IPhO 1987 · Theory problem 2

Electrons in a magnetic field

A beam of electrons emitted by a point source P enters the magnetic field \(\vec B\) of a toroidal coil (toroid) in the direction of the lines of force. The angle of the aperture of the beam \(2\cdot\alpha_0\) is assumed to be small \((2\cdot\alpha_0\ll1)\). The injection of the electrons occurs on the mean radius R of the toroid with acceleration voltage \(V_0\).

Neglect any interaction between the electrons. The magnitude of \(\vec B\), B, is assumed to be constant.

Figure 1
  1. To guide the electron in the toroidal field a homogeneous magnetic deflection field \(\vec B_1\) is required. Calculate \(\vec B_1\) for an electron moving on a circular orbit of radius R in the torus.

  2. Determine the value of \(\vec B\) which gives four focussing points separated by \(\pi/2\) as indicated in the diagram.

Note: When considering the electron paths you may disregard the curvature of the magnetic field.

  1. The electron beam cannot stay in the toroid without a deflection field \(\vec B_1\), but will leave it with a systematic motion (drift) perpendicular to the plane of the toroid.

a) Show that the radial deviation of the electrons from the injection radius is finite.

b) Determine the direction of the drift velocity.

Note: The angle of aperture of the electron beam can be neglected. Use the laws of conservation of energy and of angular momentum.

Data:

\[ \frac{e}{m}=1.76\cdot10^{11}\ \mathrm{C\cdot kg^{-1}}; \qquad V_0=3\ \mathrm{kV}; \qquad R=50\ \mathrm{mm} \]

Official English text of the problem.

Answers

Let \(e>0\) denote the magnitude of the electron charge. Take \(\hat{\boldsymbol\phi}\) along the injected beam and the toroidal field, and define \(\hat{\mathbf z}=\hat{\mathbf r}\times\hat{\boldsymbol\phi}\), pointing out of the page for the injection shown.

  • 1. The homogeneous deflection field is perpendicular to the toroid: \[ \boxed{\vec B_1=\frac{mv_0}{eR}\hat{\mathbf z} =3.69\times10^{-3}\ \mathrm T\,\hat{\mathbf z}}, \qquad v_0=\sqrt{\frac{2eV_0}{m}}=3.25\times10^7\ \mathrm{m\,s^{-1}}. \]

  • 2. In the prescribed approximation, \[ \boxed{B=\frac{4mv_0}{eR}=4B_1 =1.48\times10^{-2}\ \mathrm T}. \]

  • 3(a). The radial deviation is bounded: \[ \boxed{0\le r-R\le\frac{mv_0}{eB}}. \] With the field from part 2, this bound is \(12.5\ \mathrm{mm}\). The actual maximum deviation in the constant-\(B\) model is \(5.37\ \mathrm{mm}\).

  • 3(b). The drift is into the page, opposite to the deflection field required in part 1: \[ \boxed{\vec v_{\mathrm{drift}}\parallel-\hat{\mathbf z}}. \]

Solution

1. Deflection field

Using nonrelativistic mechanics, acceleration through the voltage \(V_0\) gives \[ \frac12mv_0^2=eV_0, \] hence \[ v_0=\sqrt{2\left(\frac em\right)V_0} =\sqrt{2(1.76\times10^{11})(3.00\times10^3)} =3.25\times10^7\ \mathrm{m\,s^{-1}}. \]

For the reference electron on the circular orbit, the velocity is parallel to the toroidal field. Consequently, that field produces no force on this electron.

The deflection field must supply the inward centripetal force. For \(\vec B_1=B_1\hat{\mathbf z}\), the electron’s Lorentz force is \[ \vec F=-e\vec v\times\vec B_1 =-ev_0B_1\hat{\mathbf r}, \] which points inward everywhere around the circle. Thus \[ ev_0B_1=\frac{mv_0^2}{R}. \]

Therefore \[ \boxed{\vec B_1=\frac{mv_0}{eR}\hat{\mathbf z}}, \] with magnitude \[ B_1= \frac{3.2496\times10^7} {(1.76\times10^{11})(0.050)} =3.69\times10^{-3}\ \mathrm T. \]

The required field is homogeneous and points out of the plane of the diagram.

2. Four focussing points

For this part, neglect the curvature of the guiding field as instructed. Locally, an electron follows a helix about the field direction. If its initial angle to the field is \(\alpha\), then \[ v_\parallel=v_0\cos\alpha\simeq v_0, \qquad v_\perp=v_0\sin\alpha\simeq v_0\alpha. \]

The transverse motion has cyclotron angular frequency and period \[ \omega=\frac{eB}{m}, \qquad T=\frac{2\pi}{\omega}=\frac{2\pi m}{eB}. \] Although the gyration radius depends on \(\alpha\), the period does not. After one complete gyration, every electron returns to its initial transverse position.

