Single-Photon Detector Efficiency Measurement: Testing Methods, Parameters, and Applications

Single-photon detector efficiency measurement determines the photon detection efficiency (PDE) and related performance parameters of ultrasensitive detectors used in quantum communication, LiDAR, fluorescence spectroscopy, and scientific instrumentation. Accredited third-party laboratories provide traceable testing and calibration services for SPADs, SNSPDs, and photomultiplier tubes in accordance with recognized standards such as ISO/IEC 17025. This article explains the tested parameters, measurement methods, sample types, and typical applications of single-photon detector efficiency determination.

Technical Overview

A single-photon detector (SPD) is an extremely sensitive photodetector capable of registering individual photons, far beyond the sensitivity of conventional photodiodes. The most important figure of merit for these devices is the photon detection efficiency (PDE), defined as the probability that a photon incident on the detector produces a registered count. It is typically expressed as:

PDE = (C − C_dark) / (Φ_ph × (1 − C × τ_d))

where C is the total count rate, C_dark is the dark count rate, Φ_ph is the incident photon flux, and the term (1 − C × τ_d) corrects for dead-time saturation. Accurate determination of this parameter is essential because the detection efficiency directly governs the secure key rate in quantum key distribution (QKD), the maximum range of photon-counting LiDAR, and the signal-to-noise ratio in time-correlated single-photon counting (TCSPC) instruments.

Efficiency measurement is technically demanding because it requires calibrated, attenuated optical fluxes at the single-photon level, traceable optical power references, precise Poisson statistics analysis, and strictly controlled detector bias and temperature conditions. Testing is typically performed by accredited third-party laboratories whose calibration capabilities comply with ISO/IEC 17025, using equipment traceable to national metrology institutes, and observing laser safety practices consistent with IEC 60825-1.

Testing Samples

Silicon Single-Photon Avalanche Diodes (Si SPADs)

Bare chips, packaged dies, or complete modules operating in the 400 nm to 1100 nm spectral range, in both linear-mode and Geiger-mode (quenching circuit included) configurations. These are the most widely submitted samples for efficiency characterization.

InGaAs/InP Single-Photon Avalanche Diodes

Telecom-wavelength detectors covering approximately 950 nm to 1700 nm, tested in gated mode or free-running mode with integrated quenching and thermoelectric cooling, as used in fiber-based QKD and near-infrared ranging systems.

Superconducting Nanowire Single-Photon Detectors (SNSPDs)

Cryogenically operated fiber-coupled or free-space-coupled devices with broad spectral coverage from the ultraviolet to the mid-infrared. System detection efficiency (SDE) is measured with the detector installed in a cryostat at its specified operating temperature and bias current.

Photomultiplier Tubes and Microchannel Plate Detectors

Photomultiplier tubes (PMTs), microchannel plate PMTs (MCP-PMTs), and silicon photomultipliers (SiPMs/MPPCs), characterized for quantum efficiency, gain, and counting linearity across their rated spectral range.

Complete Instruments and Subsystems

Single-photon counting modules, TCSPC acquisition systems, photon-counting LiDAR receivers, and QKD receiver assemblies, evaluated as integrated units to determine end-to-end system efficiency.

Testing Items

Depending on the detector technology and customer requirements, typical efficiency-determination test programs include the following parameters:

• Photon detection efficiency (PDE) or system detection efficiency (SDE) as a function of wavelength
• Spectral responsivity and relative spectral response curves
• Dark count rate (DCR) at specified operating conditions
• Afterpulsing probability and its dependence on hold-off time
• Dead time, recovery time, and count-rate linearity
• Timing jitter and instrument response function (IRF) width
• Efficiency dependence on excess bias voltage and operating temperature
• Active-area uniformity and geometric fill factor (for array and SPAD imagers)
• Breakdown voltage verification and operating point confirmation
• Optical return loss and coupling efficiency for fiber-coupled devices

Testing Methods

Attenuated Calibrated Laser Source Method

The most common approach uses a wavelength-stabilized laser whose average output power is measured with a traceable calibrated optical power meter. Calibrated optical attenuators reduce the flux so that the mean photon number per pulse, μ = P_avg / (f_rep × h × c / λ), is well below one, placing the detector in the true single-photon regime. The detected count rate is corrected for dark counts and dead-time losses, and the PDE is derived from Poisson photon statistics. Pulsed sources allow gate-synchronized measurement for gated InGaAs/InP SPADs.

