# Wave functions

1. ### redtree

64
Why are wave functions, e.g., Schrodinger's, based on the complex exponential function (e$$^{}ix$$) and not trigonometric functions (sine or cosine)?

See Euler's formula for their relationship: http://en.wikipedia.org/wiki/Euler's_formula

Furthermore, by using the complex exponential function, the probability amplitude becomes a complex-valued function (a + bi). Were sine or cosine used, the probability amplitude of the wave function would not be a complex-valued function. Is there a reason that the probability amplitude should be a complex-valued function?

Last edited: Oct 10, 2008
2. ### Ben Niehoff

1,753
You need complex-valued wavefunctions so that you can have standing waves of constant magnitude over time (such as in a potential well). Real-valued functions would have to oscillate in magnitude.

3. ### redtree

64
What do you mean by magnitude? Is it the same as amplitude?

4. ### koolmodee

55
Nothing keeps you from using sine and cos. That what Euler's identity says.

The time evolution of a state vector/ wave function must conserve the norm of the state vector, and it always has to be one.

Note e$$^{it}$$ e$$^{it}$$=1.

5. ### akhmeteli

607
My post https://www.physicsforums.com/showpost.php?p=1539835&postcount=20 and some other posts in that thread may be relevant.

6. ### dx

2,009
The simple answer is that standard quantum mechanics as we understand it requires a complex wavefunction $$\psi$$ defined on the configuration space. You can replace this with a function to $$\mathbb{R} \times \mathbb{R}$$ and change the equations accordingly, since $$\mathbb{C}$$ and $$\mathbb{R} \times \mathbb{R}$$ are isomorphic. But you can't replace $$\psi$$ with a function to just $$\mathbb{R}$$.

Also, Euler's formula doesn't turn a complex number into a real number. $$\cos \theta + i \sin \theta$$ is still a complex number, and the probability amplitude will still be a complex valued function.

7. ### redtree

64
What properties of standard quantum mechanics require a complex wavefunction?

4,725

9. ### dx

2,009
That's like asking what property of Newtonian physics requires the mechanical state of a particle to be position and momentum. It is possible to construct a theory where the time evolution of a particle depends only on initial position, but that's not the way nature is. It just happens to be so that the mechanical state of a quantum system needed to predict future probabilities is a complex function.