Spectral Response Analysis of Photovoltaic Solar Cells: Principles, Methods and Applications

Spectral response analysis is a fundamental diagnostic technique for photovoltaic cells, quantifying how efficiently a solar cell converts incident photons into electrical current across the full solar wavelength range. Performed in accordance with IEC 60904-8 and ASTM E1021, this testing provides critical insights into external quantum efficiency (EQE), internal quantum efficiency (IQE), and wavelength-dependent performance losses. This article explains the principles, sample types, test items, measurement methods, and practical applications of photovoltaic cell spectral response analysis for cell manufacturers, researchers, and quality engineers.

Technical Overview

Spectral response (SR) describes the photocurrent generated by a photovoltaic cell per unit of incident optical power as a function of wavelength, typically expressed in A/W (absolute spectral response) or as a normalized dimensionless curve (relative spectral response). Because sunlight is a broadband spectrum, the electrical output of a solar cell depends not only on the total incident power but also on how the cell responds to each wavelength band. Spectral response analysis is therefore indispensable for evaluating cell quality, diagnosing loss mechanisms, and enabling accurate energy-rating calculations.

The spectral response is closely related to external quantum efficiency (EQE), which defines the ratio of collected charge carriers to incident photons at each wavelength. The two quantities are linked by the relationship EQE(λ) = SR(λ) × 1240/λ, where λ is expressed in nanometers and SR in A/W. By additionally measuring the surface reflectance, the internal quantum efficiency (IQE) can be derived, isolating losses caused by recombination and incomplete absorption from those caused by optical reflection.

Key physical phenomena revealed by the spectral response curve include:

• Blue response (300–500 nm): sensitive to front-surface recombination, emitter doping profiles, and anti-reflection coating quality.
• Red response (700–1100 nm): sensitive to rear-surface recombination, bulk diffusion length, back-surface field (BSF) effectiveness, and light trapping.
• Near-infrared cutoff (beyond the bandgap edge, approximately 1100–1200 nm for crystalline silicon): indicates the effective bandgap and absorption properties of the absorber material.

Relevant international standards include IEC 60904-8 (measurement of spectral response of photovoltaic devices), ASTM E1021 (standard test method for measuring spectral response of photovoltaic cells), IEC 60904-7 (computation of spectral mismatch correction), and IEC 60904-2 (classification and requirements for reference cells). Measurements are typically performed by accredited third-party laboratories operating under ISO/IEC 17025 quality management systems.

Testing Samples

Crystalline Silicon Solar Cells

Monocrystalline and multicrystalline silicon cells, including mainstream architectures such as PERC, TOPCon, heterojunction (HJT), interdigitated back contact (IBC), and aluminum back-surface-field (Al-BSF) cells. The typical characterization range is 300–1200 nm. Full-format cells (e.g., 166 mm, 182 mm, 210 mm wafers), half-cut cells, and diced cell strips can all be measured.

Thin-Film Solar Cells

Amorphous silicon (a-Si), microcrystalline silicon (μc-Si), cadmium telluride (CdTe), and copper indium gallium selenide (CIGS) cells. These materials exhibit bandgaps between approximately 1.1 eV and 1.7 eV, requiring measurement ranges from 300 nm up to 1300 nm or beyond, and often require bias-light conditions matched to the cell's spectral sensitivity.

III-V and Multi-Junction Solar Cells

Gallium arsenide (GaAs) based single-junction and multi-junction (e.g., InGaP/GaAs/Ge) concentrator or space cells. These devices demand subcell-resolved spectral response measurements using selective bias light and bias voltage to activate each junction individually.

Emerging Photovoltaic Technologies

Perovskite solar cells, organic photovoltaic (OPV) cells, quantum-dot cells, dye-sensitized cells, and perovskite-silicon tandem cells. Tandem devices in particular require advanced measurement protocols with tunable bias illumination to deconvolute the spectral response of each subcell across an extended wavelength range of 300–1300 nm.

