⚛️ high-energy experiments

NTL amplification in a SuperCDMS-style detector fabricated from a University of South Dakota-grown germanium crystal

This study demonstrates that University of South Dakota-grown germanium crystals can successfully achieve Neganov-Trofimov-Luke amplification in SuperCDMS-style detectors, provided that high-quality crystal growth is paired with the deposition of aluminum bias contacts to enable voltage-dependent energy gain.

Original authors: Dongming Mei, Gerardo Gonzalez, William Baker, Mark Platt, Nader Mirabolfathi

Published 2026-10-22

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Dongming Mei, Gerardo Gonzalez, William Baker, Mark Platt, Nader Mirabolfathi

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Deep beneath the Earth's surface, in the quietest corners of the laboratory, scientists are listening for the faintest whispers of the universe. They are searching for dark matter, an invisible substance that makes up most of the universe's mass but refuses to interact with light, and for neutrinos, ghostly particles that stream through everything in vast numbers. To catch these elusive visitors, researchers use detectors made of ultra-pure crystals, cooled to temperatures colder than deep space. When a particle strikes the crystal, it deposits a tiny amount of energy, creating a ripple of heat and a burst of sound waves that travel through the solid material. The challenge is that these ripples are so small they are often drowned out by the background noise of the laboratory itself. To hear the whisper, scientists need to amplify the signal without adding their own noise. They do this by applying an electric voltage across the crystal. As the electrical charge created by the particle strike moves through the crystal under the influence of this voltage, it gains energy and creates extra sound waves, effectively turning a faint tap into a loud knock that the sensors can easily hear.

A team of researchers recently tested a new source of these crystals to see if they could work in such a sensitive setup. The crystals were grown at the University of South Dakota, a program dedicated to producing high-purity germanium, a material that is difficult to make but essential for these experiments. The team built two detectors using these crystals, shaping them into flat cylinders about the size of a large dinner plate but only an inch thick. They equipped the detectors with special sensors capable of feeling the faintest vibrations and placed them in a refrigerator cooled to ten millikelvin, a temperature just a fraction of a degree above absolute zero. The goal was simple: could a crystal grown at this university, when fashioned into a detector with a high-voltage electrode, successfully amplify the signal from a particle strike?

The first attempt did not go as planned. The researchers initially tried to apply the electric voltage without touching the crystal directly, using a method that relies on the electric field to reach through the air or a gap. In this setup, neither of the two detectors showed any sign of amplification. The signals remained faint, and the voltage seemed to have no effect. This result could have been interpreted as a failure of the crystal itself, suggesting that the material grown at the university was too impure or flawed to carry the necessary electrical charge. However, the team suspected that the problem might lie not in the crystal, but in how the electricity was being delivered.

To test this, they modified the second detector. They deposited a thin, continuous film of aluminum directly onto the back of the crystal, creating a solid electrical contact. This change transformed the experiment. When they applied the voltage again, the detector came alive. As they increased the voltage from zero to forty volts, the signal from the particle strikes grew dramatically. A signal that started at a tiny level swelled to nearly nine times its original size. The researchers measured the energy of the particles and found that the voltage was successfully boosting the sound waves generated by the moving charges. This confirmed that the crystal itself was perfectly capable of the job; it was the way the electricity was applied that had been the obstacle.

The success of this experiment is a significant step forward for the field. It proves that crystals grown at the University of South Dakota can be processed, fitted with sensors, and operated in the extreme cold required for these experiments. It shows that the material is good enough to carry the electrical current needed to amplify the signals from dark matter and neutrinos. However, the researchers are careful to note that this is a proof of concept, not a finished product. The experiment worked well on one specific channel of the detector, but not all parts of the device performed equally, and the team still needs to refine the manufacturing process to ensure every part of the crystal works perfectly. They also need to better understand how the signals are collected and to eliminate the remaining sources of noise.

This work highlights a crucial lesson for building these sensitive instruments: the quality of the final detector depends on both the raw material and the craftsmanship. A perfect crystal can fail if the electrodes are poorly designed, just as a flawed crystal can sometimes be made to work with the right engineering. By demonstrating that a university-grown crystal can produce the necessary amplification, the team has opened the door for future experiments to use these materials. It suggests that the path to catching the universe's most elusive particles may not require a single, massive supplier of crystals, but rather a diverse community of researchers capable of growing and shaping their own. The journey from a raw crystal to a working detector is long and complex, but this experiment has shown that the first, most critical link in the chain is strong.

Technical Summary: NTL Amplification in a SuperCDMS-style Detector Fabricated from a University of South Dakota-grown Germanium Crystal

Problem Statement
Low-mass dark matter searches and coherent elastic neutrino–nucleus scattering (CE $\nu$ NS) experiments require cryogenic semiconductor detectors capable of resolving energy deposits in the eV–sub-keV range. The Neganov–Trofimov–Luke (NTL) effect provides a mechanism for internal amplification of phonon signals by applying a bias voltage, allowing charge carriers to generate additional phonons as they drift. However, the successful observation of NTL gain depends on a complex interplay between bulk crystal properties (impurity concentration, carrier transport, trapping) and detector fabrication (electrode stability, surface leakage, phonon collection). A critical challenge for future large-scale experiments, such as SuperCDMS SNOLAB, is establishing a domestic supply of high-purity germanium (HPGe) crystals that are compatible with high-voltage (HV) phonon detector architectures. Previous studies indicated that crystal quality alone does not guarantee NTL performance; device fabrication and boundary conditions are equally critical. This paper addresses the specific question of whether germanium crystals grown at the University of South Dakota (USD) can support charge drift and generate measurable NTL phonons when integrated into a SuperCDMS-style detector.

