Momentum Operators and Observables Recall the time-independent Schrodinger equation: # ! 2 "2 & % ! 2m "x 2 + U(x) ( ) (x) = E) (x) $ ' E is the total energy of the particle: p2 E = KE + U = +U 2m Comparing the two equations we see that: ! 2 "2 p2 " ! = KE = # p = !i! 2 2m "x 2m "x Momentum is described by a derivative operation in the position representation. Momentum Operators and Observables Every observable quantity is described by an operator which can operate on wave functions. For position wave functions ψ(x), Position : x (operation : multiply by x) Momentum : p = !i! " "x (operation :partial derivative) For momentum wave functions ζ(p), Position : i! ! !p Momentum : p Observables that are functions of x and p are described by operators that are corresponding functions of the x and p operators Momentum Observables and corresponding operators (position wave functions) Momentum Expectation values " f (x) = # !" " f (x)P(x)dx = # f (x) $ (x,t) dx 2 !" What about the mean value of p ? Recall that p = mv = m d x dx ! p =m dt dt Using " x = # " !" xP(x)dx = # x $ (x,t) dx 2 !" And taking the time derivative, after some manipulation using the Schrodinger equation we get, $ $ ! "! p = %!* dx = % ! * p! dx i "x #$ #$ Momentum Expectation values $ $ * ! "! p = %! dx = % ! * p! dx i "x #$ #$ In general the average value of any observable O is # O == * ! $ Ô! dx "# Where Ô is the associated operator. Momentum Distribution Any function can be written as a sum of cosine and sine waves: Periodic function: Fourier Series: Spatial function at a fixed time: f (x) = " #ae n= !" ikn x n " = # a (cos(k x) + isin(k x)) n n= !" L n L n 1 1 ! ikn x an = f (x)e dx = f (x) cos(kn x) ! isin(kn x) dx " " 2L ! L 2L ! L ( ) Temporal function at a fixed position: V (t) = # $ae n= "# 1 an = 2T T # V (t)e !T ik! n t n ! i" n t = # $ a (cos(! t) + isin(! t)) n= "# 1 dt = 2T n n n T # V (t) (cos(" t) ! isin(" t)) dt n !T n Momentum Distribution Non-periodic function: Fourier Integral Spatial function at a fixed time: f (x) = a(k) = # 1 $ a(k)e 2! dk = "# # 1 2! ikx $ f (x)e" ikx dx = "# 1 2! 1 2! # $ a(k) (cos(kx) + isin(kx)) dk "# # $ "# ( ) f (x) cos(kx) " isin(kx) dx Temporal function at a fixed position: V (t) = g(! ) = $ 1 2! % g(" )ei" t d" = #$ 1 2" $ # i! t V (t)e dt = % #$ 1 2! 1 2" $ % g(" ) (cos(" t) + isin(" t)) d" #$ $ % V (t) (cos(! t) # isin(! t)) dt #$ Momentum Distribution Examples: See http://www.falstad.com/fourier/ for a demonstration of Fourier series Momentum Distribution Fourier decomposition of some functions: f(x) a(k) A cos ( k1 x ) + A cos ( k2 x ) f(x) a(k) Momentum Distribution Fourier decomposition of some functions: f(x) a(k) # !(x ! x )2 & o ( f x = C exp % %$ 2" 2 (' x # !(k ! k )2 & o ( a k = N exp % %$ 2" 2 (' k () () Gaussian Gaussian x k Recall the reciprocity relations: !x!k " 1, !#!t " 1 For Gaussian functions, the uncertainties are minimized !x!k = 1 1 , !"!t = 2 2 Heisenberg Uncertainty Principle Minimum uncertainty: !x!k = 1 1 , !"!t = 2 2 p = !k, E = !! !p!x " • • ! 2 !E!t " ! 2 Our knowledge of conjugate quantities is inherently uncertain. No matter how good our measuring instruments are we cannot simultaneously know x and p or E and t with complete precision. Gaussian For Gaussian functions: ! !p!x = 2 x ! !E!t = 2 Gaussian k Momentum Distribution Momentum space wave functions Since p=hk/2π, the functions a(k) can be easily written as functions ζ(p). Given a wave function in coordinate space representing the position of a particle, the corresponding momentum wave function is the Fourier integral of the the position wave function: ! ( p,t) = 1 e % ( 2" ! ) 1/2 #i ( xp ) ! $ (x,t)dx If the position wave function is normalized, the momentum wave function is also normalized to unity: " ! ( p,t) 2 dp = 1 Momentum Distribution Momentum space wave functions We can interpret the square of the quantity P( p,t) = ! ( p,t) dp 2 of the normalized wave function as the probability at time t of the momentum of a particle being in the volume dp around p Hence the normalization requirement is a statement that the probability of finding the particle with some value of momentum is 1. " ! ( p,t) 2 dx = " P( p,t)dp =1 Momentum Distribution Momentum space wave functions • The probability of finding a particle with momentum a < p < b is then b " ! ( p,t) dp 2 a This is the area under the momentum probability density curve between p=a and p=b: ! ( p,t) 2 p Momentum Distribution Given a wave function representing the momentum of a particle, the corresponding position space wave function is the inverse Fourier integral of the momentum wave function: ! (x,t) = 1 e # ( 2" ! ) 1/2 i ( px ) ! $ ( p,t)dp If the momentum wave function is normalized, the position wave function is also normalized to unity. Note that the position and momentum wave functions represent the same state of the system. All information about the state can be obtained from either wave function.