Reference Cells and Calibration Standards

Primary and secondary reference cells used for solar simulator calibration and spectral mismatch correction, traceable to international standards organizations such as IEC and national metrology institutes.

Testing Items

A comprehensive spectral response analysis program typically includes the following measurands:

• Absolute spectral response (A/W) as a function of wavelength.
• Relative (normalized) spectral response curve.
• External quantum efficiency (EQE) spectrum.
• Internal quantum efficiency (IQE) spectrum, derived with hemispherical reflectance measurement.
• Integrated short-circuit current density (Jsc) calculated under the AM1.5G reference spectrum (IEC 60904-3).
• Effective bandgap and near-infrared cutoff wavelength determination.
• Blue response and red response quantitative evaluation for process diagnosis.
• Reflectance and parasitic absorption analysis of front and rear surfaces.
• Current-matching verification for multi-junction and tandem cell subcells.
• Spectral response linearity verification with respect to irradiance level (per IEC 60904-10).
• Temperature-dependent spectral response for advanced characterization needs.

Testing Methods

Measurement Principle

The standard method uses a monochromatic probe beam of known irradiance generated either by a grating monochromator with a broadband light source, or by a filter wheel or tunable laser system. The cell under test is illuminated at each wavelength step, and the resulting photocurrent is recorded. A calibrated reference detector, typically a traceable silicon photodiode or reference cell, alternately measures the monochromatic power to determine the absolute photon flux. The spectral response is computed as the ratio of the cell photocurrent to the incident monochromatic power at each wavelength.

Bias Illumination and Lock-In Detection

For most modern cells, and always for thin-film and multi-junction devices, a white-light bias source approximating the AM1.5G spectrum is applied simultaneously. This ensures the cell operates under realistic carrier-injection conditions. The monochromatic probe beam is modulated by an optical chopper, and a lock-in amplifier extracts only the AC photocurrent component produced by the probe beam, rejecting the DC photocurrent generated by the bias light. For multi-junction cells, dedicated bias LEDs with wavelengths absorbed by the non-target subcells are used, combined with an appropriate bias voltage, to isolate the subcell under test.

Measurement Procedure

• Sample preparation: clean electrical contacts, mount the cell on a temperature-controlled vacuum or probe chuck (typically 25 °C), and define the illuminated test area with an aperture if required.
• System calibration: verify wavelength accuracy and signal linearity; measure the reference detector response.
• Data acquisition: scan the wavelength range (typically 300–1200 nm) in steps of 5–10 nm, recording the AC photocurrent and reference signal at each point.
• Computation: calculate absolute spectral response and EQE; if reflectance data are collected with an integrating sphere, compute IQE.
• Integration: convolve the spectral response with the IEC 60904-3 AM1.5G reference spectrum to obtain the short-circuit current density, and cross-validate against I-V measurements (IEC 60904-1).
• Uncertainty evaluation: report measurement uncertainty components including reference detector calibration, wavelength accuracy, stray light, chopper stability, and sample alignment.

Quality Control and Traceability

Accredited third-party laboratories follow IEC 60904-8 and ASTM E1021 procedures, maintain reference detectors calibrated with traceability to national metrology institutes, and operate under ISO/IEC 17025 management systems. Inter-laboratory comparisons and periodic verification against certified reference cells ensure the long-term reliability of reported results. Typical measurement uncertainties for EQE of well-behaved silicon devices are within a few percent relative, with wavelength accuracy better than ±2 nm.

Applications

Spectral response analysis serves a wide range of engineering, quality, and research purposes across the photovoltaic value chain:

