This is my umpteenth post on the same topic. ðŸ˜¦ It is obvious that this search for a sensibleÂ interpretation is consuming me. Why? I am not sure. Studying physics is frustrating. As a leading physicist puts it:
“TheÂ teaching of quantum mechanics these days usuallyÂ follows the same dogma: firstly, the student is told about the failure of classical physics atÂ the beginning of the last century; secondly, the heroic confusions of the founding fathersÂ are described and the student is given to understand that no humble undergraduate studentÂ could hope to actually understand quantum mechanics for himself; thirdly, a deus exÂ machina arrives in the form of a set of postulates (the SchrÃ¶dinger equation, the collapseÂ of the wavefunction, etc); fourthly, a bombardment of experimental verifications is given,Â so that the student cannot doubt that QM is correct; fifthly, the student learns how toÂ solve the problems that will appear on the exam paper, hopefully with as little thought asÂ possible.”
That’s obviously not the way we want to understand quantum mechanics. [With we,Â I mean, me, of course, and you, if you’re reading this blog.]Â Of course, that doesn’t mean I don’t believe Richard Feynman, one of the greatest physicists ever, when he tells us no one, including himself, understands physics quite the way we’dÂ likeÂ to understand it. Such statements should not prevent us from tryingÂ harder. So let’s look for betterÂ metaphors.Â The animation below shows the two components of the archetypal wavefunction â€“ a simple sine and cosine. They’re the same function actually, but their phases differ by 90 degrees (Ï€/2).
It makes me think of a V-2 engine with the pistons at a 90-degree angle. Look at the illustration below, which I took from a rather simple article on cars and engines that has nothing to do with quantum mechanics. Think of the moving pistons as harmonic oscillators, like springs.
We will also think of theÂ center of each cylinder as the zero point: think of that point as a point where – if we’re looking at one cylinder alone – the internal and external pressure balance each other, so the piston would not move… Well… If it weren’t for the other piston, because the second piston isÂ not at the centerÂ when the first is. In fact, it is easy to verify and compare the following positions of both pistons, as well as the associated dynamics of the situation:
Piston 1 |
Piston 2 |
Motion of Piston 1 |
Motion Piston 2 |
Top |
Center |
Compressed air will push piston down |
Piston moves down against external pressure |
Center |
Bottom |
Piston moves down against external pressure |
External air pressure will push piston up |
Bottom |
Center |
External air pressure will push piston up |
Piston moves further up and compresses the air |
Center |
Top |
Piston moves further up and compresses the air |
Compressed air will push piston down |
When the pistons move, their linear motion will be described by a sinusoidal function: a sine or a cosine. In fact, the 90-degree V-2 configuration ensures that the linear motion of the two pistons will be exactly the same, except for a phase difference of 90 degrees. [Of course, because of the sideways motion of the connecting rods, our sine and cosine function describes the linear motion only approximately, but you can easily imagine the idealizedÂ limit situation. If not, check Feynman’s description of the harmonic oscillator.]
The question is: if we’d have a set-up like this, two springs – or two harmonic oscillators – attached to a shaft through a crank, would this really work as a perpetuum mobile? We obviously talk energyÂ being transferred back and forth between the rotating shaft and the moving pistons… So… Well… Let’s model this: the totalÂ energy, potentialÂ andÂ kinetic, in each harmonic oscillator is constant. Hence, the piston only delivers or receivesÂ kineticÂ energy from the rotating mass of the shaft.
Now, in physics, that’s a bit of an oxymoron: we don’t think of negative or positive kinetic (or potential) energy in the context of oscillators. We don’t think of the direction of energy. But… Well… If we’ve got twoÂ oscillators, our picture changes, and so we may have to adjust our thinking here.
Let me start by giving you an authoritative derivation of the various formulas involved here, taking the example of the physical spring as an oscillatorâ€”but the formulas are basically the same forÂ any harmonic oscillator.
The first formula is a general description of the motion of our oscillator. The coefficient in front of the cosine function (a)Â is the maximum amplitude. Of course, you will also recognize Ï‰_{0}Â as theÂ naturalÂ frequency of the oscillator, andÂ Î” as the phase factor, which takes into account our t = 0 point. In our case, for example, we have two oscillators with a phase difference equal to Ï€/2 and, hence, Î” would be 0 for one oscillator, and â€“Ï€/2 for the other. [The formula to apply here is sinÎ¸ = cos(Î¸ â€“ Ï€/2).] Also note that we can equate our Î¸ argument to Ï‰_{0}Â·t.Â Now, ifÂ aÂ = 1 (which is the case here), then these formulas simplify to:
- K.E. = T = mÂ·v^{2}/2 =Â mÂ·Ï‰_{0}^{2}Â·sin^{2}(Î¸ + Î”) = mÂ·Ï‰_{0}^{2}Â·sin^{2}(Ï‰_{0}Â·t + Î”)
- P.E. = U = kÂ·x^{2}/2 = kÂ·cos^{2}(Î¸ + Î”)
The coefficient k in the potential energy formula characterizes the force: F = âˆ’kÂ·x. The minus sign reminds us our oscillator wants to return to the center point, so the force pulls back. From the dynamics involved, it is obvious that k must be equal to mÂ·Ï‰_{0}^{2}., so that gives us the famous T + U = mÂ·Ï‰_{0}^{2}/2 formula or, including aÂ once again, T + U = mÂ·a^{2}Â·Ï‰_{0}^{2}/2.
