Key Technical Issues and Solutions in Quantum Yield Measurement

1. Introduction

Quantum yield is the core parameter characterizing the optoelectronic conversion efficiency of luminescent materials, and is widely used in research and quality control of LED phosphors, quantum dots, organic luminescent dyes, perovskite materials and other fields. Accurate quantum-yield measurement is by no means a simple matter of "shine and calculate"; in practice, integration-time setting, radiometric calibration, self-absorption correction, scattering subtraction and many other steps can introduce significant systematic errors. Based on years of experience in instrument development and user support, this article analyzes one by one the most common technical problems in quantum-yield measurement and provides actionable solutions to help experimenters avoid detours and obtain reliable data.

2. Reasonable Setting of Integration Time

2.1 Effects of an excessively long integration time

Longer integration time is not always better. The direct consequences of an excessively long integration time include:

  • Reduced acquisition efficiency: The total acquisition time equals the integration time multiplied by the number of averages; when the integration time is set above 1 second, the single-measurement period is markedly prolonged and the experimental throughput is greatly reduced.
  • Electronic noise accumulation: The dark-current noise of CCD or CMOS detectors grows linearly with integration time; long integration lowers the signal-to-noise ratio, which is especially unfavorable for weak-signal measurements.
  • Overexposure risk at high drive: Under high drive voltage or high brightness, an excessively long integration time easily saturates the detector, causing overexposure; the spectral peaks are then flattened and the measurement result is completely distorted.

2.2 Effects of an excessively short integration time

An excessively short integration time is equally undesirable:

  • Underexposure of weak light: With a low-brightness signal and an overly short integration time, the intensity value is too low and the signal-to-noise ratio is poor; the spectral curve is visibly noisy, affecting the accuracy of subsequent optical-power and photon-count calculations.
  • Increased quantization error: The ADC has a finite number of bits; an overly low signal value reduces the number of effective bits and raises the proportion of digitization noise.

In practice, the following rules of thumb can be followed:

  1. Keep the integration time within 500ms as much as possible, with a recommended range of 300-500ms.
  2. Light the LED at maximum voltage, observe the raw spectral intensity value, and adjust the integration time so that the intensity falls between 100,000 and 200,000 counts. This range keeps the signal well away from the saturation region while providing sufficient dynamic range.
  3. If the intensity still does not reach 100,000 counts at maximum voltage, directly choose an integration time in the 300-500ms range to ensure the spectral curve is clear and discernible.
  4. For back-illuminated CCD spectrometers such as the Qepro, the shortest integration time is about 8ms; in practice it is recommended to set it above 10ms to ensure stable triggering and acquisition.
  5. High-voltage, high-brightness scenarios: If the light source is extremely bright (such as a high-power LED or laser excitation), it is better to reduce the integration time than to lower the drive voltage, so as to keep the source operating in its linear region. In this case, start from 10ms and gradually increase until the signal is appropriate.

2.4 Software acquisition-flow considerations

Ensure the measurement-flow order is correct: first set the measurement parameters, then set the drive voltage, wait until the light-source output is stable, and only then trigger spectrum acquisition and save the data. Do not acquire spectra during the transition period before the voltage has stabilized, otherwise the data cannot truthfully reflect the steady-state characteristics of the source.

3. Common Problems in Radiometric Calibration

3.1 Why radiometric calibration is needed

A fiber-optic spectrometer measures relative intensity values (counts) after ADC conversion; the spectral response distribution is affected by optical-path elements such as grating efficiency, detector quantum efficiency, fiber transmittance and coating characteristics, so the relative spectral distributions of spectrometers from different manufacturers and models all differ. The purpose of radiometric calibration is to convert relative intensity values into absolute optical-power distributions with physical meaning.

In short, the physical meaning of the radiometric calibration coefficient $f(\lambda)$ is: the optical-power value corresponding to a unit count at each wavelength within a unit integration time. With $f(\lambda)$, the relative spectrum $S(\lambda)$ measured at any integration time $t$ can be converted into the absolute optical power $I(\lambda)$:

$$ I(\lambda) = f(\lambda) \times \frac{S(\lambda)}{t} $$

3.2 Stability of the calibration light source

The stability of the radiometric calibration source (such as the HL-3plus-CAL series) directly determines calibration accuracy. Before use, ensure that:

  • The calibration source has been fully warmed up (usually 15-30 minutes) and its output has reached thermal equilibrium.
  • The lamp file (calibration file) of the calibration source is intact and matches the current lamp; different lamps must not be mixed.
  • The output of the calibration source is periodically checked for drift with a standard detector. As a tungsten lamp ages, its short-wave output gradually declines; if the deviation exceeds 2%, consider updating the calibration or replacing the bulb.

