1: \documentclass[11pt]{article}
2: \newcommand{\bmu}{\mbox{\boldmath$\mu$}}
3: \newcommand{\balpha}{\mbox{\boldmath$\alpha$}}
4: \newcommand{\bomega}{\mbox{\boldmath$\omega$}}
5: \newcommand{\bsigma}{\mbox{\boldmath$\sigma$}}
6: \newcommand{\bi}[1]{\mbox{\boldmath$#1$}}
7: \usepackage{graphicx}
8: \newcommand{\wfigl}[4]{\begin{figure}[!hbtp]\centering%
9: \includegraphics[width=#2]{#3}\caption{\small{#4}}\label{#1}\end{figure}}
10: \newcommand{\cfigl}[3]{\begin{figure}[!hbtp]\centering%
11: \includegraphics[width=.4\textwidth]{#2}\caption{\small{#3}}\label{#1}\end{figure}}
12:
13: \addtolength{\topmargin}{-1in}\addtolength{\textheight}{4cm}\addtolength{\textwidth}{2.5cm}
14: \addtolength{\evensidemargin}{-0.5in}\addtolength{\oddsidemargin}{-0.5in}
15: %\renewcommand{\baselinestretch}{1.6} %double space
16:
17: \begin{document}
18:
19: %\begin{frontmatter}
20:
21: \title{Are the electron spin and magnetic moment\\ parallel or antiparallel vectors?}
22: \author{{Mart\'{\i}n Rivas}\thanks{wtpripem@lg.ehu.es}\\Dpto. de F\'{\i}sica Te\'orica,\\The University of the Basque Country,\\
23: Apdo.~644, 48080 Bilbao, Spain}
24: \date{\today}
25:
26: \maketitle
27:
28: \begin{abstract}
29: A direct measurement of the relative orientation between the
30: spin and magnetic moment of the electron seems to be never performed.
31: The kinematical theory of elementary particles developed by the author
32: and the analysis of the expectation value of Dirac's magnetic moment operator show that,
33: contrary to the usual assumption,
34: spin and magnetic moment of electrons and positrons might have the same relative orientation.
35: Two plausible experiments for their relative orientation measurement are proposed.
36: \end{abstract}
37: \vspace{1cm}
38: %\begin{keyword}
39: {\sl Keywords}: Semiclassical theories; Electron; Electric and magnetic moments
40:
41: {PACS:} 03.65.S; 14.60.C; 13.40.Em
42: %\end{keyword}
43:
44: %\end{frontmatter}
45: \vspace{2cm}
46: %\newpage
47: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
48: \section{Introduction}
49: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
50:
51: The usual assumption concerning the magnetic dipole structure of the electron,
52: states that if the electron is a spinning particle of negative charge which
53: rotates along the spin direction then
54: this motion will produce a magnetic moment opposite to the spin. In the case of the positron both magnitudes,
55: spin and magnetic moment, will therefore have the same direction.
56: But this interpretation is not supported by a classical analysis of spin, but rather by the
57: guess that presumably spin and angular velocity are directly related.
58:
59: Dirac's analysis \cite{Dirac} of the relativistic electron shows that the spin
60: and magnetic moment operators are related by
61: \begin{equation}
62: \bmu=\frac{q\hbar}{2m}\bsigma,
63: \label{eq:mus}
64: \end{equation}
65: where $q$ is the electric charge and $\bsigma$ the spin matrix operator. For
66: the electron $q=-e$, $e>0$, and therefore the spin
67: and magnetic moment are antiparallel vectors while they are parallel for positrons since
68: $q=+e$.
69: However, if we make this discussion for the expectation values we obtain
70: some indefiniteness in this relative orientation. Let us consider the particle analysed
71: in the center of mass frame. Let
72: \[
73: u_1=\pmatrix{1\cr 0\cr0\cr0\cr},\quad
74: u_2=\pmatrix{0\cr 1\cr0\cr0\cr},\quad
75: v_1=\pmatrix{0\cr 0\cr1\cr0\cr},\quad
76: v_2=\pmatrix{0\cr 0\cr0\cr1\cr},
77: \]
78: the usual Dirac's spinors in this frame. Spinors $u_1$ and $u_2$ are positive
79: energy solutions and $v_1$ and $v_2$ the negative energy ones.
