PHYSICS RESULTS

Summary

The GammeV experiment is a gamma to milli-eV axion-like particle (ALP) search using a “light shining through a wall” technique.

We have taken data in a scalar (or pseudoscalar) configuration depending on whether we rotate (or not) the polarized light from the laser. We have taken data in two configurations of our plunger … “wall” in the center or with the “wall” located 2m away from the center of the 6m magnet.

So, in the end, we will have four data samples to search for a signal, two polarizations and two plunger positions. We do not see candidates above the expected background levels and therefore can set limits on the axion-like particle coupling to photons as a function of the mass of an axion-like particle. We completely exclude (by more than 5σ) the region of interest suggested by the PVLAS anomaly, we extend and fill in regions of insensitivity on the previous BFRT result.

LEAKY Mirror Data

Click here to learn more details about our “Leaky Mirror” calibration running where we attenuate our high power laser down to detectable single photons to provide the absolute timing window for our signal and to verify the laser polarization states. When correcting for offsets, the signal window is about 50ns after the laser fires due to the speed of light time to travel through our apparatus.

Findings

Here is a table of our results

The four configurations represent whether we use horizontally or vertically polarized photons to test the scalar or pseudoscalar axion-like particle hypothesis. The number of photons in each configuration calculated from continuously monitoring the power reading on the laser pulse that returns into the laser box after reflecting off of the mirror (“wall”). We correct for measured intensity losses in the entrance vacuum window. The numbers makes sense since each pulse is a few x 10¹⁷ and the laser pulses at 20Hz and each configuration collected about 20 hours of data. The estimated background comes from looking in a 10000ns wide window around our signal region and counting the number of background photons that are observed. Since our search window is only 10ns wide, we divide the typically 1600 background events by 1000 to obtain a typical expected background of 1.6 events in a 10ns wide search window. The Candidates are the number of candidate events that fall within our 10ns wide search window for each of the configurations. The number of the candidates can be compared with background expectations. If the central value of the PVLAS anomaly were due to axion-like particles, we would have expected ~150 candidates in either the horizontal or vertical configurations. The final colume represents the limit that we set for the largest possible axion-like particle to photon coupling constant that is excluded at 3σ confidence in the limit of a very small axion-like particle mass.

Here is a figure that shows these results

The figure shows our data in the four configurations in a 100ns wide window approximately centered about our search window. The lowest panel shows the leaky mirror calibration data which is exactly what a real signal would look like. The leaky mirror data is at 50ns which is where you would expect those photons to be since they travel at the speed of light. The shape of the leaky mirror data is the expected shape due to an approximately 5ns wide laser pulse. The two vertical bands represent our search window. Between these bands, we count the number of possible signal candidates. No excess above the background expectations are observed. Outside the bands shows some of the background events that do not fall within the search window.

Here is a figure that show the limits

The figures show the 3σ limits that we set using our data for the coupling of an axion-like particle to photons for either the scalar (left) or pseudoscalar (right) hypothesis as a function of the mass of a new particle. Shown on the figure are the previous regeneration results from BFRT (beige) that show regions of insensitivity due to the specific length magnet that was used. The PVLAS anomalous results as interpreted as an axion-like particle is shown in the (pink) band. For the data taken when the plunger position was near the middle of the magnet, we exclude the region above the (dashed blue) curve. It has a double spike feature because the position of the plunger was actually 10cm away from the center of the magnet. For the data taken when the plunger was near one of the ends, we exclude the region above the (dashed red) curve. Using the sensitivity from these two plunger positions, we exclude the region above the (black) curve where we follow a statistical procedure to combine limits (including systematic uncertainties) from both plunger positions. We clearly exclude the PVLAS suggested region of interest that happened to fall in the region of minimal sensitivity of the previous BFRT data. We also extend the excluded BFRT region by several factors over the entire mass range. In the case of limits for the pseudoscalar case, we also show the 3σ exclusion from the published BMV experiment.

Other Results:

The GammeV ALP measurements have been used to set limits on massive paraphotons. [link]