Experimental investigation of the Landau-Pomeranchuk-Migdal effect in low-$Z$ targets

In the CERN NA63 Collaboration we address the question of the potential inadequacy of the commonly used Migdal formulation of the Landau-Pomeranchuk-Migdal (LPM) effect by measuring the photon emission by 20 and 178 GeV electrons in the range of 100 MeV–4 GeV, in targets of low-density polyethylene, C, Al, Ti, Fe, Cu, Mo, and, as a reference target, Ta. For each target and energy, a comparison between simulated values based on the LPM suppression of incoherent bremsstrahlung is shown, taking multiphoton effects into account. For these targets and energies, we find that Migdal’s theoretical formulation is adequate to a precision of better than about 5%, irrespective of the target substance.

Figure 1

A schematic drawing of the setup used in the experiment. Two magnetic dipoles, each 2 m along the beam, are used at a low field, $B\simeq 0.16\mathrm{T}$, to reduce the influence of synchrotron radiation in the MeV region. This, combined with the necessity of deflecting the electrons outside the BGO calorimeter, into the lead glass calorimeter, forces a long lever arm. Right in front of the BGO, a veto scintillator ScV was mounted to reject events where the photon has converted or events where an electron may have interacted with the vacuum chamber, generating a shower. The total length of the setup was about 70 m.

Figure 2

The BGO spectrum for 580 MeV electrons obtained from ASTRID compared to a Geant4 simulation of the energy deposit in the BGO crystal.

Figure 3

BGO calibration made at ASTRID with electrons from 100 to 580 MeV and fitted with a linear function.

Figure 4

The BGO spectrum measured at a PMT HV of 480 V. One can observe the pedestal at channel 85, a 4.4 MeV peak from the AmBe source, and a broad muon peak from cosmics.

Figure 5

The peak channel as a function of PMT HV for the 4.4 MeV peak, the muon peak, and the pedestal. The AmBe data have been fitted to a function describing the PMT amplification in two intervals.

Figure 6

Geant4 simulations and measurements of the background radiation and the aluminum reference target. See the text for more details.

Figure 7

The measured (markers) and simulated (line) ratio between the aluminum reference and the background.

Figure 8

Detection efficiency found by dividing the measured radiation background to the simulated as described in the text. The markers are the actual ratio and the line is the smoothed efficiency used in the analysis.

Figure 9

Calculated radiation power spectra for 178 GeV electrons traversing a 1.97% $X_0$ carbon target. The radiation spectra are normalized to the target thickness in units of $X_0$. The Migdal line is valid for a semi-infinite target and calculated with the approximations of Stanev et al. BK refers to the theory of Baier and Katkov. The Blankenbecler and Drell curves (BD and $\mathrm{BD}\mathrm{\text{−}}\delta$) are calculated for a finite size target. The $\delta$ refers to formulas in [18] including an extra correlation term.

Figure 10

Calculated radiation power spectra for 20 GeV electrons traversing a 2.58% $X_0$ copper target. See caption of Fig. 9 for more information.

Figure 11

Power spectra of radiation emission from 178 GeV electrons penetrating amorphous targets. The spectra are normalized to the number of incoming electrons. The lower, red dotted line shows the simulated contribution from the background, with data points representing the values obtained in the experiment. The middle, blue line shows the simulated contribution from the targets, including the LPM effect, and with data points representing the values obtained in the experiment. The upper, black dashed line shows the simulated contribution from the targets, excluding the LPM effect. Only statistical uncertainties are shown.

Figure 12

178 GeV electrons in carbon target. See caption for Fig. 11. Additionally, simulations based on the BD and $\mathrm{BD}\mathrm{\text{−}}\delta$ theory are included.

Figure 13

20 GeV electrons in copper target. See caption for Fig. 11. Additionally, simulations based on the BD and $\mathrm{BD}\mathrm{\text{−}}\delta$ theory are included.