80: The $\sigma_z$ operator takes the form
81: \[
82: \sigma_z=\pmatrix{1&0&0&0\cr 0&-1&0&0\cr 0&0&1&0\cr 0&0&0&-1\cr},
83: \]
84: so that $u_1$ and $v_1$ are spin up states and $u_2$ and $v_2$ spin down.
85: If we take the expectation value of $\mu_z$ in all these states we obtain that
86: for positive and negative energy states the magnetic moment has the opposite orientation
87: to the corresponding expectation value of the spin.
88: But if the negative energy states are considered to describe the antiparticle states
89: then particle and antiparticle have the same relative orientation of spin
90: and magnetic moment.
91:
92: Now, let $C=i\gamma_2\gamma_0$ the charge conjugate operator. In
93: the Pauli-Dirac representation it takes the form
94: \[
95: C=\pmatrix{0&0&0&-1\cr 0&0&1&0\cr 0&-1&0&0\cr 1&0&0&0\cr},
96: \]
97: and the charge conjugate spinors $\tilde{u}_i=Cu_i$ and $\tilde{v}_i=Cv_i$ are given by
98: \[
99: \tilde{u}_1=\pmatrix{0\cr 0\cr0\cr 1\cr},\quad
100: \tilde{u}_2=\pmatrix{0\cr 0\cr-1\cr0\cr},\quad
101: \tilde{v}_1=\pmatrix{0\cr 1\cr0\cr0\cr},\quad
102: \tilde{v}_2=\pmatrix{-1\cr 0\cr0\cr0\cr},
103: \]
104: so that the $\tilde{v}$ spinors represent positive energy states and $\tilde{u}$ negative
105: energy states of a system which satisfies Dirac's equation for a
106: particle of opposite charge. Now $\tilde{u}_2$ and $\tilde{v}_2$
107: are spin up states and the others spin down. But for this system we have to take
108: for the magnetic moment operator the corresponding expression (\ref{eq:mus})
109: with $q=+e$ and therefore
110: the expectation values show that for positive and negative energy states of the positron
111: spin and magnetic moment are parallel vectors.
112: Then, although the expression of the magnetic moment
113: operator obtained in Dirac's theory is unambiguous,
114: the analysis of the expectation values leads to some contradiction and
115: in any case it seems that particle and antiparticle
116: may have the same relative orientation between both magnitudes.
117: We shall see in section \ref{sec:clas} that the analysis of a classical model
118: of electron, which satisfies when quantised Dirac's equation, leads to the same indefiniteness
119: in this relative orientation.
120:
121: No explicit direct measurement of the relative orientation
122: between spin and magnetic moment of the free electron,
123: known to the author, can be found in the literature
124: although very high precision experiments
125: are performed to measure the magnitude of the magnetic moment and
126: the absolute value of $g$, the gyromagnetic ratio.
127: In the review article
128: by Rich and Wesley \cite{RichWesley} the two main methods for measuring the anomaly factor
129: of leptons $a=|g|/2-1$, are analysed: one kind involves precession methods which measure
130: the difference between the spin precession
131: frequency and its cyclotron frequency in a uniform magnetic field.
132: The other are resonance experiments, like the ones developed by Dehmelt on a single electron
133: in a Penning trap \cite{Dehmelt} where the cyclotron motion, magnetron motion and axial
134: oscillation are monitored.
135: All these measurements are in fact independent of whether
136: spin and magnetic moment are indeed parallel or antiparallel, because they
137: involve measurements of the spin precession
138: frequencies in external magnetic fields.
139:
140: All attempts of Stern-Gerlach type on unpolarised beams to separate
141: electrons in inhomogeneous magnetic fields have failed
142: and Bohr and Pauli claimed that this failure was
143: a consequence of the Lorentz force on charged particles
144: which blurred the splitting. Nevertheless
145: Seattle experiments on a single electron \cite{Dehmelt2}
146: show a ``continuous Stern-Gerlach type'' of interaction producing an ``axial'' oscillation
147: of the particle in the direction orthogonal to its cyclotron motion.
