

|
In order to emphasize various things I have coloured operators, eigenfunctions etc.: Rotation operators are so named because when they operate on a state vector they produce a 'rotation' in the state space. As nmr spectroscopists we are generally interested in what is happening to our spin systems as time passes and, along with the density operator, rotation operators are very helpful to us. They are very easy to visualize for an object in Cartesian coordinate space. For simpliciy, let's look at the rotation of a 2D vector: ![]() This is actually a rotation of coordinate axes rather than a vector rotation but it is entirely equivalent. Our problem is, given starting values of x0 and y0 in the coordinate axes x and y, what will the new values x1 and y1 be in the new coordinate axes x' and y'. Looking at the diagram and the red line and using a little basic trigonometry you should be able to see that: A = cos(
and since A+B=x1 we can write: x1 = cos(
Similarly, for y1 we find: y1 = -sin(
We can write this in matrix notation: ( x1 ) = ( cos(
( y1 ) ( -sin( or even more compactly: V' = AV
Yet another way to look at this is to characterize the vector with a complex number: ![]() Algebraically we write: r = x + iy
Simple trigonometry gives us: x = |r|cos(
y = |r|sin( and substituting into the first equation: r = |r|cos(
= |r|(cos( Using the Euler identity: e(i
we rewrite: r = |r|e(i
So we now have an exponential representation of our vector. What about rotation? Let's multiply |r|e(i r = |r|e(i
= |r|e(i This must represent a new vector that has phase angle ![]() So, our original vector, at phase angle Rotation operators for angular momentum operators are slightly more complicated but their form is exponential as in the previous example. We start with the time-dependent Schrodinger equation: d
where This is just a simple 1st order differential equation with a simple solution: |
where | Now let's see what happens if we calculate the expectation value for the operator, <Ix> = <
= <(exp(i
= <
= <
where I have defined: g(
The first derivative of g is: g'(
Remembering our angular momentum commutators: [
we substitute: g'(
Next, the second derivative: g''(
= -exp(-i
= -g(
or, in standard differential equation form: g''(
The solution to this equation is: g(
A equals g(0) which is just g(
So: Similarly: Does this look familiar? Look back up this page to the very beginning where vectors in Cartesian space were being discussed. This equation has exactly the same form as the simple rotation equations. Evidently, an expression such as: exp(-i
corresponds to a rotation about the z-axis by R(
|