bhobba said:
Vacuum energy and virtual particles are different things. Vacuum energy in QFT is actually infinite and one of the first indications of a sickness in QFT and the need for renormalisation, although it can be eliminated by what's called normal ordering.
Thanks
Bill
Ironically "vacuum energy" (i.e., closed bubbles without external legs in terms of Feynman diagrams) are just what's canceled when calculated connected scattering matrix elements which are encoding what's really observable, namely transition rates and cross sections. Have a look in any textbook on QFT (the keyword is the Lehmann-Symanzik-Zimmermann (LSZ) reduction formula), which is mathematically a tricky business.
What's often paraphrased as a vacuum "filled with virtual particles" is indeed just some popular-science myth. What's behind it are indeed quantum fluctuations of quantum fields which become however not observable just in vacuo but you need to probe the vacuum with something, e.g., you need an electron (and some equipment) to measure its anomalous magnetic moment leading to a deviation from the leading-order ("tree level" in the language of Feynman diagrams) prediction, these deviations are due to quantum fluctuations ("radiation corrections", i.e., "loop diagrams").
Another example is the deviation of the electric field of a charge from the classical (tree level) Coulomb field, which is also due to quantum fluctuations, called "vacuum polarization", and this puts it in much better terms than "virtual particles". Indeed due to the quantum fluctuations of the fields there's a kind of polarization of the vacuum, but this polarization is not due to the vacuum but the reaction of the vacuum to the point charge making up the electrostatic field.
Last but not least there are pure quantum effects like the Casimir effect. Also the Casimir effect is due to the presence of charges within an overall neutral material. The usual calculation in the first pages of QFT books, where a boundary-value problem is solved and two zero-point energies (both infinite by the way) are subtracted is in fact an idealization in the limit of infinite coupling constant (when ideal-conductor boundary conditions are empolyed as usual in this very simplified treatment). I think the famous paper by Jaffe has been cited alread in this thread.
You can go on and on with such examples: Whenever something is argued with "virtual particles", in fact it's something induced due to the presence of real particles and/or fields. Whenever you read about "virtual particles" in the real physics book or paper it's a paraphrase for "internal lines in Feynman diagrams", i.e., the asymptotic formal power series in the coupling (number of vertices in the Feynman diagrams) or ##\hbar## (number of loops in the Feynman diagrams), i.e., some formal expressions in terms of propagators, vertices and (often divergent) integrals!