148: But these experiments are not able to determine the relative orientation
149: between these magnitudes. Batelaan et al. \cite{Batelaan}, propose an alternative
150: device which according to Dehmelt's suggestions minimize
151: the Lorentz force by using an external magnetic field along the electron velocity
152: but maximize the spin force by using large magnetic field gradients in that direction.
153: They obtain numerically a polarization of the electron beam
154: along the direction of motion. Perhaps an experimental setup in these terms will be able to clarify
155: the relative orientation between these two vector magnitudes.
156:
157: After giving in the next section a short review to the kinematical formalism of
158: spinning particles
159: developed by the author, in section {\ref{sec:directexperi}} a
160: direct experiment and in \ref{sec:indirexper} an indirect
161: experiment will be suggested to check the relative orientation
162: of spin and magnetic moment.
163:
164: %\newpage
165: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
166: \section{Clasical spinning particles}
167: \label{sec:clas}
168: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
169:
170: The kinematical theory of elementary spinning particles \cite{Rivas1}
171: produces a classical description of spin and an elementary
172: particle in this formalism is a pointlike object. In this point we locate the charge of the particle
173: and its motion can be interpreted as a translational motion of its center of mass
174: and a harmonic motion of the center of charge around the center of mass.
175: Once the spin direction is fixed, the motion
176: of the point charge is completely determined.
177: If we consider as {\em the particle} the positive energy
178: solution and of negative electric charge,
179: then the spin and magnetic moment for
180: both the electron and positron are
181: described by parallel vectors. If we consider that the particle has positive
182: charge we get the opposite orientation. We thus obtain the same indefiniteness
183: as in the previous analysis of the expectation value of Dirac's magnetic moment operator.
184:
185: Let us review the main highlights of the mentioned approach:
186: \begin{itemize}
187: \item The classical variables that characterise the initial and final state
188: of a classical elementary spinning particle in a Lagrangian approach are precisely
189: the variables used as parameters of the kinematical group of space-time symmetries
190: or of any of its homogeneous spaces.
191: Any element of the Poincar\'e group can be parametrised in terms of the time and
192: space translation and the relative velocity and orientation
193: among inertial observers. Therefore,
194: a relativistic spinning particle is described by the variables
195: time $t$, position ${\bi r}$, velocity ${\bi v}$ and orientation $\balpha$.
196: We shall call to these variables the kinematical variables and the manifold they span
197: the kinematical space of the system.
198:
199: \item A classical spinning particle is thus described as a point with orientation.
200: The particle moves and rotates in space. Point ${\bi r}$ describes its position in space
201: while $\balpha$ describes its spatial orientation.
202: But what point ${\bi r}$ describes is the position of the charge,
203: which is in general a different
204: point than its center of mass ${\bi q}$, and in general
205: ${\bi r}$ describes a harmonic motion around ${\bi q}$,
206: usually called this motion the {\em zitterbewegung}.
207:
208: \item When expressed the Lagrangian in terms of the kinematical variables
209: it becomes a homogeneous function of first degree
210: of the derivatives of the kinematical variables and consequently
211: it also depends on $\dot{\bi v}$, the acceleration of point ${\bi r}$, and on
212: $\dot{\balpha}$ or equivalently on the angular
213: velocity $\bomega$. It turns out that it can be written as
214: \begin{equation}
215: L=T\dot{t}+{\bi R}\cdot\dot{\bi r}+{\bi V}\cdot\dot{\bi v}+{\bi W}\cdot\bomega,
216: \label{eq:L}
217: \end{equation}
218: where $T=\partial L/\partial\dot{t}$, ${\bi R}=\partial L/\partial\dot{\bi r}$,
219: ${\bi V}=\partial L/\partial\dot{\bi v}$ and ${\bi W}=\partial L/\partial\bomega$.
