Electrons escaping a metal surface

Click For Summary
In low temperature conditions, the Fermi-Dirac distribution approximates the number density of electrons, and the current density from escaping electrons is calculated using their momentum. The discussion focuses on deriving the escape rate of electrons from a metal surface, emphasizing the relationship between momentum and energy. An integral is set up to calculate the escape rate per unit area, considering only electrons with sufficient momentum to escape. The factor of 2 accounts for electron spin, allowing two electrons per momentum state. The calculations and setup presented are confirmed as correct.
ergospherical
Science Advisor
Homework Helper
Education Advisor
Insights Author
Messages
1,100
Reaction score
1,387
Homework Statement
Electrons in a semi-infinite slab (z < 0) of metal behave as an ideal non-relativistic Fermi gas. They escape the surface if ##p_z^2/(2m) > E_F + V##, where ##E_F## is the Fermi energy and ##V## is a potential barrier - what is the current density of escaping electrons? Assume ##E_F \gg k_B T## and ##V \gg k_B T##.
Relevant Equations
N/A
In the low temperature limit ##\mu \approx E_F## and the Fermi-Dirac distribution is ##n(E) \approx g(E)/(e^{\beta(E-E_F)}+1)##. An escaping electron contributes ##\Delta j_z = -ev_z = -ep_z/m## to the current density. How can I calculate the rate that electrons escape at? I can't see how to relate ##p_z## to the Fermi-Dirac distribution (apart from ##E = p^2/(2m) = (p_x^2 + p_y^2 + p_z^2)/(2m)##, in which case I don't know what to say about the transverse component of the momentum).
 
Physics news on Phys.org
ergospherical said:
How can I calculate the rate that electrons escape at?

##g(E) \large \frac{dE}{e^{\beta(E-E_F)}+1}## gives the number density of electrons with energy between ##E## and ##E + dE##.

In terms of momentum, verify that the number density of electrons with momentum in the range ##(p_x, p_y, p_z)## to ##(p_x+dp_x, p_y+dp_y, p_z + dp_z)## is $$\frac{2}{h^3} \frac{dp_x dp_y dp_z}{e^{\beta[(p_x^2+p_y^2+p_z^2)/(2m)-E_F]}+1}$$ Use this to set up an integral that gives the rate ##R## at which electrons will escape from a unit area of the surface of the metal. $$R = \int_{??}^\infty dp_z \int_{-\infty}^\infty dp_y\int_{-\infty}^\infty dp_x \rm {\,[\, integrand \,\, left \,\, for \,\, you \, :) \,]}$$
 
  • Like
Likes ergospherical
Cheers. Where does the factor 2 come from? I figured that if the (momentum) phase space volume of a state is ##h^3##, then ##g(E) dE \sim d^3 p / h^3##.

Then I thought about a section of the metal with small surface area ##dA##. In time ##dt##, electrons with velocities between ##v_z## and ##v_z + dv_z## reach the surface if they are within a depth ##v_z dt##, i.e. within a volume ##(p_z/m) dt dA##. There are ##dn (p_z/m) dt dA## such electrons, where ##dn = dn(p_x,p_y,p_z)## is the number density of electrons given above. So the rate of escape per unit area, given the restriction that only ##p_z > \sqrt{2m(E_F + V)}## can escape, is \begin{align*}
R = \frac{2}{h^3 m} \int_{\sqrt{2m(E_F + V)}}^{\infty} dp_z \int_{-\infty}^{\infty} dp_y \int_{-\infty}^{\infty} dp_x \ \frac{p_z}{e^{\beta(p^2/(2m) - E_F)} + 1}
\end{align*}Does that look right to you?
 
Last edited:
ergospherical said:
Cheers. Where does the factor 2 come from? I figured that if the (momentum) phase space volume of a state is ##h^3##, then ##g(E) dE \sim d^3 p / h^3##.
Electron spin allows two electrons to be in each momentum state. This is easy to forget.

ergospherical said:
Then I thought about a section of the metal with small surface area ##dA##. In time ##dt##, electrons with velocities between ##v_z## and ##v_z + dv_z## reach the surface if they are within a depth ##v_z dt##, i.e. within a volume ##(p_z/m) dt dA##. There are ##dn (p_z/m) dt dA## such electrons, where ##dn = dn(p_x,p_y,p_z)## is the number density of electrons given above. So the rate of escape per unit area, given the restriction that only ##p_z > \sqrt{2m(E_F + V)}## can escape, is \begin{align*}
R = \frac{2}{h^3 m} \int_{\sqrt{2m(E_F + V)}}^{\infty} dp_z \int_{-\infty}^{\infty} dp_y \int_{-\infty}^{\infty} dp_x \ \frac{p_z}{e^{\beta(p^2/(2m) - E_F)} + 1}
\end{align*}Does that look right to you?
Yes. Very nice.
 
So is there some elegant way to do this or am I just supposed to follow my nose and sub the Taylor expansions for terms in the two boost matrices under the assumption ##v,w\ll 1##, then do three ugly matrix multiplications and get some horrifying kludge for ##R## and show that the product of ##R## and its transpose is the identity matrix with det(R)=1? Without loss of generality I made ##\mathbf{v}## point along the x-axis and since ##\mathbf{v}\cdot\mathbf{w} = 0## I set ##w_1 = 0## to...

Similar threads

  • · Replies 11 ·
Replies
11
Views
2K
  • · Replies 5 ·
Replies
5
Views
3K
Replies
4
Views
2K
  • · Replies 1 ·
Replies
1
Views
4K
Replies
1
Views
4K
  • · Replies 1 ·
Replies
1
Views
2K
Replies
1
Views
1K
  • · Replies 3 ·
Replies
3
Views
2K
  • · Replies 1 ·
Replies
1
Views
2K
  • · Replies 6 ·
Replies
6
Views
3K