# Do photons create gravity?

When we have pair production , a photon turns into an electron and positron , I would like to think that the gravitational field of the electron and positron came from the photon,
And vice versa if we collide an electron and positron and we get a photon out, It seems like the G field of the electron and positron would get transferred to the photon.

D H
Staff Emeritus
I do not believe the topic is whether photons gravitate, the topic is whether a single photon can create a gravitational field, these are in my mind two different things.
Caveat: I am not an expert in GR.

My understanding is that a single photon does not create a gravitational field. A collection of photons can, however. The intrinsic mass of a collection of particles is

$$m = \frac{\sqrt{E^2 - p^2c^2}}{c^2}$$

where E is the total energy of the collection of particles and p2 is the square of the net momentum of the collection. While energy and momentum are both frame-dependent quantities, the difference E2 - p2c2 is frame independent.

For a single photon, E=pc, so the intrinsic mass of a single photon is zero. Now consider photons created by an electron-positron annihilation. In the rest frame of the (former) electron-positron pair, the created photons have zero net momentum. The total energy of these photons is equal to the 2mec2 plus any kinetic energy of the electron-positron pair. The intrinsic mass of the photons created by this annihilation event is identically equal to the intrinsic mass of the electron + positron system prior to the annihilation event.

Last edited:
Dale
Mentor
2021 Award
No, im saying you are all wrong.
If you believe that GR is correct and that we are wrong then you do not understand GR. I recommend further study. In particular, you should learn about pp-wave spacetimes.

Dale
Mentor
2021 Award
But we don't need a quantum theory of gravity to answer this question. All we need is QED in curved spacetime, and that is no problem at all.

1. Observationally, we know single photons fall.
2. GR says that momentum is conserved (in this case).
3. Therefore, single photons must gravitate.

That conclusion can only be escaped by asserting GR is incorrect.
I like this. That is a good point.

Alternatively, one can put in a single photon in T and again GR shows the effect on G.
This one is the dangerous approach. How can you define T when the photon does not have a definite energy, momentum, and location at any given time? Actually, I am not even sure that spacetime can be represented by a manifold at sufficiently small scales due to quantum effects.

Staff Emeritus
2021 Award
A single photon does create a gravitational field, because T is non-zero. If T is non-zero, so is G.

Dale
Mentor
2021 Award
If T is non-zero then what is T?

"everything" distorts/curves/interrupts etc space-time

thats what gravity is

pervect
Staff Emeritus
While it is true that photons do not have mass, it is incorrect to conclude that because they don't have a mass, they can't cause gravity. Photons don't need mass to cause gravity, all they need is energy and momentum, which they do have, because it's the stress energy tensor that causes gravity and not "mass".

While GR is a classical theory, and doesn't have anything to do directly with photons per se, you can model a photon in GR as a packet of light. Which is essentially what Tolman et al did in the 1931 paper I mentioned earlier, with the results I mentioned - parallel light packets don't attract each other, there is no "self focusing" effect of light due to gravity, but beams in opposite directions do interact gravitationally.

Please!!! In this thread there has been nothing but manipulation of Albert’s general relativity to produce a theoretical scenario where imaginary particle states can produce gravitation. Its not the photon with gravitation that is bugging me, it’s the use of Albert’s notions in producing general relativity to prove your hypothesis of bosons produce gravity is false because the theory in its entirety didn’t explain theoretical zero mass particles producing gravity and your existence to interpret the information in such a way to try and change theoretical understandings of a particle. What are you suggesting number 1? Where in Einstein’s formulas does it produce an explanation of zero mass particles producing gravity 2. And please note that a wave’s propagation can be seen as action at a distance but not gravity. How can you compress a field for density not field compression or amplification which will create distortion not density . Propagation is not gravity. Waves like light (sigh), do not produce gravity

This thread is in the relativity section, and according to GR, "light" does produce gravity.

If you looked at a completely "empty" Universe that was void of any form of energy, matter, radiation, etc., in a 4 dimensional spacetime where

$$\mathbf{R(X,Y)Z}=R^{a}_{\ bcd}X^{c}Y^{d}Z^{b}=0=R_{bd}$$

and then let a wave of radiation pass through then you get the interesting scenario where calculating the Ricci tensor will still give you zero
$$R_{bd}=0$$
but the Riemann tensor isn't required to vanish and you can find calculations where $$R^{a}_{\ bcd}X^{c}Y^{d}Z^{b}\neq 0$$, ie., the radiation causes curvature of the spacetime (in 4 or greater dimensions).

The fact that GR predicts that light alters the spacetime geometry is much more general, but that is a situation where nothing exists to cause curvature, so "gravitational" curvature is entirely from the radiation.

Staff Emeritus
2021 Award
If T is non-zero then what is T?

It's been a while, but I believe the upper left 2x2 is 1 and the other components are 0.

bcrowell
Staff Emeritus
Gold Member
Please!!! In this thread there has been nothing but manipulation of Albert’s general relativity to produce a theoretical scenario where imaginary particle states can produce gravitation.

