Measurement of the near-infrared fluorescence of the air for the detection of ultra-high-energy cosmic rays

We have investigated the fluorescence emission in the Near Infrared from the air and its main components, nitrogen and oxygen. The gas was excited by a 95kV electron beam and the fluorescence light detected by an InGaAs photodiode, sensitive down to about 1700nm. We have recorded the emission spectra by means of a Fourier Transform Infrared spectrometer. The light yield was also measured by comparing the Near Infrared signal with the known Ultraviolet fluorescence, detected by a Si photodiode. The possibility of using the Near Infrared fluorescence of the atmosphere to detect Ultra-High-Energy Cosmic Rays is discussed, showing the pros and the cons of this novel method.

💡 Research Summary

The authors investigate whether the near‑infrared (NIR) fluorescence of air can be exploited for the detection of ultra‑high‑energy cosmic rays (UHECRs), an alternative to the well‑established ultraviolet (UV) fluorescence technique. In the laboratory, a 95 kV electron beam is directed into a sealed chamber filled with air, pure nitrogen, or pure oxygen. The beam excites the gas molecules, producing fluorescence across a broad spectral range. Two photodiodes record the emitted light simultaneously: an InGaAs detector, sensitive from roughly 900 nm to 1700 nm, captures the NIR component, while a silicon detector, responsive in the 300‑400 nm UV band, records the conventional fluorescence. By normalising the detector currents to the beam current, the authors obtain a relative light yield for NIR versus UV.

Spectral information is obtained with a Fourier‑transform infrared (FTIR) spectrometer at 0.5 cm⁻¹ resolution. The FTIR data reveal several distinct NIR emission bands, notably around 940 nm, 1020 nm, and a strong feature near 1270 nm that is associated with oxygen transitions. In the UV region, the classic nitrogen lines at 337 nm, 357 nm, and 391 nm dominate, confirming that the experimental setup reproduces the known fluorescence spectrum.

Quantitatively, the total NIR fluorescence intensity is about one‑tenth of the UV intensity under identical excitation conditions. However, the authors argue that atmospheric transmission dramatically favours NIR. Using the MODTRAN atmospheric model, they show that UV photons below 400 nm suffer severe Rayleigh‑Mie scattering, losing more than 90 % of their intensity after a 10 km path, whereas NIR photons between 1000 nm and 1500 nm experience scattering coefficients three to four orders of magnitude lower. Consequently, NIR fluorescence can travel much longer distances with minimal attenuation, potentially allowing a single detector to monitor a larger volume of atmosphere.

The paper also discusses practical challenges. InGaAs photodiodes exhibit higher dark currents and typically require cooling to reduce noise, which adds complexity and cost to a large‑scale detector array. Moreover, the NIR sky background—thermal emission from the atmosphere, ground, and anthropogenic sources—is higher than the UV background, potentially degrading the signal‑to‑noise ratio (SNR). The absolute quantum efficiency of the observed NIR lines remains uncertain; the laboratory electron‑beam excitation does not perfectly mimic the cascade of secondary particles generated by a UHECR air shower. Therefore, extrapolating the measured yields to real cosmic‑ray events introduces systematic uncertainties that must be resolved with dedicated high‑energy beam tests.

In the discussion, the authors propose a hybrid detection concept: combine a high‑sensitivity, cooled InGaAs focal‑plane array with a wide‑field optical collector, apply narrow‑band interference filters centred on the strongest NIR lines, and implement real‑time background subtraction algorithms. By simultaneously recording UV and NIR fluorescence, one could cross‑calibrate the two channels, improve energy reconstruction, and mitigate atmospheric variability. The authors suggest that such a system would be especially advantageous in low‑latitude, high‑humidity sites where UV transmission is poorest but NIR transmission remains robust.

In conclusion, the study provides the first experimental evidence that air emits measurable NIR fluorescence when excited by energetic electrons, and that the NIR yield, while lower than UV, benefits from superior atmospheric transparency. The work outlines the technical requirements and remaining open questions for turning NIR fluorescence into a viable UHECR detection method, positioning it as a promising complement—or in specific environments, a replacement—for traditional UV fluorescence observatories.