{
  "id": "quantum-efficiency",
  "title": "Quantum Efficiency",
  "lang": "en",
  "source_lang": "en",
  "summary": "Quantum efficiency measures how effectively a device or process converts incident or absorbed photons into useful charge carriers, emitted photons, or other specified outcomes.",
  "category": "science",
  "subcategory": "physics",
  "type": "concept",
  "aliases": [
    "QE"
  ],
  "generated_at": "2026-11-05",
  "model": "pro/gpt-6.1-sol",
  "url": "https://aiwiki.page/en/quantum-efficiency/",
  "translations": {
    "zh-Hans": {
      "url": "https://aiwiki.page/zh/quantum-efficiency/",
      "json": "https://aiwiki.page/zh/quantum-efficiency.json"
    }
  },
  "sections": [
    {
      "heading": "",
      "text": "Quantum efficiency is a dimensionless measure of the number of specified output events produced per input photon. In a light detector, it commonly means the number of collected electrons per incident photon; in a light-emitting device, it commonly means the number of emitted photons per injected electron. Its precise meaning depends on the process being measured and on whether the denominator counts incident photons, absorbed photons, or injected charge carriers. Quantum efficiency is therefore a particle-counting measure, not directly a measure of energy conversion efficiency."
    },
    {
      "heading": "Definition and physical basis",
      "text": "For a photodetector, external quantum efficiency at wavelength \\(\\lambda\\) is commonly defined as\n\n\\[\\eta_{\\mathrm{ext}}(\\lambda) = \\frac{N_{\\mathrm{collected}}}{N_{\\mathrm{incident}}},\\]\n\nwhere \\(N_{\\mathrm{collected}}\\) is the number of elementary charge units collected as signal and \\(N_{\\mathrm{incident}}\\) is the number of photons arriving at the specified entrance surface during the same interval. The ratio may also be expressed using rates under steady illumination.\n\nA quantum efficiency of \\(0.80\\), or 80%, means that 100 incident photons produce, on average, 80 collected elementary charges when internal multiplication is absent. It does not mean that 80% of the incident optical energy becomes useful electrical energy.\n\nThe energy of an individual photon is\n\n\\[E_{\\gamma}=\\frac{hc}{\\lambda},\\]\n\nwhere \\(h\\) is the Planck constant and \\(c\\) is the speed of light in vacuum. Thus, a given optical power contains different numbers of photons at different wavelengths.\n\nThe concept is rooted in the quantum description of light and the photoelectric effect. Modern quantum-efficiency measurements extend this counting approach to semiconductor detectors, photovoltaic devices, light emitters, and photochemical processes."
    },
    {
      "heading": "External and internal quantum efficiency",
      "text": "External quantum efficiency\n\nExternal quantum efficiency, or EQE, includes losses between the defined external input and the useful output. For a detector, these may include:\n\n- Reflection at the entrance surface.\n- Absorption in windows, coatings, electrodes, or inactive material.\n- Transmission through the active region without absorption.\n- Recombination or trapping of generated charge carriers before collection.\n\nThe reference surface matters. An efficiency specified for a bare sensor need not equal the efficiency of a camera containing a lens, filters, and a protective window.\n\nInternal quantum efficiency\n\nInternal quantum efficiency, or IQE, normally refers to useful output per photon absorbed in the active material:\n\n\\[\\eta_{\\mathrm{int}}(\\lambda) = \\frac{N_{\\mathrm{collected}}}{N_{\\mathrm{absorbed,active}}}.\\]\n\nIf \\(A_{\\mathrm{active}}(\\lambda)\\) is the fraction of incident photons absorbed in that material, consistent definitions give\n\n\\[\\eta_{\\mathrm{ext}}(\\lambda) = A_{\\mathrm{active}}(\\lambda)\\eta_{\\mathrm{int}}(\\lambda).\\]\n\nThis distinction separates optical losses from losses occurring after absorption. High IQE can coexist with low EQE if much of the incident light never reaches or is not absorbed by the active region.\n\nFor a simple structure with negligible parasitic absorption, active absorptance may be approximated by \\(1-R-T\\), where \\(R\\) and \\(T\\) are reflectance and transmittance. That approximation is unsuitable when inactive layers absorb appreciable light: absorption somewhere in a device does not necessarily contribute to its useful signal."