3.3 Calibration frequency

As long as the optical-path components (spectrometer, fiber, integrating-sphere interface) are unchanged, a single radiometric calibration can be used for a long time. However, recalibration is required in the following cases:

  • Replacing or re-plugging the fiber
  • Replacing the grating or slit
  • Moving the system or subjecting it to mechanical shock
  • The calibration-source bulb has accumulated more than 500 hours of use
  • The ambient temperature changes by more than ±5°C

It is recommended to keep a calibration log and save the calibration spectrum and coefficient file after each calibration for later traceability and comparison.

4. Identification and Correction of the Self-Absorption Effect

4.1 Phenomenon and causes

The self-absorption effect is one of the most insidious error sources in fluorescence quantum-yield measurement. When the sample concentration is high or the path length is long, part of the fluorescence emitted by the sample is re-absorbed by the sample itself before leaving the integrating sphere, causing the measured number of emitted photons to be too low and the PLQY result to be underestimated.

Typical manifestations of self-absorption:

  • A visible dip or shape distortion on the short-wavelength side of the emission spectrum (the region overlapping the absorption spectrum)
  • For the same sample at different concentrations, the measured PLQY decreases systematically as the concentration increases
  • Red shift of the emission peak position (short-wavelength components are preferentially absorbed)

4.2 Correction methods

For different degrees of self-absorption, the following strategies can be tried in order:

  1. Dilution method (first choice): Dilute the sample solution step by step, measure the PLQY at different concentrations, and extrapolate to the limiting value at zero concentration. This method is direct and effective and is suitable for solution samples.
  2. Thin-layer sample preparation: For solid powders, use the thinnest possible sample layer to reduce the transmission path length of fluorescence inside the sample. The powder can be uniformly dispersed in an inert matrix (such as BaSO₄) and pressed into a pellet.
  3. Mathematical correction models: When self-absorption cannot be fully eliminated experimentally, a correction model based on the overlap integral of the absorption and emission spectra can be used. The iterative correction method proposed by Mével et al. is widely used in the literature; the core idea is to estimate the self-absorption loss from the absorption coefficient and sample geometry parameters and correct the measured PLQY accordingly.

4.3 Practical suggestions

In routine measurement, if the PLQY is found to be sensitive to concentration, self-absorption should be suspected first. Whether self-absorption exists can be verified by comparing the PLQY at different excitation wavelengths (which changes the penetration depth).

5. Effect of Excitation-Light Scattering on Absorbance Calculation

5.1 Problem description

In the integrating-sphere method, the absorbance is calculated by comparing the excitation-light scattering signal with and without the sample (or with a blank substrate). However, the difference in scattering characteristics between the sample and the blank substrate — for example, differences in sample surface roughness, particle size or refractive index — causes the scattered fraction of excitation light to differ, in turn making the calculated absorbance deviate from the true value.

5.2 Handling strategies

  • Use a matched blank substrate: The physical form of the blank substrate should be as close as possible to the sample. Use pure solvent for liquid samples, a blank pellet for solid powders, and a blank substrate for film samples.
  • Subtract the scattering background: Subtract the excitation-light scattering by comparing with the scattering signal of the blank substrate. Note that the integration region of the scattering peak should cover the complete excitation-wavelength range to avoid missing the scattering wings.
  • Verify the self-consistency of the absorbance: For the same batch of samples, the absorbance measured at different excitation wavelengths should be consistent; if significant deviation appears, it indicates a problem with the scattering subtraction.

6. Maintenance and Calibration of the Integrating Sphere

6.1 Daily maintenance of the PTFE integrating sphere

The high-purity PTFE (polytetrafluoroethylene) sintered integrating sphere has good diffuse-reflectance characteristics and chemical inertness, but it is easily contaminated in actual use:

  • Dust protection: Dust adhering to the inner wall lowers the local reflectance and disrupts the spatial uniformity of the light field. When not in use, always cover the sample port and viewing port to avoid leaving them open for long periods.
  • Splash and moisture protection: Liquid samples must be placed in sealed cuvettes; never allow liquid to splash onto the sphere's inner wall. Sweat and conductive powders can also seriously contaminate the wall and even create a risk of electrical short circuit.
  • Do not touch the inner wall: PTFE is soft; any scraping by hard objects causes irreversible damage. Handle samples gently and do not touch the wall with your fingers.
  • Cleaning method: If there is floating dust on the inner wall, gently blow it off with dry, clean compressed air; if there is slight contamination, gently wipe with a cotton swab moistened with anhydrous ethanol and air-dry naturally. Never soak or clean with organic solvents (such as acetone), which would destroy the microporous structure of the PTFE.