The distance along the beam to the next focus is therefore, to first order in the small aperture angle, \[ \ell=v_\parallel T\simeq v_0T =\frac{2\pi mv_0}{eB}. \] Differences in \(v_\parallel\) enter only at order \(\alpha^2\), which is neglected.

Four equally spaced focussing points require \[ \ell=\frac{2\pi R}{4}=\frac{\pi R}{2}. \] Equating these distances gives \[ \frac{2\pi mv_0}{eB}=\frac{\pi R}{2}, \] so \[ \boxed{B=\frac{4mv_0}{eR}=4B_1 =1.48\times10^{-2}\ \mathrm T}. \]

The successive refocussings occur at azimuthal advances \(\pi/2,\pi,3\pi/2,2\pi\); the last coincides with P. Thus there are four distinct focussing positions around the toroid, including P.

3(a). Finite radial deviation without the deflection field

Now remove \(\vec B_1\) and neglect the beam aperture. Use cylindrical coordinates about the toroid’s symmetry axis: \[ \vec B=B\hat{\boldsymbol\phi}, \qquad \vec v=\dot r\,\hat{\mathbf r} +r\dot\phi\,\hat{\boldsymbol\phi} +\dot z\,\hat{\mathbf z}. \] Initially, \[ r=R,\qquad \dot r=0,\qquad \dot z=0,\qquad r\dot\phi=v_0. \]

A magnetic field does no work, so the speed remains constant: \[ \dot r^{\,2}+r^2\dot\phi^{\,2}+\dot z^{\,2}=v_0^2. \tag{1} \]

The magnetic force has no azimuthal component. Therefore angular momentum about the symmetry axis is conserved: \[ mr^2\dot\phi=mRv_0, \qquad r\dot\phi=\frac{Rv_0}{r}. \tag{2} \] Since the azimuthal speed cannot exceed the total speed, equation (2) immediately implies \[ r\ge R. \tag{3} \]

To obtain an upper bound, use the axial component of the Lorentz force: \[ -e\vec v\times\vec B =eB\dot z\,\hat{\mathbf r} -eB\dot r\,\hat{\mathbf z}. \] Thus \[ m\ddot z=-eB\dot r. \] Because \(B\) is constant, integration with the initial conditions gives \[ \dot z=-\frac{eB}{m}(r-R)=-\omega(r-R). \tag{4} \]

Substituting equations (2) and (4) into energy conservation yields \[ \boxed{ \dot r^{\,2} +\frac{R^2v_0^2}{r^2} +\omega^2(r-R)^2 =v_0^2. } \tag{5} \] Every term on the left is nonnegative. Consequently, \[ \omega^2(r-R)^2\le v_0^2. \] Together with equation (3), this proves \[ \boxed{0\le r-R\le\frac{v_0}{\omega} =\frac{mv_0}{eB}}. \]

For the field found in part 2, \(\omega=4v_0/R\), giving the finite bound \[ r-R\le\frac R4=12.5\ \mathrm{mm}. \]

Equation (5) also gives the actual outer turning point. Set \(\dot r=0\) and write \(x=(r-R)/R\). Then \[ 16x^2+\frac{1}{(1+x)^2}=1. \] Besides the initial turning point \(x=0\), the positive root satisfies \[ 16x(1+x)^2=x+2, \] giving \[ x=0.107403, \qquad \boxed{r_{\max}-R=5.37\ \mathrm{mm}}. \]

3(b). Direction of the drift

Equations (3) and (4) give \[ \dot z=-\omega(r-R)\le0. \] Thus the axial velocity never points in the \(+\hat{\mathbf z}\) direction.

The motion actually leaves \(r=R\): initially the magnetic force vanishes because the velocity is parallel to \(\vec B\), and the radial equation gives \[ \ddot r(0)=\frac{v_0^2}{R}>0. \] Hence \(r>R\) shortly after injection and \(\dot z<0\). As the radial motion oscillates between its turning points, the axial displacement continues to accumulate in the negative direction: \[ \left\langle\dot z\right\rangle =-\omega\left\langle r-R\right\rangle<0. \]

Therefore the systematic drift is \[ \boxed{\text{into the page, opposite to }\vec B_1.} \]

The solution was checked against the official solution.