• Process optimization: identifying front-surface passivation quality from blue response, and bulk or rear-surface recombination losses from red response, guiding emitter diffusion, passivation layer, and anti-reflection coating process tuning.
• Cell efficiency diagnosis: locating current-loss mechanisms that cannot be separated in standard I-V measurements, supporting root-cause analysis for low-performing production batches.
• New material and device development: characterizing perovskite, tandem, CIGS, CdTe, and III-V devices during R&D, including stability studies under UV exposure, thermal stress, and damp-heat aging.
• Multi-junction current matching: verifying that subcells in tandem and multi-junction architectures are current-matched under the operating spectrum, which is essential for maximizing device efficiency.
• Calibration and spectral mismatch correction: providing spectral data required for IEC 60904-7 mismatch corrections when calibrating solar simulators and reference cells, and supporting module power rating under IEC 61215 design qualification and IEC 61853 energy rating frameworks.
• Energy yield prediction: combining spectral response data with site-specific spectral irradiance models to improve annual energy yield estimates for PV power plants.
• Incoming inspection and supplier qualification: comparing spectral response signatures of purchased cells or wafers against agreed acceptance criteria in commercial contracts.
• Failure analysis: detecting degradation signatures such as yellowing encapsulant effects, UV-induced degradation (UVID), and light- and elevated-temperature-induced degradation (LeTID) in post-aging investigations.

FAQ

What is the difference between spectral response and quantum efficiency?

Spectral response expresses photocurrent per unit of incident optical power (A/W), while quantum efficiency expresses the number of collected electrons per incident photon (percentage or dimensionless ratio). They contain equivalent physical information and are directly convertible through the wavelength: EQE(λ) = SR(λ) × 1240/λ, with λ in nanometers and SR in A/W. Quantum efficiency is often preferred for device physics interpretation because each photon is weighted equally.

Why is external quantum efficiency lower than internal quantum efficiency?

EQE counts all incident photons, including those reflected from the cell surface or transmitted without absorption, while IQE counts only photons that actually enter and are absorbed by the absorber layer. The difference is therefore caused primarily by reflection losses at the front surface, grid shading, and incomplete long-wavelength absorption. Comparing EQE and IQE helps separate optical losses from recombination-related electrical losses.

Why is bias illumination necessary during the measurement?

Many photovoltaic materials, especially thin-film, perovskite, and multi-junction devices, exhibit nonlinear response at low injection levels, so measurements performed with only a weak monochromatic beam would not represent real operating conditions. A white bias light approximating one-sun AM1.5G illumination biases the cell at its actual operating point. The chopped probe beam and lock-in detection then separate the small probe-induced signal from the large DC bias photocurrent.

What wavelength range is typically measured for silicon solar cells?

For crystalline silicon cells, the standard range is approximately 300–1200 nm, covering the UV-blue edge through the silicon bandgap cutoff near 1100–1150 nm. For thin-film and tandem devices, the range may extend down to 300 nm and up to 1300–1800 nm depending on the absorber bandgaps. Extending measurements slightly beyond the bandgap edge is useful for evaluating free-carrier absorption and verifying stray-light performance of the instrument.

How can spectral response results be used to improve cell efficiency?

Regions of the EQE curve map directly to physical device regions: weak blue response points to front-surface recombination or poor emitter passivation, a dip in the visible range may indicate anti-reflection coating mismatch, and weak red response points to short bulk diffusion length or poor rear passivation. Engineers use these signatures to target specific process improvements, such as adjusting passivation layer thickness, emitter sheet resistance, or hydrogen annealing conditions.

How is the measurement validated against I-V test results?

The spectral response curve is integrated with the AM1.5G reference spectral irradiance defined in IEC 60904-3 to calculate the short-circuit current density Jsc. This value is cross-checked against the Jsc obtained from calibrated sun simulator I-V measurement per IEC 60904-1. Agreement within approximately 1–2% confirms the consistency of both measurements and the validity of the spectral mismatch correction applied.

What sample formats can be tested?

Formats range from small-area research cells (a few square millimeters) to full-size production cells of 210 mm or larger, as well as cut cells, mini-modules, and encapsulated samples with appropriate optical correction. Proper area definition, contact probing, and temperature control are required to ensure accurate and reproducible absolute results.

How long does a typical spectral response measurement take?

A single EQE scan over 300–1200 nm with standard resolution typically takes several minutes per sample, depending on the wavelength step size and signal averaging. Additional time is required for reference calibration, reflectance measurement when IQE is needed, and multi-bias procedures for tandem and multi-junction devices.