Physical Address
304 North Cardinal St.
Dorchester Center, MA 02124
💡 IEC 61066: Thermoluminescence Dosimetry from Solid-State Physics to Personal and Environmental Monitoring Engineering
Among the tens of millions of personal dosimeters worn by occupationally exposed workers worldwide each year, one technology holds a commanding share: the thermoluminescence dosimeter (TLD). Whether pinned to the chest of a nuclear power plant operator, an interventional radiologist, an industrial radiographer, or a uranium miner, that unassuming white card or cylinder almost invariably contains a lithium fluoride crystal doped with magnesium and titanium — LiF:Mg,Ti, the classic TLD-100 — as its radiation-sensing core. The operating principle is elegantly physical: ionising radiation promotes electrons into metastable “trap” energy levels within the crystal lattice; upon controlled heating, the trapped electrons are released and recombine with luminescence centres (holes), emitting visible light whose total yield is proportional to the deposited radiation energy — that is, the dose. The international standard that governs type testing, performance classification, and calibration traceability for this entire system is IEC 61066:2006, “Thermoluminescence dosimetry systems for personal and environmental monitoring.” This article, grounded in the standard’s second edition (which replaced the 1991 first edition), systematically unpacks the phosphor physics, reader design, calibration metrology, environmental resilience, and the recurring operational errors that plague external dosimetry services.
0.01 mSv ~ 10 Sv
Dose Range (Routine & Accident)
12 keV ~ 7 MeV
Photon Energy Range Hp(10) / H*(10)
LiF:Mg,Ti
Most Widely Deployed TLD Phosphor
Hp(10) / Hp(0.07) / H*(10)
Three Operational Quantities
⚛️ 1. Thermoluminescence Physics: Traps, Glow Curves, and the Quantification of Dose
1.1 How Ionising Radiation Leaves a Record in a LiF Crystal
The ability of thermoluminescence to serve as a radiation archive is rooted in crystal defect engineering. Pure LiF is a wide-bandgap insulator (band gap ~14 eV) that possesses no intrinsic charge-storage capability. But when ~100~200 ppm of Mg2+ and a small amount of Ti3+ are introduced during melt growth, the Mg2+ ions substitute for Li+ lattice sites, creating sites of excess positive charge that stabilise neighbouring Li+ vacancies to form Mg2+-vacancy dipoles. These dipoles exist as non-luminescent recombination centres before irradiation. When photon or beta radiation interacts with the crystal:
- Electron-hole pair production: The incident radiation excites valence-band electrons across the band gap into the conduction band, simultaneously leaving holes in the valence band. Each MeV of absorbed energy generates approximately 3.5 x 104 electron-hole pairs in LiF — this is the upper physical limit of dose response.
- Carrier trapping: Free electrons migrate through the conduction band and are captured at Mg-related defect complexes (principally trimer Mg-Li-Mg clusters, the so-called “traps”), forming trapped-electron centres. Concurrently, holes are captured at Ti3+-associated defects, forming trapped-hole centres. The two trapping processes together store the radiation energy in metastable form within the lattice.
- Trap depth governs storage stability: Different Mg defect clusters create traps of varying depths. The classic glow curve of TLD-100 reveals at least 10 glow peaks (trap levels), of which the most critical are: Peak 5 (the main dosimetric peak at ~210 deg.C, depth ~1.38 eV), Peak 4 (~180 deg.C), Peak 3 (~160 deg.C), and Peak 2 (~120 deg.C, depth ~0.95 eV). Deeper traps hold electrons more securely: at room temperature, the probability of thermal detrapping from Peak 5 is vanishingly small (half-life ~80 years), whereas Peak 2 depopulates within hours. This is why Peak 5 is the dosimetric workhorse and Peak 2 is a nuisance to be removed.
1.2 The Glow Curve: A TLD System’s Fingerprint
Place the irradiated TLD chip on a metallic heating planchet, ramp the temperature at a controlled rate (e.g., 10 deg.C/s) from room temperature to approximately 300~400 deg.C, and simultaneously record the emitted light intensity as a function of temperature (or time) using a photomultiplier tube (PMT). The resulting trace is the glow curve. For LiF:Mg,Ti, it presents a sequence of partially overlapping peaks whose peak-maximum temperatures (Tmax) are determined by the trap depth, in accordance with first-order Randall-Wilkins kinetics:
I(T) = n0 s exp(-E/kT) exp[ -(s/β) ∫ exp(-E/kT’) dT’ ]
where n0 is the initial trapped-electron population (proportional to dose), E is trap depth (eV), s is the frequency factor (s-1), T is absolute temperature, β is the heating rate, and k is Boltzmann’s constant. The area under each peak (or peak height, for well-separated peaks) quantifies the dose component stored in that trap. In practice, dose evaluation uses the combined area of Peaks 4+5, since Peak 4 (~180 deg.C) and Peak 5 (~210 deg.C) overlap substantially and cannot be fully deconvolved in routine production reads.
