/)
electron has many collisions before
direction change
Ohms Law: J follows E, real
High frequency (>>1/)
electron has nearly 1 collision or less
when direction is changed
J imaginary and 90 degrees out of phase
with E, is imaginary
Qualitatively:
<<1, electrons in phase, re-irradiate, E
i
=E
r
+E
t
, reflection
>>1, electrons out of phase, electrons too slow, less interaction,transmission =
r

r
=1
10
14
sec, =c, =
3x10
10
cm / sec
10
14
Hz
5000x10
8
cm
E-fields with frequencies greater than visible light frequency expected to be
beyond influence of free electrons
Response of light to interaction with
material
Need Maxwells equations
from experiments: Gauss, Faraday, Amperes laws
second term in Amperes is from the unification
electromagnetic waves!
SI Units (MKS) Gaussian Units (CGS)
r
r
D =
D =4
r
r
B =0
B =0
r
xE
r
=
B
r
xE
r
=
1
c

B
t
t r
r r
4
r
1 D
r r
D xH = J +
xH =J + c c t
r r r
t
r r r r D =E +4P
D =
0
E +P = E
r r r
r r r r
B =H +4M
B =
0
H +
0
M = H
=
r

0
; =
r

0
Waves in Materials
Non-magnetic material, =
0
Polarization non-existent or swamped by free electrons,
P=0
r
xE
r
=
B
For a typical wave,
t
E =E e
i (k r t )
=E e
ik r
e
it
=E(r)e
it
r
0 0
r r
E
xB =
0
J +
0

0
2
E(r) = i
0
E(r)
0

2
E(r)
t
r

2
E(r) =

2
2
()E(r)
Wave Equation
r
xB
c x(xE) =
t
i
() =1+

2
E =

[
0
J +
0

0
E
]
t t
E(r) =E
0
e
ik r

2
E =
0

E
+
0

2
E
t t
2
k
2
=
2
()
c
c
v = =
k
()
Waves in Materials
Waves slow down in materials (depends on ())
Wavelength decreases (depends on ())
Frequency dependence in ()
() =1+
i
=1+
i
0

0

0
(1i )
i
2

() =1+
p
2
i
2

2
p
=
ne
Plasma Frequency

0
m
For >>>1, ()goes to 1
For an excellent conductor (
0
large), ignore 1, look at case for <<1
i
2
i
2

()
p
2

p
i
Waves in Materials

0
k = () = i
c c
0

0

0

1+i
0
=

+i k =
2

2
0
c
2
2
0
c
2

For a wave
E =E
0
e
i (kzt )
Let k=k
real
+k
imaginary
=k
r
+ik
i
E =E
0
e
i[k
r
zt ]
e
k
i
z
The skin depth can be defined by
2 c
2
1 2
o
= = =
k
i o o o

Waves in Materials
For a material with any
0
, look at case for >>1

2
<
p
, is negative, k=k
i
, wave reflected
( ) =1
2

p

>
p
, is positive, k=k
r
, wave propagates
R

p
Success and Failure of Free e- Picture
Success
K/=thermal conduct./electrical conduct.~CT
Metal conductivity
1
2
Wiedmann-Franz law

v
2 m
Hall effect valence=1
=
3
c
v
v
therm

3
2
3k
b
T
Skin Depth
c
v
=

= nk
b
; v
therm
=
Examples of Failure
=
1

3
nk
b

3k
b
T

=
3 nk
b
2
T
Insulators, Semiconductors
3

2 m
Hall effect valence>1
ne
2

=
Thermoelectric effect
m
Colors of metals
3
k
b

2
Therefore :

=
2

T
Luck: c
vreal
=c
vclass
/100; ~C!
v
real
2
=v
class
2
*100
Wiedmann-Franz Success
Thermoelectric Effect
Exposed Failure when
c and v
2
are not both
v
in property
E =QT
3
c
v

2
nk
b
nk
b
Thermopower Q is
Q = = =
3ne 3ne 2e
Thermopower is about 100 times too large!
Element
k
(watt cm-K)
k
(watt cm-K)
k T
(watt-ohm K
2
)
k T
(watt-ohm K
2
)
Li
Na
K
Rb
Cu
Ag
Au
Be
Mg
0.71
1.38
1.0
0.6
3.85
4.18
3.1
2.3
2.38
1.5
0.52
0.80
1.13
1.0
0.88
0.5
0.64
0.38
0.09
0.18
2.12
2.23
2.42
2.20
2.31
2.32
2.36
2.14
2.14
2.90
2.61
2.28
2.49
2.58
2.75
2.48
2.64
3.53
2.57
Nb
Fe
Zn
Cd
Al
In
Tl
Sn
Pb
Bi
Sb
2.22 x 10
8
3.82
4.17
3.1
1.7
2.30
1.5
0.54
0.73
1.1
1.0
0.80
0.45
0.60
0.35
0.08
0.17
0.73
2.29
2.38
2.36
2.42
2.19
2.25
2.78
2.88
2.30
2.60
2.75
2.54
2.53
3.35
2.69
2.43 x 10
8
273K 373K
Experimental thermal conductivities and Lorenz numbers of selected metals
Table by MIT OpenCourseWare.