    },
    {
      "heading": "Quantum efficiency in photodetectors",
      "text": "A photodetector converts optical radiation into an electrical signal. Its quantum efficiency generally depends on wavelength, operating conditions, and device structure.\n\nIn a semiconductor, photon absorption can create an electron–hole pair. In a photodiode, an internal or applied electric field helps separate the carriers and transport them to electrical contacts. One fully collected pair contributes one elementary charge to the external circuit, not two.\n\nThe band gap helps determine the long-wavelength absorption limit of many semiconductor detectors. Photons below the relevant absorption threshold are not efficiently absorbed through ordinary band-to-band transitions. At shorter wavelengths, carriers may be generated close to a surface, where recombination can reduce their collection probability.\n\nOther influences include:\n\n- Absorbing-layer thickness: thicker layers may increase absorption but can increase carrier transit distances.\n- Surface and interface quality: defects can trap carriers or promote recombination.\n- Bias voltage: this can change collection efficiency, depletion depth, and multiplication.\n- Temperature: this affects material properties, carrier transport, and unwanted background current.\n- Illumination geometry: angle of incidence, polarization, and position can affect optical coupling and collection.\n\nIn imaging sensors, backside illumination and suitable coatings can improve optical access to the active material. Pixel microlenses can redirect light away from inactive areas. Such improvements do not eliminate other limitations, including saturation, electrical noise, or nonuniform response.\n\nRelation to responsivity\n\nResponsivity expresses electrical output per incident optical power. For current responsivity,\n\n\\[\\mathcal{R}(\\lambda)=\\frac{I_{\\mathrm{ph}}}{P_{\\mathrm{opt}}},\\]\n\nwhere \\(I_{\\mathrm{ph}}\\) is the light-induced current and \\(P_{\\mathrm{opt}}\\) is incident optical power.\n\nFor monochromatic illumination and no internal gain,\n\n\\[\\mathcal{R}(\\lambda) = \\eta_{\\mathrm{ext}}(\\lambda)\\frac{q\\lambda}{hc},\\]\n\nwhere \\(q\\) is the elementary charge. Equivalently,\n\n\\[\\eta_{\\mathrm{ext}}(\\lambda) = \\mathcal{R}(\\lambda)\\frac{hc}{q\\lambda}.\\]\n\nAt \\(500\\ \\mathrm{nm}\\), a detector with 100% EQE has a responsivity of approximately \\(0.403\\ \\mathrm{A\\,W^{-1}}\\).\n\nResponsivity and quantum efficiency are therefore not interchangeable. Even at constant quantum efficiency, current responsivity increases with wavelength because a unit of optical power contains more photons at longer wavelengths."
    },
    {
      "heading": "Photovoltaic devices",
      "text": "In a solar cell, EQE is usually the number of electrons collected through the external circuit per incident photon at a specified wavelength and operating condition. Its spectrum shows which parts of the incident light contribute to current and helps distinguish optical losses from collection losses.\n\nIf \\(\\Phi_{\\mathrm{inc}}(\\lambda)\\) is incident photon flux per unit area per unit wavelength, the short-circuit current density can be estimated from\n\n\\[J_{\\mathrm{sc}} = q\\int \\eta_{\\mathrm{ext}}(\\lambda) \\Phi_{\\mathrm{inc}}(\\lambda)\\,d\\lambda,\\]\n\nprovided that the measured EQE applies under the illumination conditions of interest.\n\nQuantum efficiency is distinct from power-conversion efficiency. The latter depends on both current and voltage, as well as resistive and other losses. A cell can collect nearly one electron per absorbed photon yet lose substantial photon energy through carrier relaxation and deliver much less electrical energy than the incident optical energy.\n\nFor multijunction cells, spectral-response measurements require particular care because the response of one junction can be constrained by the others. Bias illumination and electrical bias may be used to isolate the response of a selected junction."
