# Negative Energy

1. Feb 27, 2005

### Thomas Ruedel

In the transition from "basic" quantum theory to quantum field theory, both in the derivation of the Gordon-Klein and of the Dirac-equations, there is a big problem with solutions with negative energies. Can anyone explain me why this is being considered a big problem in the first place (it seems obvious to textbook authors). After all, energy is anyhow only determined up to a constant. And electrons within atoms are generally assigned negative energies.
This is probably one of those stupid questions with obvious answers which people just forget to mention. The more grateful I would be for an answer.

2. Feb 27, 2005

### marlon

There is no problem with that. But just to be clear, negative energy solutions really means the negative square root coming from the Einstein energy relation : E² = p²c² + m'²c^4 where m' is the restmass.

These negative solutions correspond to anti-particles. For example positrons.

In the case of electrons, you are mixing this with potential energy. Indeed the latter is negative (Coulombic potential). However, electrons obey the positive square root relation for energy.

Keep in mind that the Einstein energy formula does not account for the interactions that particles can undergo. It just expresses the energy of a particle with certain rest energy and certain momentum p. This, however does NOT change the negative Coulombic potential between charged particles

marlon

Last edited: Feb 27, 2005
3. Feb 27, 2005

### dextercioby

Einstein's formula is valid for particles which are on their mass sheet.For example,free particles are always on theor mass sheet.Virtual particles,not.

Daniel.

4. Feb 27, 2005

### marlon

As a matter of fact, there are two (connected) ways to determin whether a particle is virtual or not.

1) if a particle does NOT obey the Einstein-energy relation
2) if in a Feynmann diagramma, the lines of a particle start and end at the same vertex, the particle is virtual. Though, it can be promoted to a real particle for a short time, thanks to HUP

marlon

5. Feb 27, 2005

### Thomas Ruedel

In the case of electrons, you are mixing this with potential energy. Indeed the latter is negative (Coulombic potential). However, electrons obey the positive square root relation for energy.

marlon[/QUOTE]

Thanks a lot. My problem is really quite basic, I´m afraid. Are you saying something like the following?
" Traditionally one would have thought that potential energy can be negative but kinetic energy cannot. And indeed kinetic energy cannot be negative for the electron. However, the fact that there are solutions for free particles displaying negative (kinetic) energies has led to the discovery of anti-particles, such as the positron. And they can have negative kinetic energies."
Is this a correct understanding?

6. Feb 27, 2005

### marlon

The E in the Einstein relation is NOT kinetic energy. So no, you are wrong. Besides in QM electrons can have negative kinetic energy. This happens with QM-tunnelling effects

marlon

7. Feb 27, 2005

### Sterj

"Einstein's formula is valid for particles which are on their mass sheet.For example,free particles are always on theor mass sheet.Virtual particles,not."

Why aren't virtual particles on their mass sheets? Can you please explain me that?

8. Feb 27, 2005

### dextercioby

Because you can't cave both energy-momentum conservation and ONLY real particles in interaction.Take for example the vertex in QED and try to impose at the same time that the 2 fermions and the photon to be on their mass sheets and 4momentum to be conserved in vertex.

Tell me what u get.

Daniel.

9. Feb 27, 2005

### Sterj

@dexter: I Don't really know what you want (what shall I try to calculate?). Can you give me a beginning of what I should calculate?

10. Feb 27, 2005

### dextercioby

Yes,I've told you what to do.I think it was clear enough.Consider the vertex in QED and try to see what happens if you combine 4momentum conservation+the condition that all 3 particles are real/on their mass sheets...

Daniel.

11. Feb 27, 2005

### Sterj

Ok, I'll try it, I'll give you a solution tomorrow.

12. Feb 27, 2005

### dextercioby

As you wish.I'm not asking you for anything.It was just an example to show u what happens when we have virtual particles and the connection to "mass sheet",the latter being a key concept in SR.

Daniel.