Now, if we normalizeÂ our functions by equating k to one (k = 1), thenÂ the motion ofÂ our first oscillator is given by the cosÎ¸ function, and its kinetic energy will be equal toÂ sin^{2}Î¸. Hence, the (instantaneous)Â changeÂ in kinetic energy at any point in time will be equal to:
d(sin^{2}Î¸)/dÎ¸ = 2âˆ™sinÎ¸âˆ™d(sinÎ¸)/dt = 2âˆ™sinÎ¸âˆ™cosÎ¸
Let’s look at the second oscillator now. Just think of the second piston going up and down in our V-twin engine. Its motion is given by theÂ sinÎ¸ function which, as mentioned above, is equal to cos(Î¸âˆ’Ï€ /2). Hence, its kinetic energy is equal toÂ sin^{2}(Î¸âˆ’Ï€ /2), and how itÂ changesÂ – as a function of Î¸ – will be equal to:
2âˆ™sin(Î¸âˆ’Ï€ /2)âˆ™cos(Î¸âˆ’Ï€ /2) =Â = âˆ’2âˆ™cosÎ¸âˆ™sinÎ¸ = âˆ’2âˆ™sinÎ¸âˆ™cosÎ¸
We have our perpetuum mobile! While transferring kinetic energy from one piston to the other, the rotating shaft moves at constant speed. Linear motion becomes circular motion, and vice versa, in a frictionless Universe. We have the metaphor we were looking for!
Somehow, in this beautiful interplay between linear and circular motion, energy is being borrowed from one place to another, and then returned. From what place to what place? I am not sure. We may call it the real and imaginary energy space respectively, but what does that mean? One thing is for sure, however: the interplay between the real and imaginary part of the wavefunction describes how energy propagates through space!
How exactly? Again, I am not sure. Energy is, obviously, mass in motion – as evidenced by the E = mÂ·c^{2}Â equation, and it may not have any direction (when everything is said and done, it’s a scalarÂ quantity without direction), but the energy in a linear motion is surely different from that in a circular motion, and our metaphor suggests we need to think somewhat more along those lines. Perhaps we will, one day, able toÂ square this circle. ðŸ™‚
SchrÃ¶dinger’s equation
Let’s analyze the interplay between the real and imaginary part of the wavefunction through an analysis of SchrÃ¶dinger’s equation, which we write as:
iÂ·Ä§âˆ™âˆ‚Ïˆ/âˆ‚t = â€“(Ä§^{2}/2m)âˆ™âˆ‡^{2}Ïˆ + VÂ·Ïˆ
We can do a quick dimensional analysis of both sides:
- [iÂ·Ä§âˆ™âˆ‚Ïˆ/âˆ‚t] = Nâˆ™mâˆ™s/s = Nâˆ™m
- [â€“(Ä§^{2}/2m)âˆ™âˆ‡^{2}Ïˆ] = Nâˆ™m^{3}/m^{2} = Nâˆ™m
- [VÂ·Ïˆ] = Nâˆ™m
Note the dimension of the ‘diffusion’ constantÂ Ä§^{2}/2m: [Ä§^{2}/2m] = N^{2}âˆ™m^{2}âˆ™s^{2}/kg = N^{2}âˆ™m^{2}âˆ™s^{2}/(NÂ·s^{2}/m) = Nâˆ™m^{3}. Also note that, in order for the dimensions to come out alright, the dimension of V â€“ the potential â€“ must be that of energy. Hence, Feynmanâ€™s description of it as the potential energy â€“ rather than the potential tout courtÂ â€“ is somewhat confusing but correct: V must equal the potential energy of the electron. Hence, V is not the conventional (potential) energy of the unit charge (1 coulomb). Instead, the natural unit of charge is used here, i.e. the charge of the electron itself.