    },
    {
      "heading": "Light-emitting devices and luminescence",
      "text": "In a light-emitting diode, the direction of conversion is reversed. External quantum efficiency is commonly defined as\n\n\\[\\eta_{\\mathrm{ext,LED}} = \\frac{N_{\\mathrm{photons,escaped}}} {N_{\\mathrm{electrons,injected}}}.\\]\n\nInternal quantum efficiency describes photon generation inside the device per injected charge, with the exact convention depending on whether injection efficiency is included or treated separately.\n\nA useful factorization is\n\n\\[\\eta_{\\mathrm{ext,LED}} = \\eta_{\\mathrm{inj}} \\eta_{\\mathrm{rad}} \\eta_{\\mathrm{out}},\\]\n\nwhere \\(\\eta_{\\mathrm{inj}}\\) describes injection into the intended active region, \\(\\eta_{\\mathrm{rad}}\\) is the fraction of relevant recombination events that generate photons, and \\(\\eta_{\\mathrm{out}}\\) is the fraction of generated photons that escape. The factors must be defined consistently to avoid counting a loss twice.\n\nLight extraction can be limited by total internal reflection, absorption, and optical trapping. Consequently, efficient internal photon generation does not guarantee high external efficiency.\n\nFor optically excited emission, especially fluorescence, the closely related term photoluminescence quantum yield denotes emitted photons per absorbed excitation photon. More generally, quantum yield is widely used in photochemistry for specified molecular events per absorbed photon. These terms overlap conceptually with quantum efficiency but should not be substituted without stating the counted events.\n\nElectrical-to-optical power efficiency, often called wall-plug efficiency, is different again: it compares output optical power with input electrical power rather than counting photons and electrons."
    },
    {
      "heading": "Measurement and uncertainty",
      "text": "Quantum-efficiency measurements typically combine a known photon input with a measured output signal. For spectral measurements, a wavelength-selective source illuminates the device, and a reference detector with established calibration determines the incident optical power.\n\nImportant controls include:\n\n- Subtraction of dark signal and background light.\n- Adequate spectral purity and known wavelength bandwidth.\n- Measurement of the illuminated area and beam uniformity.\n- Operation within the device’s linear range.\n- Control of temperature, bias, and exposure time.\n- Correction for gain and readout conversion factors.\n\nFor IQE, determining absorption in the active region introduces additional uncertainty, especially in multilayer or strongly scattering structures. A value inferred from a simplified optical model is not equivalent to a direct measurement of active-layer absorption.\n\nReported results should specify whether they describe a component or an entire instrument, whether they include multiplication gain, and which operating conditions were used. Measurement uncertainty can arise from both optical-power calibration and electrical-signal measurement."
    },
    {
      "heading": "Limits and interpretation",
      "text": "Values above 100%\n\nQuantum efficiency is not universally restricted to 100%.\n\nIn an ordinary detector without multiplication, one absorbed photon generally produces at most one collected electron–hole pair. However, sufficiently energetic photons can generate multiple pairs. Photoconductive or avalanche gain can also make collected charge per incident photon exceed one elementary charge.\n\nSuch gain does not mean that the probability of detecting an individual photon exceeds one. A gain-inclusive charge yield, a primary conversion efficiency, and a photon-detection probability are different quantities. Multiplication may also introduce additional fluctuations.\n\nLikewise, some photon-conversion and photochemical processes can produce more than one specified output event per absorbed photon. Whether a value above unity is meaningful depends on the stated event count and the available energy.\n\nEfficiency is not complete detector performance\n\nQuantum efficiency alone does not determine sensitivity, image quality, or measurement precision. Detectors with similar QE can differ in background current, readout noise, response speed, saturation capacity, and stability.\n\nIn a simple photon-counting model with independent incident photons and fixed detection probability \\(\\eta\\), the number detected from exactly \\(N\\) incident photons has mean \\(N\\eta\\) and variance \\(N\\eta(1-\\eta)\\). Under Poisson-distributed illumination and ideal independent detection, the detected count is also Poisson-distributed, with mean \\(\\eta\\) times the incident mean. These statistical fluctuations remain even when instrumental noise is absent.\n\nA quantum-efficiency value is therefore most informative when accompanied by its wavelength, reference boundary, operating conditions, and definition of the useful output."
    }
  ],
  "links": [
    "photon",
    "electron",
    "energy",
    "planck-constant",
    "speed-of-light",
    "light",
    "photoelectric-effect",
    "reflection-optics",
    "semiconductor",
    "electron-hole",
    "photodiode",
    "electric-field",
    "band-gap",
    "temperature",
    "electric-current",
    "elementary-charge",
    "solar-cell",
    "light-emitting-diode",
    "total-internal-reflection",
    "fluorescence",
    "calibration",
    "measurement-uncertainty"
  ],
  "sources": [],
  "license": "https://creativecommons.org/licenses/by/4.0/"
}