Reference Detector Substitution Method

In the substitution configuration, the photon flux is alternately measured by a calibrated reference photodiode (traceable to national metrology institutes) and by the device under test at the same optical plane. This geometry-matched comparison suppresses uncertainties from beam divergence and connector losses, and is well suited to fiber-coupled detector modules. Measurement practices follow applicable parts of the IEC 60747 series for optoelectronic semiconductor devices and relevant GB/T photodetector test standards.

Heralded-Photon (Klyshko) Correlation Method

For the highest accuracy, twin photons generated by spontaneous parametric down-conversion are used. Detection of the herald photon in a reference channel guarantees the presence of exactly one photon at the device under test, enabling absolute efficiency calibration without relying on a separate calibrated detector or absolute attenuation values. This method significantly reduces systematic uncertainty and is frequently used to validate other calibration chains.

Cryogenic System Efficiency Testing for SNSPDs

Superconducting detectors are characterized inside closed-cycle or bath cryostats at temperatures typically below 4 K. Fiber-coupled optical feedthroughs deliver the calibrated flux, and detection efficiency is recorded versus bias current to identify the saturation plateau. The reported system detection efficiency includes all coupling losses, which is the parameter most relevant to end users.

Measurement Conditions and Uncertainty Evaluation

All measurements are performed under controlled environmental conditions with recorded temperature, humidity, and electromagnetic shielding. A complete uncertainty budget is prepared in accordance with ISO/IEC 17025, covering power calibration, attenuation uncertainty, wavelength accuracy, statistical counting error, and dead-time corrections. Typical expanded uncertainties (k = 2) for PDE determination range from approximately 1.5% to 5% relative, depending on wavelength and detector type. Test reports include measurement conditions, raw data, correction models, and traceability statements.

Applications

Verified detector efficiency data is critical wherever single-photon sensitivity determines system performance. Major application fields include:

• Quantum key distribution (QKD), where PDE directly affects secure key generation rate and transmission distance
• Photonic quantum computing and boson sampling experiments
• Time-correlated single-photon counting (TCSPC) for fluorescence lifetime imaging and spectroscopy
• Single-photon LiDAR and time-of-flight 3D ranging
• Astronomy, including adaptive optics, occultation timing, and lunar laser ranging
• Medical imaging, such as PET detector development and Cherenkov imaging
• Distributed fiber-optic sensing and quantum-enhanced metrology
• Semiconductor inspection and failure analysis using ultra-weak luminescence detection
• Research and development quality control for detector manufacturers and module integrators

FAQ

What is the difference between quantum efficiency and photon detection efficiency?

Quantum efficiency (QE) generally describes the probability that an absorbed photon generates a photoelectron, while photon detection efficiency (PDE) is the end-to-end probability that an incident photon produces a registered count. PDE therefore also includes absorption probability, avalanche triggering probability, and, for arrays, the geometric fill factor. For SNSPD systems, the equivalent reported quantity is usually the system detection efficiency (SDE), which additionally includes fiber-coupling losses.

Which wavelength ranges can be tested?

Standard laboratories cover approximately 400 nm to 1700 nm for SPAD and SiPM devices using the attenuated-source or substitution methods. SNSPD testing in cryostats can extend coverage from the ultraviolet through the visible and telecom bands into the mid-infrared, depending on the available calibrated sources and detectors.

What measurement uncertainty can be expected?

Typical expanded uncertainties of 1.5% to 5% relative (coverage factor k = 2) are achievable, depending on wavelength, flux level, detector dark count rate, and the chosen method. Heralded-photon calibration generally delivers the lowest systematic uncertainty.

How should samples be prepared and submitted?

Customers should provide the device datasheet, required operating conditions (bias voltage or current, gate width and frequency, temperature), requested wavelengths, and, for modules, the electrical interface and power supply details. Fiber-coupled devices should be submitted with connector specifications so that reference measurements can be made in a matched geometry.

Are the results traceable to national standards?

Yes. Reputable accredited third-party laboratories maintain optical power and wavelength references that are traceable to national metrology institutes, and calibration activities are conducted under an ISO/IEC 17025 quality system with documented traceability chains stated in the test report.

How long does efficiency testing usually take?

Standard single-wavelength PDE verification is typically completed within a few working days after sample acceptance. Full spectral sweeps, cryogenic SNSPD characterization, or multi-parameter test programs involving jitter, afterpulsing, and linearity may require one to two weeks, depending on sample count and measurement complexity.