220:
221: \item For a free relativistic particle, when analyzing the
222: invariance under the different one-parameter subgroups of the Poincar\'e group,
223: Noether's theorem
224: determines the usual constants of the motion which take the following form
225: in terms of the above magnitudes: Energy,
226: \[
227: H=-T-{\bi v}\cdot\frac{d{\bi V}}{dt},
228: \]
229: linear momentum,
230: \begin{equation}
231: {\bi P}={\bi R}-\frac{d{\bi V}}{dt},
232: \label{eq:p}
233: \end{equation}
234: kinematical momentum
235: \begin{equation}
236: {\bi K}=\frac{H}{c^2}{\bi r}-{\bi P}t-\frac{1}{c^2}{\bi S}\times{\bi v},
237: \label{eq:k}
238: \end{equation}
239: and angular momentum
240: \begin{equation}
241: {\bi J}={\bi r}\times{\bi P}+{\bi S},
242: \label{eq:j}
243: \end{equation}
244: where the observable ${\bi S}$, takes the form
245: \begin{equation}
246: {\bi S}={\bi v}\times{\bi V}+{\bi W}.
247: \label{eq:s}
248: \end{equation}
249:
250: \item The linear momentum (\ref{eq:p}) is not lying along the velocity
251: ${\bi v}$ of point ${\bi r}$. Point ${\bi r}$
252: does not represent the center of mass position. If in terms of the last term in (\ref{eq:k})
253: we define the position vector
254: \[
255: {\bi k}=\frac{1}{H}\,{\bi S}\times{\bi v},
256: \]
257: then the center of mass position can be defined as ${\bi q}={\bi r}-{\bi k}$,
258: such that the kinematical momentum can be written as
259: \[
260: {\bi K}=\frac{H}{c^2}{\bi q}-{\bi P}t.
261: \]
262: Taking the time derivative of this expression leads for the linear momentum to
263: \[
264: {\bi P}=\frac{H}{c^2}\,\frac{d{\bi q}}{dt},
265: \]
266: which is the usual relativistic expression of the linear
267: momentum in terms of the center of mass velocity.
268: Observable ${\bi k}$ is the relative position of
269: point ${\bi r}$ with respect to the center of mass ${\bi q}$.
270:
271: \item The observable ${\bi S}$ is the classical equivalent of Dirac's spin operator, because
272: it satisfies the free dynamical equation
273: \[
274: \frac{d{\bi S}}{dt}={\bi P}\times{\bi v},
275: \]
276: which does not vanish because ${\bi P}$ and ${\bi v}$ are not parallel vectors
277: and where the velocity operator in Dirac's theory becomes ${\bi v}=c\balpha$ in
278: terms of Dirac's $\balpha$ matrices.
279:
280: \item The structure of the spin observable is twofold. One, ${\bi v}\times{\bi V}$,
281: is related to the zitterbewegung
282: or charge motion around its center of mass and another ${\bi W}$ to the rotation of the particle.
283:
284: \item The magnetic moment is produced by the charge motion and is thus related
285: only to the zitterbewegung part of the spin. It is because the spin has
286: another contribution coming from the rotation
287: that a pure kinematical interpretation of the gyromagnetic ratio
288: \cite{Rivas3} has been given.
289:
290: \item The classical system that when quantised satisfies Dirac's equation \cite{Rivas2} corresponds
291: to a particle whose charge is moving at the speed of light, and therefore if $v=c$ is constant,
292: the acceleration is always orthogonal to the velocity.
293:
294: \item If we take in (\ref{eq:k}) the time derivative of this expression
295: and afterwards its scalar product with vector ${\bi v}$, since $v=c$ we get the relationship
296: \[
297: H-{\bi P}\cdot{\bi v}-\frac{1}{c^2}\left({\bi S}\times\frac{d{\bi v}}{dt}\right)\cdot{\bi v}=0
298: \]
299: which is the classical equivalent of Dirac's equation.
300:
301: \item The center of mass observer is defined by the conditions ${\bi P}=0$ and ${\bi K}=0$.
302: For this observer we see from (\ref{eq:j})
303: that the spin ${\bi S}$ is a constant of the motion. If the system has positive energy $H=+mc^2$,
304: from (\ref{eq:k}) we get
305: \[
306: m{\bi r}=\frac{1}{c^2}{\bi S}\times{\bi v},
307: \]
308: so that the charge of the particle is describing circles in a plane orthogonal to ${\bi S}$
309: as depicted in part a) of figure \ref{fig:1}.
310: Part b) is the time reversed motion
311: of this particle which corresponds to its antiparticle or to
312: a particle that in the center of mass frame has energy $H=-mc^2$.