13. Feb 28, 2005

### SymmetryBreaking

I don't know if its related to the topic but does this negative energy really cancels out the positive one? Like for example, we are experimentally getting less of the expected total vacuum energy because of this effect(pls correct me if i'm wrong). :uhh:

14. Feb 28, 2005

### Sterj

@dexter: yeah, It's clear now, why virtual particles aren't in their mass sheet. (Because they receive their energy from the Heisenberg's uncertainty principle. And this energy most disappear after a short time).

But I've another question. I asked a physicist, why for example virtual photons annihilate after a short time (I mean the antiparticle collides with the particle but for this the antiparticle has to know the position of the particle and to collide the particle and the antiparticle have to go in the same direction, but what is if they don't go in the same direction?). You know what I mean?

--> He's answers didn't sadisfy me. He said, that it is so, ... (I think he didn't know an answer, perhaps there isn't really an answer for that).

What do you mean?

15. Feb 28, 2005

### reilly

I suggest that a little history is in order. Back in Dirac's time, the existence of negative energy states, down to - infinity, indicated a complete collapse of electrons, protons as wel, etc. unless the negative energy states were filled creating a rather extensive Fermi Sea. The rest is history, positrons were holes in the big Fermi Sea, and finally, in modern QM, particles are anti particles are afforded equal protection under the law. All of this stuff is, or should be, discussed in great detail in any text or treatise on relativistic QM, Weinberg, Schweber's QED and the Men Who Made It -- just a couple of suggestions.
Regards,
Reilly Atkinson

16. Feb 28, 2005

### Sterj

I know that "reilly". But I can't find my answer in any text book of qm or qed.

again:
"""""""But I've another question. I asked a physicist, why for example virtual photons annihilate after a short time (I mean the antiparticle collides with the particle but for this the antiparticle has to know the position of the particle and to collide the particle and the antiparticle have to go in the same direction, but what is if they don't go in the same direction?). You know what I mean?

--> He's answers didn't sadisfy me. He said, that it is so, ... (I think he didn't know an answer, perhaps there isn't really an answer for that).

What do you mean?""""""

17. Feb 28, 2005

### marlon

Again, if you look at a particle that interacts with an anitiparticle, don't imagine just two particles that are annihilated. In QM you need to look at this as two BEAMS of particles that interact (read : scatter) with each other. The probability of such an interaction is expressed by quentities like the cross section of the beam and the scattering amplitude.

Virtual particles exist for a short while because of the uncertainty principle. Suppose you look at the vacuum (ie the lowest energy state). It has a NON-zero energy value because otherwise you cannot detect any particles in this state (you know, because of :$$\Delta x * \Delta p = constant$$)Now Since the energy value is NONZERO we have mass (because of E=mc²) and therefore the vaccuum is filled with particles. These particles (of energy E)exist for a short period of time because of $$\Delta E * \Delta t = constant$$. These particles are the socalled vacuumfluctuations. These particles are indeed virtual because they only exist for an intermediate stage and they can be used in interactions between real particles that are on mass shell

regards
marlon

18. Feb 28, 2005

### Sterj

@Marlon, that's what I know. But it exists only a probability that the particle and antiparticle interact (annihilate). So how can a particle find its "friend" (I know that we can say that particle and antiparticle are a kind of wave with probability to be at a position).

19. Feb 28, 2005

### marlon

Clearly you did not get my point. A particle does NOT find or recognize another anti-particle. Again i refer to the scattering processes and the probability of interaction (annihilation)

marlon

ps : for all that is interested, i posted a text on virtual particles in my journal. Check it out, it is the 5th entry on the designated page (on vacuum fluctuations and virtual particles)
https://www.physicsforums.com/journal.php?s=&action=view&journalid=13790&perpage=10&page=2

Last edited: Feb 28, 2005
20. Mar 1, 2005

### Sterj

Your text makes it a little bit simpler to understand. With "number operator" you mean this N=ab with a the creation operator and b the annihilation operator, right?

"suppose each such particle is represented by one LOWEST energy quantum of a harmonic oscillator"
You mean each particle has energy of: E=h(bar)w*1/2=h/(4pi)*w=h*f/2. Shouldn't the energy of a photon be: E=hf? Or is the ground state of the electro magnetic field described by the harmonic oscillator?