💡 Engineering Insight — Preheating to Eliminate Low-Temperature Peak Noise
Peaks 2 and 3 fade severely within tens of hours post-irradiation. Different dosimeter batches spend varying durations in postal transit, storage, and wear cycles, meaning the contribution of these low-temperature peaks is inherently unstable. Standard operating protocol requires a 100 deg.C / 10-minute pre-read anneal, or a low-temperature preheat phase built into the readout ramp (typically 135~160 deg.C), which empties Peaks 2 and 3 without any effect on Peaks 4 and 5. The precision of this operation directly controls measurement reliability in the low-dose region (<0.1 mSv). Many automated readers brand this feature as a “preheat plateau” and it is a standard prerequisite in the readout algorithm.
Peaks 2 and 3 fade severely within tens of hours post-irradiation. Different dosimeter batches spend varying durations in postal transit, storage, and wear cycles, meaning the contribution of these low-temperature peaks is inherently unstable. Standard operating protocol requires a 100 deg.C / 10-minute pre-read anneal, or a low-temperature preheat phase built into the readout ramp (typically 135~160 deg.C), which empties Peaks 2 and 3 without any effect on Peaks 4 and 5. The precision of this operation directly controls measurement reliability in the low-dose region (<0.1 mSv). Many automated readers brand this feature as a “preheat plateau” and it is a standard prerequisite in the readout algorithm.
1.3 From Light to Dose — The Reader’s Core Optical Chain
A TLD reader is, at its essence, a high-gain, low-noise photon-counting or photocurrent-integrating optical system. IEC 61066 imposes a set of stability requirements on the reader independent of the dosimeter. A typical reader’s optical chain is composed of the following critical modules in series:
- Heating system: Usually a metallic planchet (platinum or nichrome, resistively heated) or a hot-nitrogen gas stream. Heating rate must be tightly regulated (±2%), because glow-peak temperature shifts with heating rate, and rate fluctuation translates directly into dose-readout fluctuation. A rate of 10 deg.C/s is the de facto global standard across most dosimetry laboratories: faster rates compress and overlap peaks; slower rates extend measurement time and worsen signal-to-noise ratio.
- Light collection optics: A mirror assembly directs the combined light output — the thermoluminescence signal from the phosphor plus thermal blackbody radiation from the hot planchet — onto the PMT photocathode. To prevent the incandescent blackbody radiation (which becomes intense above ~300 deg.C) from overwhelming the TL signal, an infrared-absorbing optical filter must be inserted in the light path (typically Schott KG-1, KG-3, or Corning 7-59 blue-green glass). The transmission window of the filter stack must be matched to the main emission band of LiF:Mg,Ti (~410 nm, blue region).
- Photomultiplier tube (PMT): Converts single photons into nanosecond current pulses, which are then shaped, discriminated, and counted to yield total photon counts (photon-counting mode), or integrated as DC current via a transimpedance amplifier (current-integration mode). Photon counting offers superior signal-to-noise at microsievert-level doses. PMT gain exhibits a temperature coefficient of approximately -0.3%/deg.C, which is why IEC 61066 Clause 11.6 explicitly requires reader stability testing and Clause 11.7 requires ambient-temperature testing on the reader itself.
- Reference light source: Most readers incorporate an internal stability reference — a 14C-doped scintillator or an LED reference source — that fires before or after each measurement to track and correct for PMT gain drift. This is the minimum baseline configuration for any production TLD reader.
✅ Best Practice — Institutionalise an In-Reader Fixed-Interval Quality-Control Protocol
IEC 61066 Clause 11.6 requires reader drift not to exceed ±5% over a 24-hour period. In daily operation, insert a set of reference dosimeters (calibration detectors irradiated to a known dose) and a set of background dosimeters after every 50 routine dosimeters. If the mean reference reading drifts outside the control limits of a Shewhart chart, stop the batch immediately and trace the root cause to PMT high-voltage drift or filter contamination. This is the concrete translation of a “standard requirement” into a daily laboratory SOP.
IEC 61066 Clause 11.6 requires reader drift not to exceed ±5% over a 24-hour period. In daily operation, insert a set of reference dosimeters (calibration detectors irradiated to a known dose) and a set of background dosimeters after every 50 routine dosimeters. If the mean reference reading drifts outside the control limits of a Shewhart chart, stop the batch immediately and trace the root cause to PMT high-voltage drift or filter contamination. This is the concrete translation of a “standard requirement” into a daily laboratory SOP.