313: If the particle is negatively charged then particle and antiparticle have their magnetic moment
314: along the spin direction. If we consider as the particle the positively charged one
315: we obtain the opposite orientation for the magnetic moment. In any case the magnetic moment of the particle and antiparticle
316: have the same relative orientation with respect to the spin.
317:
318: \cfigl{fig:1}{fig1.eps}{Charge motion of the electron a) and positron b) in the center of mass frame.}
319:
320: The radius
321: of this motion is
322: \[
323: R_o=\frac{S}{mc}=\frac{\hbar}{2mc}= \frac{1}{2}\lambda_C=3.8\times10^{-11}\;\;{\rm cm}.
324: \]
325: The frequency of this motion is
326: \[
327: \omega_o=\frac{mc^2}{S}=\frac{2mc^2}{\hbar}=1.55\times 10^{21}\, {\rm s}^{-1}.
328: \]
329:
330: \item It turns out that although the particle is pointlike, because of the zitterbewegung
331: the charge has a localized region of influence of size $2R_o$, which is Compton's wave length.
332: The latest LEP experiments at CERN establish an upper bound of $10^{-17}$cm for the radius
333: of the charge of the electron,
334: which is consistent with this pointlike interpretation,
335: while its quantum mechanical behaviour is produced
336: for distances of its Compton wave length, six orders of magnitude larger.
337:
338: \item Properly speaking what this formalism shows is that
339: the magnetic moment $\bmu$ of the electron is not an
340: intrinsic property like the charge.
341: It is produced by the motion of the charge and therefore is orthogonal
342: to the zitterbewegung plane. But at the same time, in the center of mass frame,
343: the electron has an oscillating electric dipole
344: of magnitude $eR_o$ lying on the zitterbewegung plane.
345: Its time average value vanishes and in low energy interactions the effect
346: of this electric dipole is negligible but in high energy physics we have to take into account
347: the detailed position of the charge and thus the electric dipole contribution
348: is not negligible. This electric dipole is not related
349: to a loss of spherical symmetry of some charge distribution. The charge distribution
350: is spherically symmetric because it is just a point. This dipole
351: is just the instantaneous electric dipole moment
352: of the charge with respect to the center of mass.
353:
354: \end{itemize}
355:
356: As an approximation we can consider the classical electron
357: as a point, its center of mass, where we also locate the charge. But at the same time
358: we have to assign to this point two electromagnetic properties,
359: a magnetic moment lying along or opposite to the spin
360: direction and an oscillating electric dipole, of frequency $\omega_o$,
361: on a plane orthogonal to the spin.
362:
363: A more detailed analysis of the dynamics gives rise to the dynamical equations for the center of
364: mass ${\bi q}$ and center of charge ${\bi r}$ in an external electromagnetic field as given by \cite{Rivas4}:
365: \begin{eqnarray}
366: m\ddot{\bi q}&=&\frac{e}{\gamma(\dot{q})}\left[{\bi E}+\dot{\bi r}\times{\bi B}-\dot{\bi q}\left(\left[{\bi E}
367: +\dot{\bi r}\times{\bi B}\right]\cdot\dot{\bi q}\right)\right],\label{eq:mq2}\\
368: \ddot{\bi r}&=&\frac{1-\dot{\bi q}\cdot\dot{\bi r}}{({\bi q}-{\bi r})^2}\left({\bi q}-{\bi r}\right).\label{eq:r2enq}
369: \end{eqnarray}
370: Here, an over dot means a time derivative and the external fields are defined at the charge position ${\bi r}$,
371: and it is the velocity of the charge that produces the magnetic force term.
372:
373: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
374: \section{A direct measurement}
375: \label{sec:directexperi}
376: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
377:
378: Since the classical model depicted in figure {\ref{fig:1}} satisfies
379: when quantised Dirac's equation \cite{Rivas2}
380: it is legitimate to use it for analysing the interaction with an external electromagnetic field
381: from a classical viewpoint. We can alternatively use the Bargmann-Michel-Telegdi equation
382: for the spin evolution\cite{BMT}, but this approach assumes a minimal
383: coupling for the charge and also an anomalous
384: magnetic moment coupling. However in our approach since the spinning electron is a point charge
385: such that the magnetic moment is a consequence of the zitterbewegung, we only have to consider
386: a minimal coupling prescription which is closer to quantum electrodynamics
387: in which no anomalous magnetic moment coupling for the electron is present.