Now, SchrÃ¶dingerâ€™s equation â€“ without the VÂ·Ïˆ term â€“ can be written as the following pair of equations:
- Re(âˆ‚Ïˆ/âˆ‚t) = âˆ’(1/2)âˆ™(Ä§/m)âˆ™Im(âˆ‡^{2}Ïˆ)
- Im(âˆ‚Ïˆ/âˆ‚t) = (1/2)âˆ™(Ä§/m)âˆ™Re(âˆ‡^{2}Ïˆ)
This closely resembles the propagation mechanism of an electromagnetic wave as described by Maxwell’s equation for free space (i.e. a space with no charges), but E and B are vectors, not scalars. How do we get this result. Well… Ïˆ is a complex function, which we can write as a + iâˆ™b. Likewise, âˆ‚Ïˆ/âˆ‚t is a complex function, which we can write as c + iâˆ™d, and âˆ‡^{2}Ïˆ can then be written as e + iâˆ™f. If we temporarily forget about the coefficients (Ä§, Ä§^{2}/m and V), then SchrÃ¶dingerâ€™s equation – including VÂ·Ïˆ term – amounts to writing something like this:
iâˆ™(c + iâˆ™d) = â€“(e + iâˆ™f) + (a + iâˆ™b) â‡” a + iâˆ™b = iâˆ™c âˆ’ d + e+ iâˆ™f Â â‡” a = âˆ’d + e and b = c + f
Hence, we can now write:
- Vâˆ™Re(Ïˆ) = âˆ’Ä§âˆ™Im(âˆ‚Ïˆ/âˆ‚t) + (1/2)âˆ™( Ä§^{2}/m)âˆ™Re(âˆ‡^{2}Ïˆ)
- Vâˆ™Im(Ïˆ) = Ä§âˆ™Re(âˆ‚Ïˆ/âˆ‚t) + (1/2)âˆ™( Ä§^{2}/m)âˆ™Im(âˆ‡^{2}Ïˆ)
This simplifies to the two equations above for V = 0, i.e. when there is no potential (electron in free space). Now we can bring the Re and Im operators into the brackets to get:
- Vâˆ™Re(Ïˆ) = âˆ’Ä§âˆ™âˆ‚Im (Ïˆ)/âˆ‚t + (1/2)âˆ™( Ä§^{2}/m)âˆ™âˆ‡^{2}Re(Ïˆ)
- Vâˆ™Im(Ïˆ) = Ä§âˆ™âˆ‚Re(Ïˆ)/âˆ‚t + (1/2)âˆ™( Ä§^{2}/m)âˆ™âˆ‡^{2}Im(Ïˆ)
This is very interesting, because we can re-write this using the quantum-mechanical energy operator H = â€“(Ä§^{2}/2m)âˆ™âˆ‡^{2} + VÂ· (note the multiplication sign after the V, which we do not have â€“ for obvious reasons â€“ for the â€“(Ä§^{2}/2m)âˆ™âˆ‡^{2} expression):
- H[Re (Ïˆ)] = âˆ’Ä§âˆ™âˆ‚Im(Ïˆ)/âˆ‚t
- H[Im(Ïˆ)] = Ä§âˆ™âˆ‚Re(Ïˆ)/âˆ‚t
A dimensional analysis shows us both sides are, once again, expressed in Nâˆ™m. Itâ€™s a beautiful expression because â€“ if we write the real and imaginary part of Ïˆ as râˆ™cosÎ¸ and râˆ™sinÎ¸, we get:
- H[cosÎ¸] = âˆ’Ä§âˆ™âˆ‚sinÎ¸/âˆ‚t = Eâˆ™cosÎ¸
- H[sinÎ¸] = Ä§âˆ™âˆ‚cosÎ¸/âˆ‚t = Eâˆ™sinÎ¸
Indeed, Î¸Â = (Eâˆ™t âˆ’ pâˆ™x)/Ä§ and, hence, âˆ’Ä§âˆ™âˆ‚sinÎ¸/âˆ‚t = Ä§âˆ™cosÎ¸âˆ™E/Ä§ = Eâˆ™cosÎ¸ and Ä§âˆ™âˆ‚cosÎ¸/âˆ‚t = Ä§âˆ™sinÎ¸âˆ™E/Ä§ = Eâˆ™sinÎ¸.Â Now we can combine the two equations in one equation again and write:
H[râˆ™(cosÎ¸ + iâˆ™sinÎ¸)] = râˆ™(Eâˆ™cosÎ¸ + iâˆ™sinÎ¸) â‡” H[Ïˆ] = Eâˆ™Ïˆ
The operator H â€“ applied to the wavefunction â€“ gives us the (scalar) product of the energy E and the wavefunction itself. Isn’t this strange?
Hmm… I need to further verify and explain this result… I’ll probably do so in yet another post on the same topic… ðŸ™‚
Post scriptum: The symmetry of our V-2 engine – or perpetuum mobileÂ – is interesting: its cross-section has only one axis of symmetry. Hence, we may associate some angle with it, so as to define its orientation in the two-dimensional cross-sectional plane. Of course, the cross-sectional plane itself is at right angles to the crankshaft axis, which we may also associate with some angle in three-dimensional space. Hence, its geometry defines two orthogonal directions which, in turn, define a spherical coordinate system, as shown below.
We may, therefore, say that three-dimensional space is actually being impliedÂ byÂ the geometry of our V-2 engine. Now thatÂ isÂ interesting, isn’t it? ðŸ™‚