Dale
Mentor
2021 Award
It's been a while, but I believe the upper left 2x2 is 1 and the other components are 0.
I believe that is correct for a classical pulse of light at a definite location and of known (unit) energy density propagating in the +x direction. But that is not the same as an EM quantum with uncertain location and non-definite energy. What you are describing is simply GR for classical pulses of light, not for photons. Do you see the difference and why I am reluctant to make conclusions for a photon based on classical pulses of light?

I like this. That is a good point.

This one is the dangerous approach. How can you define T when the photon does not have a definite energy, momentum, and location at any given time? Actually, I am not even sure that spacetime can be represented by a manifold at sufficiently small scales due to quantum effects.

The QM theory gives only probability distributions of photon location, its momentum and energy, but photon must have a definite energy, momentum and location in time that we cannot know by QM.

pervect
Staff Emeritus
Energy, momentum, etc would only be expectations for a photon. But we expect that any quantum theory of light is going to have to include gravity in the classical limit, because we know that collections of photons do gravitate under certain circumstances, and we know which ones.

Furthermore, I already posted a link to a book by Zee which discusses how you get quantum gravity for the photon perturbatively - though it's outside the scope of GR.

So, I'd suggest continuing the QM side of the debate in the "beyond the standard model" forum.

I'd also encourage people interested to look at the reference by Zee, perhaps it got lost in the shufle.

No. According to GR the source of gravity is the stress-energy tensor. There are 10 independent components in the stress-energy tensor. Energy is only one of those 10 components.

http://en.wikipedia.org/wiki/Stress-energy_tensor

We do not have a working theory of quantum gravity at the time so I cannot answer your question wrt photons, however I can answer it wrt classical pulses of light. Pulses of light have energy, they also have momentum, so several of the components of stress energy tensor will be non-zero. So light can be a source of gravity.

I wonder what you think of the following thought experiment by Dimitry67 that seems to show that parallel beams of light will not converge. Thus it is as though light can't be a source of gravity.

Consider two massive objects, separated by some distance, flying in the same direction at velocity v according to an observer. In their inertial system they collide, say, in 1s. For the observer this process takes longer because of the time dilation. The faster the two objects are flying the longer it takes. In the limit where v --> c they never converge according to the observer.

Dale
Mentor
2021 Award
I wonder what you think of the following thought experiment by Dimitry67 that seems to show that parallel beams of light will not converge. Thus it is as though light can't be a source of gravity.
Dimitry67 is correct, but you should not take the quote out of context. The bolded conclusion above does not follow, as you can clearly see by considering the full quote:

This is correct, light beams create gravity.

However, when it was discussed here about 1 or 2 y ago, I remember that someone (with much deeper knowledge of GR - I am just a layman) told me that:

2 parallel light beams going in the same direction do not attract (even they attract the surrounding objects)
2 parallel light beams going in opposite directions do attract.

The first fact might be clear if we look at 2 massive objects, separated by some distance, flying in the same direction. In their inertial system they collide, say, in 1s. For an external observer, this process would take longer because of the time dilation. The faster 2 objects are flying the longer it takes. You can think about the case N1 as a limit where v --> c (it takes forever)

So,

What percentage of the universe's net gravitational field would be due to electromagnestism?

GrayGhost

Dimitry67 is correct, but you should not take the quote out of context. The bolded conclusion above does not follow, as you can clearly see by considering the full quote:

I stand corrected - I did take Dimitry67's argument out of context.

I did it for two personal reasons I guess.

1/ I like the argument.

2/ I have a "pet theory" that the inertia of a particle with rest mass is caused by retarded gravitational waves impinging on the particle from all the other massive particles in the Universe. I believe that initially massless particles respond to this by following circular orbits whose rotation energy gives half the mass/energy of the particle (the rest being in the mutual gravitational energy between the particle and the rest of the Universe). I am trying to formalise Mach's Principle. I want to argue that light is different so that it does not pick up an inertia. I probably need to think about my theory more before I can decide what it says about light. I'm not using GR itself but an approximation to it called gravitomagnetism that is like Maxwell's theory. In fact I should probably just stick to trying to understand electromagnetically induced inertia for the moment as i feel on safer ground with EM. (I only have at best an undergraduate understanding of physics).

Last edited:
Dale
Mentor
2021 Award
I have a "pet theory" ... I only have at best an undergraduate understanding of physics
Hmm.

Dale
Mentor
2021 Award
Threadmark is wrong. Was there any ambiguity?

bcrowell
Staff Emeritus
Gold Member
What percentage of the universe's net gravitational field would be due to electromagnestism?

If you mean the gravitational field as defined in Newtonian mechanics, that isn't an unambiguously well defined thing in GR. By the equivalence principle, the gravitational field at a given point can be anything you like, depending on your frame of reference. In the frame of an observer at rest with respect to the cosmic microwave background, the gravitational field is zero by symmetry (in a homogeneous and isotropic cosmological model).

Interpreting your question more loosely, the answer is that the universe was radiation-dominated at one time, then matter-dominated, and is now vacuum-dominated.

No,

All I meant was that if photons gravitate, then they produce gravitation. The cosmos has some net collective gravitational feild. Given photons are everywhere, I was just wondering what percentage of the cosmic gravitational field would owe to EM?

GrayGhost

bcrowell
Staff Emeritus