Room Temperature The Self-Calibrating Dual-Mode Detector - Say Goodby to the Long Traceability Chain

Low cost, high-accuracy calibration standards are requested by the radiometry community [1]. In the EMPIR-funded project chipS·CALe, we are developing self-calibrating dual-mode detectors for high-accuracy optical power measurements to meet this need. The dual-mode detector combines two primary standard techniques (PQEDs and electrical substitution radiometer [2]) into one device. This means the detector can be calibrated against its own internal primary reference. This eliminates the usually long and cumbersome traceability chain, and makes the detector suitable for calibrations in remote locations.

The absorbing element for both measurement modes is an induced-junction photodiode [3]. A photodiode has internal losses that make it deviate from ideal responsivity. These losses vary with temperature and wavelength, and change over time as material properties of the photodiode change. As they affect the responsivity of the photodiode, the internal losses must be determined before high-accuracy measurements can be done. In the dual-mode detector these internal losses are determined by using thermal mode as a reference. In addition, an independent method for determining the internal losses is available, by fitting a charge carrier simulation model to IV curves [4].

To have the dual-mode detector work in thermal mode, special packaging of the photodiode is required [5]. The silicon photodiode is mounted on a carrier with a weak thermal heat link. The thermal design of the detector is optimised to minimise non-equivalence between optical and electrical heating, by the use of COMSOL heat transfer simulations. Vacuum conditions are necessary, as heat convection through air introduces complicated and unpredictable effects on the thermal equivalence. The largest contributor to non-equivalence is radiation losses, due to different thermal gradients in electrical and optical heating mode.

Halfway through the project, we have already reduced the type A uncertainty below the project aim of 0.05 % in room temperature, getting close to uncertainty levels comparable to primary standard cryogenic radiometers. We are continuously making improvements in thermal packaging, thermal readout, electrical readout and calculation algorithms, and the latest results will be presented at the conference.

This project 18SIB10 chipS·CALe has received funding from the EMPIR programme co-financed by the Participating States and from the European Union’s Horizon 2020 research and innovation programme.

Figure 1: Two of our dual-mode detector modules. During operation they are mounted in a trap configuration in the construction visible in the background, to minimise reflection losses.

References:

[1] CCPR Strategy Document, Ch. 5.2.1 https://www.bipm.org/utils/en/pdf/CCPR-strategy-document.pdf

[2] BIPM. Mise en pratique for the definition of the candela in the SI, Ch 5.1 http://www.bipm.org/en/publications/mises-en-pratique, 2019.

[3] T. E. Hansen. Silicon UV-Photodiodes Using Natural Inversion Layers. Physica Scripta, 18:471-475, 1978.

[4] J. Gran, T. Tran and T. Donsberg. Three dimensional modelling of photodiode responsivity. 14th International Conference on New Developments and Applications in Optical Radiometry (NEWRAD 2021)

[5] E. Bardalen, M. U. Nordsveen, P. Ohlckers, and J. Gran. Packaging of silicon photodiodes for use as cryogenic electrical substitution radiometer. 14th International Conference on New Developments and Applications in Optical Radiometry (NEWRAD 2021)