388:
389: The proposed experiment is to send a beam of transversally polarised electrons
390: or positrons and check the interaction of their magnetic moment with an external
391: magnetic field.
392: As suggested by Batelaan et al. \cite{Batelaan}
393: we shall consider a region of low or negligible magnetic field but with a
394: non-negligible field gradient such that the deflection of the beam
395: is mainly due to the magnetic dipole structure.
396:
397: We consider a beam moving along the positive direction of
398: the $OY$ axis and with the transversal spin pointing along the positive $OZ$ axis.
399: The external magnetic field will be produced by two conducting wires parallel to the $OY$ axis,
400: separated by a distance $2b$ and contained
401: in the $YOZ$ plane of a cartesian frame (see figure \ref{fig:2}).
402: If they carry a current in the same direction,
403: then the magnetic field vanishes along the $OY$ axis and is very low in its neighborhood.
404: The square depicted in the figure represents the region, using for computation,
405: where the initial position of the center of mass of the electrons in the beam is contained.
406: We have found no experimental
407: evidence of one such a device which could be useful to analyse the magnetic moment
408: of free charged particles as an alternative to the Stern-Gerlach magnets
409: which do not work properly with charged particles.
410:
411: \cfigl{fig:2}{fig2.eps}{A transversally polarised electron beam of
412: square cross section is sent along the $OY$ axis
413: into the magnetic field created by two conducting wires, separated
414: a distance $2b$, perpendicular to
415: the figure and carrying a current $I$ in the same direction.}
416:
417: The magnetic field produced by the current takes the form:
418: \begin{eqnarray*}
419: B_x&=&\frac{k(z+b)}{x^2+(z+b)^2}+\frac{k(z-b)}{x^2+(z-b)^2}, \quad
420: B_y=0,\\
421: B_z&=&-\frac{kx}{x^2+(z+b)^2}-\frac{kx}{x^2+(z-b)^2},
422: \end{eqnarray*}
423: where $k=I/2\pi\epsilon_0c^2$, $I$ is the intensity of the current
424: and $\epsilon_0$ the permittivity of the vacuum.
425:
426: To compute numerically the motion of a polarised electron beam in an external
427: magnetic field, we shall use the above dynamical equations (\ref{eq:mq2}) and (\ref{eq:r2enq})
428: with initial conditions such that the zitterbewegung plane of each electron
429: is the $XOY$ plane and the charge motion
430: produces a magnetic moment pointing along the positive $OZ$ axis.
431: For the center of mass position
432: we shall consider the electrons
433: uniformly distributed in the mentioned square region.
434:
435: \cfigl{fig:3}{fig3.eps}{Position of the beam after travelling a distance $d$, $1.5d$ and $1.6d$
436: along the $OY$ direction in the magnetic field created by the two parallel currents. The separation
437: between the wires is $200\lambda_C$, the velocity of the center of mass of electrons is $0.003c$
438: and the current $I=3$A.}
439:
440: In figure \ref{fig:3} we depict the situation of the polarised electron beam after travelling some distance
441: inside the magnetic field region. The dots represent the center of mass position of a sample of particles
442: which in the incoming beam are distributed uniformly in the shown square region.
443: With the magnetic moment of the
444: electrons pointing upwards we see a deflection (and also a focusing effect)
445: to the left. The deflection is of the same amount to the right
446: for electrons with the magnetic moment pointing down. It is checked that the deflection is independent
447: of the initial position of the electron charge compatible with the initial center of mass position.
448: We also obtain an equivalent deflection for many other values of the separation between wires,
449: current and initial beam velocity.
450:
451: We thus expect from this experiment that if the beam is deflected to the left
452: then spin and magnetic moment are parallel vectors, while they are antiparallel for right deflection.
453: The interaction does not modify the spin orientation so that we can check the polarization
454: of the beam at the exit by some direct method of electron absorption like the one devised
455: for measuring the spin of the photon in circularly polarised light beams \cite{Beth}.
456:
457: Although this device is considered for analysing charged particles the deflection
458: is produced in the low magnetic field region so that it
459: is mainly due to the interaction with the magnetic moment. This is one of the reasons
460: to consider this device as an alternative to the Stern-Gerlach magnets
461: to separate unpolarised beams.