6.2 Spectral-response calibration of the integrating sphere

The reflectance of the integrating-sphere wall varies with wavelength (PTFE reflectance declines in the ultraviolet region); in addition, geometric factors such as sphere openings and baffle position also introduce spectral bias. PLQY measurement systems are usually calibrated in the following ways:

  • Use a standard white plate with known reflectance (such as a Spectralon standard reflectance plate) to normalize the spectral response of the integrating sphere.
  • Load the integrating-sphere calibration file in the software to automatically correct system deviations at different wavelengths.
  • Periodically (e.g., quarterly) verify the system with standard fluorescent samples (such as rhodamine 6G and sodium fluorescein, whose PLQY values are known) to ensure the calibration remains valid.

7. The Re-excitation Effect

7.1 What is the re-excitation effect

In integrating-sphere measurement, the fluorescence emitted by the sample, after multiple diffuse reflections in the cavity, may strike the sample again and be absorbed, thereby exciting additional fluorescence emission. This cascade process of "fluorescence exciting fluorescence" is the re-excitation effect, which causes the measured number of emitted photons to be inflated and the PLQY to be overestimated.

7.2 Severity and judgment

The severity of the re-excitation effect depends on:

  • The size of the sample's Stokes shift: the smaller the shift, the more the absorption and emission spectra overlap and the greater the re-excitation risk.
  • The sample's quantum yield: the higher the PLQY, the stronger the emitted fluorescence and the greater the probability of re-excitation.
  • The integrating-sphere reflectance: the higher the wall reflectance, the greater the diffuse flux and the more re-excitation cycles.

How to judge the re-excitation effect: if the PLQY of the same sample measured at different excitation powers rises with increasing power, or if the PLQY measured in a high-reflectance integrating sphere is markedly higher than in a low-reflectance sphere, re-excitation may be present.

7.3 Mitigation measures

  • Using an integrating sphere with appropriately reduced wall reflectance (such as a partially sprayed gray coating) can weaken re-excitation, but the signal-to-noise loss must be weighed at the same time.
  • Reduce the excitation power to keep the fluorescence emission intensity at a lower level.
  • If the Stokes shift is extremely small (<20nm), consider a non-integrating-sphere method (such as an alternative scheme based on an optical integrating cavity).

8. Dealing with Oxygen Quenching

8.1 Mechanism of oxygen quenching

Ground-state oxygen molecules (³O₂) are highly efficient fluorescence quenchers that can undergo energy transfer or electron transfer with excited fluorescent molecules, causing the excited state to deactivate non-radiatively and lowering the PLQY. This effect is especially pronounced in phosphorescent materials, and the PLQY of many fluorescent materials is also highly sensitive to oxygen.

8.2 Experimental countermeasures

  • Inert-atmosphere measurement: For oxygen-sensitive samples, measure in an inert-gas glove box (N₂ or Ar environment, O₂ < 1 ppm). A glove box with fiber feedthroughs can be used, with the integrating sphere and sample placed inside and the spectrometer connected by fiber outside.
  • Solution deoxygenation: Liquid samples can be deoxygenated by bubbling high-purity N₂ or Ar for 15-30 minutes, followed by measurement in a sealed cuvette.
  • Comparative verification: Measure the same sample in air and in an inert atmosphere; if the PLQY differs significantly, the sample exhibits oxygen quenching and the data under inert atmosphere should prevail.

9. Sample Preparation Considerations

9.1 Liquid samples

  • Use quartz cuvettes with high transmittance and low fluorescence, and avoid ordinary glass cuvettes, which produce background fluorescence under ultraviolet excitation.
  • The solution concentration should be moderate: too low gives a weak signal and poor signal-to-noise ratio; too high causes severe self-absorption. It is recommended to determine the optimal concentration through preliminary experiments — an emission intensity between 50,000 and 150,000 counts is appropriate.
  • The outer walls of the cuvette must be wiped clean; fingerprints and dust scatter the excitation light and interfere with absorbance calculation.
  • Before each measurement, use pure solvent as a blank control and subtract the solvent background.