Methodology
The study utilized two USD-grown germanium substrates with nominal dimensions of 75 mm diameter and 10 mm thickness, (100) orientation, and a net electrically active impurity concentration of approximately $10^{10}$ cm $^{-3}$ . The etch-pit density ranged from 300 to 7000 cm $^{-2}$ . Both crystals were instrumented as four-channel athermal-phonon detectors using Transition Edge Sensor (TES) / Quasiparticle-Excitation Transducer (QET) technology, operating at a bath temperature of approximately 10 mK.

The experimental campaign proceeded in two phases:

Contact-Free Bias Tests: Both detectors were initially tested with a contact-free electrode configuration. Detector 1 was tested at 0, 10, and 20 V, and Detector 2 at 0, 5, 10, and 15 V. No resolvable NTL gain was observed in either case.
Contacted-Electrode Final Run: For the final phase, Detector 2 was modified by depositing a continuous 600 nm aluminum bias film on the face opposite the phonon sensors using a modified Perkin Elmer 4400 sputter coater. This replaced the capacitive boundary condition with a direct electrical contact. The detector was irradiated with a 1 $\mu$ Ci $^{241}$ Am source (positioned near channel B) and tested at bias voltages of 0, 5, 10, 15, 20, 30, and 40 V.

Data acquisition involved digitizing waveforms at 1 MHz. Event selection utilized optimal filtering for timing and low-pass-filter (LPF) amplitude extraction for pulse height. The analysis focused on the dominant peak in the LPF amplitude spectrum of Channel B, fitting Gaussian distributions to determine centroid shifts. The absorbed energy ( $E_{abs}$ ) was reconstructed from the current pulse integral, and the apparent total phonon collection efficiency ( $\eta_{tot}$ ) was calculated relative to the ideal NTL energy.

Key Contributions and Results
The primary contribution of this work is the demonstration of NTL amplification in a USD-grown crystal within a SuperCDMS-style detector architecture, following a transition from a null result in a contact-free configuration to a positive result in a contacted configuration.

Observation of Gain: In the contacted configuration, Detector 2 exhibited clear, voltage-dependent NTL amplification. The reconstructed Channel B absorbed energy increased from approximately 3.9 keV at 0 V to 34.9 keV at 40 V. This corresponds to an enhancement factor of approximately 8.9.
Statistical Precision: The statistical uncertainty on the fitted LPF peak centroids was below 0.3% across all seven bias settings, confirming that the observed displacement is robust against statistical noise.
Phonon Response Coefficients: A phenomenological fit to the low-bias data (0–20 V) yielded effective prompt-phonon and NTL-phonon response coefficients of $(6.77 \pm 0.82)\%$ and $(4.77 \pm 0.20)\%$ , respectively. The quoted uncertainties are regression standard errors and do not include full calibration or systematic uncertainties.
Channel Asymmetry: Significant gain was observed only in Channel B, which was positioned directly under the radiation source. Channel C showed little gain, consistent with position-dependent phonon partition and the localized nature of the source.
Efficiency Trends: The apparent total phonon collection efficiency ( $\eta_{tot}$ ) decreased from 6.56% at 0 V to 4.09% at 40 V, indicating that prompt and NTL phonons are not collected with identical efficiency in this geometry.

Significance and Claims
The paper claims proof-of-principle compatibility between USD-grown germanium and SuperCDMS-style high-voltage phonon detectors. The successful generation of NTL phonons demonstrates that the tested crystal possesses the necessary bulk properties (sufficient carrier mobility and low trapping) to support charge drift through an applied potential when integrated into a properly fabricated device.

Crucially, the paper emphasizes that crystal quality and detector fabrication are inseparable in NTL testing. The initial null results in the contact-free configuration did not indicate a failure of the crystal material; rather, the transition to a positive result upon adding a continuous aluminum contact highlights that electrode geometry, surface boundary conditions, and leakage control are decisive factors. The authors state that a null NTL result in one configuration does not disqualify a crystal, while measurable gain in a functioning device serves as a direct functional test of the material-device combination.

The work is explicitly framed as a foundational milestone rather than a production qualification. The authors note limitations including the reliance on a single functional channel (Channel B), the absence of direct leakage current monitoring, incomplete traceability of crystal IDs and boule positions, and the lack of a full systematic uncertainty budget for energy reconstruction. Consequently, the results support continued fabrication studies using USD-grown crystals but do not yet constitute a general acceptance test for all such crystals or a fully qualified SuperCDMS detector. The paper concludes that future validation requires controlled, traceable fabrication protocols on multiple crystals with full channel yield and comprehensive uncertainty propagation.

More like this