462:
463: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
464: \section{An indirect measurement}
465: \label{sec:indirexper}
466: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
467:
468: As an indirect experiment we shall measure the relative
469: orientation of spin and magnetic moment of electrons bounded in atoms.
470:
471: Let us consider a material system formed by atoms of some specific
472: substance. Let us send a beam of circularly polarised light of such an energy
473: to produce electron transitions on these atoms from an S-state ($l=0$ orbital angular momentum)
474: into another S-state.
475: Let us assume that the photons of the circularly polarised beam have their spins pointing forward.
476: In this case the transition only affects to the electrons with the spin pointing backwards
477: such that after the transition the excited electrons have their spins in the forward direction.
478: If we now introduce a magnetic field in the forward direction to observe the Zeeman splitting then
479: the measurement of the additional interaction energy $-\bmu\cdot{\bi B}$ will give us only one of the
480: two expected transition lines of the emission spectrum from which we determine the relative orientation
481: between $\bmu$ and ${\bi B}$ and therefore between $\bmu$ and ${\bi S}$.
482:
483:
484: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
485: \section*{Acknowledgments}
486: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
487:
488: I like to acknowledge the use of the computer program {\em Dynamics Solver} developed
489: by Aguirregabiria \cite{Aguirre} with which all the numerical computations contained in this
490: paper have been done. I am very much indebted to Aguirregabiria's help.
491:
492: I also acknowledge the suggestion by Prof. Dieter Trines of DESY, Hamburg, during
493: the Prague SPIN-2001 congress, that leads to the proposed indirect measurement experiment.
494:
495: This work has been partially supported by the Universidad del Pa\'{\i}s Vasco/Eus\-kal Herriko Unibertsitatea
496: under contract UPV/EHU 172.310 EB150/98 and by General Research Grant UPV 172.310-G02/99.
497:
498:
499: \begin{thebibliography}{99}
500:
501: \bibitem{Dirac} P.A.M. Dirac, {\em The Principles of Quantum Mechanics}, 4th ed. (Oxford U. P. 1958).
502:
503: \bibitem{RichWesley} A. Rich and J.C. Wesley, Rev. Mod. Phys. {\bf 44} (1972) 250.
504:
505: \bibitem{Dehmelt} H. Dehmelt, Rev. Mod. Phys. {\bf 62} (1990) 525.
506:
507: \bibitem{Dehmelt2} H. Dehmelt, Z. Phys. {\bf D10} (1988) 127.
508:
509: \bibitem{Batelaan} H. Batelaan, T.J. Gay and J.J. Schwendiman,
510: Phys. Rev. Lett. {\bf 79} (1997) 4517.
511:
512: \bibitem{Rivas1} M. Rivas, J. Math. Phys., {\bf 30} (1989) 318; {\bf 35} (1994) 3380. For a more detailed
513: exposition see M. Rivas, {\em Kinematical theory of spinning particles}, (Kluwer, Dordrecht 2001).
514:
515: \bibitem{Rivas3} M. Rivas, J.M. Aguirregabiria and A. Hern\'andez,
516: Phys. Lett. {\bf A 257} (1999) 21.
517:
518: \bibitem{Rivas2} M. Rivas, J. Math. Phys., {\bf 35} (1994) 3380.
519:
520: \bibitem{Rivas4} M. Rivas, To be published elsewhere (physics/0112005).
521:
522: \bibitem{BMT} V. Bargmann, L. Michel and V.L. Telegdi, Phys. Rev. Lett., {\bf 2} (1959) 435.
523:
524: \bibitem{Beth} R. A. Beth, Phys. Rev. {\bf 50} (1936) 115.
525:
526: \bibitem{Aguirre} J.M. Aguirregabiria, {\em Dynamics Solver}, computer program for solving
527: different kinds of dynamical systems, which is available from his author
528: through the web site {\tt <http://tp.lc.ehu.es/jma.html>}
529: at the server of the Theoretical Physics Dept. of The University of the Basque Country,
530: Bilbao, Spain.
531:
532: \end{thebibliography}
533:
534: \end{document}
535:
536: %End of file: mu2.tex
537:
538:
539:
540: