diff options
author | Prefetch | 2022-10-14 23:25:28 +0200 |
---|---|---|
committer | Prefetch | 2022-10-14 23:25:28 +0200 |
commit | 6ce0bb9a8f9fd7d169cbb414a9537d68c5290aae (patch) | |
tree | a0abb6b22f77c0e84ed38277d14662412ce14f39 /source/know/concept/debye-length |
Initial commit after migration from Hugo
Diffstat (limited to 'source/know/concept/debye-length')
-rw-r--r-- | source/know/concept/debye-length/index.md | 150 |
1 files changed, 150 insertions, 0 deletions
diff --git a/source/know/concept/debye-length/index.md b/source/know/concept/debye-length/index.md new file mode 100644 index 0000000..9b87585 --- /dev/null +++ b/source/know/concept/debye-length/index.md @@ -0,0 +1,150 @@ +--- +title: "Debye length" +date: 2021-10-18 +categories: +- Physics +- Plasma physics +layout: "concept" +--- + +If a charged object is put in a plasma, +it repels like charges and attracts opposite charges, +leading to a **Debye sheath** around the object's surface +with a net opposite charge. +This has the effect of **shielding** the object's presence +from the rest of the plasma. + +We start from [Gauss' law](/know/concept/maxwells-equations/) +for the [electric field](/know/concept/electric-field/) $\vb{E}$, +expressing $\vb{E}$ as the gradient of a potential $\phi$, +i.e. $\vb{E} = -\nabla \phi$, +and splitting the charge density into ions $n_i$ and electrons $n_e$: + +$$\begin{aligned} + \nabla^2 \phi(\vb{r}) + = - \frac{1}{\varepsilon_0} \Big( q_i n_i(\vb{r}) + q_e n_e(\vb{r}) + q_t \delta(\vb{r}) \Big) +\end{aligned}$$ + +The last term represents a *test particle*, +which will be shielded. +This particle is a point charge $q_t$, +whose density is simply a [Dirac delta function](/know/concept/dirac-delta-function/) $\delta(\vb{r})$, +and is not included in $n_i$ or $n_e$. + +For a plasma in thermal equilibrium, +we have the [Boltzmann relations](/know/concept/boltzmann-relation/) +for the densities: + +$$\begin{aligned} + n_i(\vb{r}) + = n_{i0} \exp\!\bigg( \!-\! \frac{q_i \phi(\vb{r})}{k_B T_i} \bigg) + \qquad \quad + n_e(\vb{r}) + = n_{e0} \exp\!\bigg( \!-\! \frac{q_e \phi(\vb{r})}{k_B T_e} \bigg) +\end{aligned}$$ + +We assume that electrical interactions are weak compared to thermal effects, +i.e. $k_B T \gg q \phi$ in both cases. +Then we Taylor-expand the Boltzmann relations to first order: + +$$\begin{aligned} + n_i(\vb{r}) + \approx n_{i0} \bigg( 1 - \frac{q_i \phi(\vb{r})}{k_B T_i} \bigg) + \qquad \quad + n_e(\vb{r}) + \approx n_{e0} \bigg( 1 - \frac{q_e \phi(\vb{r})}{k_B T_e} \bigg) +\end{aligned}$$ + +Inserting this back into Gauss' law, +we arrive at the following equation for $\phi(\vb{r})$, +where we have assumed quasi-neutrality such that $q_i n_{i0} = q_e n_{e0}$: + +$$\begin{aligned} + \nabla^2 \phi + &= - \frac{1}{\varepsilon_0} + \bigg( q_i n_{i0} - n_{i0} \frac{q_i^2 \phi}{k_B T_i} + q_e n_{e0} - n_{e0} \frac{q_e^2 \phi}{k_B T_e} + q_t \delta(\vb{r}) \bigg) + \\ + &= \bigg( \frac{n_{i0} q_i^2}{\varepsilon_0 k_B T_i} + \frac{n_{e0} q_e^2}{\varepsilon_0 k_B T_e} \bigg) \phi + - \frac{q_t}{\varepsilon_0} \delta(\vb{r}) +\end{aligned}$$ + +We now define the **ion** and **electron Debye lengths** +$\lambda_{Di}$ and $\lambda_{De}$ as follows: + +$$\begin{aligned} + \boxed{ + \frac{1}{\lambda_{Di}^2} + \equiv \frac{n_{i0} q_i^2}{\varepsilon_0 k_B T_i} + } + \qquad \quad + \boxed{ + \frac{1}{\lambda_{De}^2} + \equiv \frac{n_{e0} q_e^2}{\varepsilon_0 k_B T_e} + } +\end{aligned}$$ + +And then the **total Debye length** $\lambda_D$ is defined as the sum of their inverses, +and gives the rough thickness of the Debye sheath: + +$$\begin{aligned} + \boxed{ + \frac{1}{\lambda_D^2} + \equiv \frac{1}{\lambda_{Di}^2} + \frac{1}{\lambda_{De}^2} + = \frac{n_{i0} q_i^2 T_e + n_{e0} q_e^2 T_i}{\varepsilon_0 k_B T_i T_e} + } +\end{aligned}$$ + +With this, the equation can be put in the form below, +suggesting exponential decay: + +$$\begin{aligned} + \nabla^2 \phi(\vb{r}) + &= \frac{1}{\lambda_D^2} \phi(\vb{r}) + - \frac{q_t}{\varepsilon_0} \delta(\vb{r}) +\end{aligned}$$ + +This has the following solution, +known as the **Yukawa potential**, +which decays exponentially, +representing the plasma's **self-shielding** +over a characteristic distance $\lambda_D$: + +$$\begin{aligned} + \boxed{ + \phi(r) + = \frac{q_t}{4 \pi \varepsilon_0 r} \exp\!\Big( \!-\!\frac{r}{\lambda_D} \Big) + } +\end{aligned}$$ + +Note that $r$ is a scalar, +i.e. the potential depends only on the radial distance to $q_t$. +This treatment only makes sense +if the plasma is sufficiently dense, +such that there is a large number of particles +in a sphere with radius $\lambda_D$. +This corresponds to a large [Coulomb logarithm](/know/concept/coulomb-logarithm/) $\ln\!(\Lambda)$: + +$$\begin{aligned} + 1 \ll \frac{4 \pi}{3} n_0 \lambda_D^3 = \frac{2}{9} \Lambda +\end{aligned}$$ + +The name *Yukawa potential* originates from particle physics, +but can in general be used to refer to any potential (electric or energetic) +of the following form: + +$$\begin{aligned} + V(r) + = \frac{A}{r} \exp(-B r) +\end{aligned}$$ + +Where $A$ and $B$ are scaling constants that depend on the problem at hand. + + + +## References +1. P.M. Bellan, + *Fundamentals of plasma physics*, + 1st edition, Cambridge. +2. M. Salewski, A.H. Nielsen, + *Plasma physics: lecture notes*, + 2021, unpublished. |