9.2 Solid film samples

  • The film should be flat and uniform in thickness, avoiding wrinkles and bubbles.
  • Use a blank substrate (such as a quartz plate) for control measurement to ensure that the fluorescence contribution of the substrate itself has been subtracted.
  • Face the front of the film toward the excitation-light incidence direction and keep the sample mounting position consistent.

9.3 Powder samples

  • The powder should be thoroughly ground to ensure uniform particle size; large particles cause severe scattering.
  • Use a powder holder or pellet-pressing method to make the powder into a flat measurement surface; a loose powder pile makes scattering uncontrolled.
  • For strongly absorbing powders, mix and grind them with an inert diluent (such as BaSO₄ or KBr powder) in a certain ratio, press into a pellet and measure, then extrapolate the concentration.
  • Pay attention to the consistency of the pressing pressure; the packing density of the powder differs at different pressures, affecting the scattered fraction.

10. Error Analysis in Data Processing

10.1 Dark-noise subtraction

Before each measurement, a dark spectrum should be acquired (turn off the excitation source or block the entrance port) and subtracted during data processing. The integration time and number of averages of the dark spectrum should be consistent with those of the sample measurement. Improper dark-noise subtraction is one of the most common mistakes made by beginners; it directly affects baseline accuracy and the calculation of weak signals.

10.2 Choice of wavelength range and integration regions

  • PLQY calculation involves integrating the numbers of excitation photons and emission photons over their respective bands. The boundaries of the excitation and emission regions should be chosen carefully to avoid cross-interference caused by overlap between the two.
  • The integration range of the emission region should cover all wavelengths from the red side of the excitation wavelength down to where the emission signal returns to zero; missing part of the integral makes the PLQY too low.
  • For grating spectrometers with second-order diffraction peaks, ensure that the emission integration region does not include the second-order diffraction signal of the excitation wavelength.

10.3 Multiple averaging and repeatability

  • Under the same measurement conditions, acquire at least 3 times and take the average to reduce random noise.
  • Different batches of samples should be prepared and measured independently to evaluate the batch-to-batch variability introduced by sample preparation.
  • When reporting PLQY, the standard deviation should be given at the same time to demonstrate measurement repeatability.

10.4 Temperature control

The PLQY of most fluorescent materials decreases as temperature rises, due to the increased rate of non-radiative transitions and enhanced molecular-vibration deactivation. If the ambient temperature of the experiment fluctuates greatly (e.g., a laboratory without air conditioning), the measurement temperature should be noted in the report. For precise measurement, a temperature-controlled sample holder is recommended to stabilize the temperature within ±1°C.

11. Summary and Best-Practice Recommendations

Quantum-yield measurement is a meticulous systems-engineering task; the factors affecting result accuracy run through the entire chain of "light source → integrating sphere → sample → detector → data processing". The key best practices are summarized as follows:

  1. Integration time: Keep it within 300-500ms, and adjust the integration time so that the spectral intensity falls in the 100,000-200,000 counts range, avoiding both saturation and underexposure.
  2. Radiometric calibration: Use a fully warmed-up, stable calibration source and, after calibration, keep the optical path fixed; recalibration is required at key points (changing fiber, changing grating, long periods of disuse).
  3. Self-absorption correction: Prefer diluting the sample or preparing a thin-layer sample; use a mathematical correction model when necessary.
  4. Scattering subtraction: Use a matched blank substrate and accurately subtract the excitation-light scattering signal.
  5. Integrating-sphere maintenance: Protect against dust, liquid and contact; periodically verify the system with standard samples.
  6. Re-excitation effect: Watch for re-excitation risk in samples with small Stokes shift and high PLQY; reduce the excitation power or change the integrating sphere when necessary.
  7. Oxygen quenching: Oxygen-sensitive samples must be measured in an inert atmosphere (N₂/Ar).
  8. Sample preparation: Use quartz cuvettes and optimize the concentration for liquids; grind powders thoroughly and press them uniformly; ensure films are flat and consistent.
  9. Data processing: Subtract dark noise properly, choose integration regions correctly, average multiple times, and report the standard deviation and measurement temperature.

Only by handling every step well can one obtain quantum-yield data that stand up to scrutiny.

Compiled and written by Pynect.