📏 2. IEC 61066 Performance Requirements and Calibration Traceability
2.1 The Three Operational Quantities and Their Technical Distinctions
One of the defining revisions in the 2006 second edition of IEC 61066 was the full adoption of operational quantities as defined in ICRU Report 51. The standard covers three dose monitoring quantities whose physical meanings and application contexts are fundamentally different:
| Quantity | Definition | Reference Depth | Typical Application | Photon Energy (IEC 61066) | Beta Energy |
|---|---|---|---|---|---|
| Hp(10) | Personal Dose Equivalent — Deep | 10 mm below body surface | Whole-body effective dose assessment (deep organs) | 12 keV ~ 7 MeV | N/A |
| Hp(0.07) | Personal Dose Equivalent — Shallow | 0.07 mm below body surface | Skin dose assessment (extremities, superficial) | 8 keV ~ 250 keV | 0.07 MeV ~ 1.2 MeV |
| H*(10) | Ambient Dose Equivalent | 10 mm in ICRU sphere | Area/environmental radiation monitoring (not worn on body) | 12 keV ~ 7 MeV | N/A |
⚠️ Critical Distinction — Hp(10) Is Never Equal to H*(10)
This is the single most common conceptual error in radiation protection training. Hp(10) is defined on a human body proxy (the ICRU slab phantom for the trunk, the rod phantom for fingers) — the body itself backscatters and attenuates the radiation field, altering the actual fluence received by the detector. H*(10) is a pure environmental quantity defined inside the ICRU sphere phantom — used for area monitoring, never attached to a person. The same TLD detector calibrated on a slab phantom versus on a sphere phantom may differ in response by 5~15% for the same radiation field (depending on energy and angle). IEC 61066 explicitly requires the manufacturer to declare which operational quantity the system is designed to measure, and the corresponding calibration must use the appropriate phantom.
This is the single most common conceptual error in radiation protection training. Hp(10) is defined on a human body proxy (the ICRU slab phantom for the trunk, the rod phantom for fingers) — the body itself backscatters and attenuates the radiation field, altering the actual fluence received by the detector. H*(10) is a pure environmental quantity defined inside the ICRU sphere phantom — used for area monitoring, never attached to a person. The same TLD detector calibrated on a slab phantom versus on a sphere phantom may differ in response by 5~15% for the same radiation field (depending on energy and angle). IEC 61066 explicitly requires the manufacturer to declare which operational quantity the system is designed to measure, and the corresponding calibration must use the appropriate phantom.
2.2 IEC 61066 Performance Requirements Quick-Reference Table
The following table summarises the core performance limits for Hp(10) dosimeters as extracted from IEC 61066 Table 3, covering coefficient of variation, nonlinearity, energy and angle response, and environmental durability:
| Performance Item | Test Conditions | IEC 61066 Limit | Engineering Interpretation |
|---|---|---|---|
| Coefficient of Variation (CV) | 10 repeated irradiations of same batch at reference conditions | ≤ 7.5% (Hp(10) above H0) | Standard deviation / mean. Reflects combined detector-reader system precision. If CV exceeds limit, first check heating-planchet temperature uniformity, then investigate phosphor doping batch consistency. |
| Non-Linear Response | From 0.1 mSv to 1 Sv span; deviation from reference-dose response | 0.91 ≤ R ≤ 1.11 (factor 1.11; i.e., within ±9% above H0; ±30% below H0) | In the high-dose region of 1~10 Sv, LiF:Mg,Ti exhibits supralinearity — response can exceed the low-dose response by 20~30%, which must be corrected algorithmically. |
| Radiation Energy and Angle of Incidence | 80 keV ~ 1.25 MeV (minimum required range), incidence 0deg;~60deg; | Response ratio relative to reference between 0.71~1.67 (factor 1.5) | For low-energy photons (~30~80 keV X-rays), photoelectric cross-section ∝ Z4 causes LiF (Zeff≈8.2) to over-respond relative to tissue; packaging compensation is required (see Section 2.3). |
| Fading | Post-irradiation storage at 30 deg.C/65%RH for 30 days | Response change ≤ ±20% | This is one of the most stringent daily operational constraints on a dosimetry service (discussed in Section 3.2). |
| Light Exposure | Post-irradiation exposure to 1000 lx visible light for 24 h | Response change ≤ ±10% | TLD packaging must strike a delicate balance between optical transparency and protection against light-induced fading. |
| Drop | Free fall from 1.5 m onto hardwood floor | Response change ≤ ±10% | Dosimeters are worn, removed, and pocketed daily — routine mechanical shock is a normal operating condition. |
| Reader Stability | Continuous operation under fixed conditions for 24 h | Reading drift ≤ ±5% | Governed jointly by PMT dark-current drift, HV module temperature drift, and planchet temperature regulation. |
2.3 The Culprit and the Cure: Low-Energy Photon Energy Response and Encapsulation Compensation
One of the core reasons LiF is widely adopted is that its effective atomic number (Zeff ≈ 8.2) is very close to soft tissue (Zeff ≈ 7.4), a dramatic improvement over earlier phosphors such as CaSO4:Dy (Zeff ≈ 15.3) and CaF2:Mn (Zeff ≈ 16.3), whose over-response to low-energy photons could reach 10~15x. Nevertheless, even with LiF’s Zeff only modestly above that of tissue, in the 30~80 keV low-energy X-ray band the photoelectric cross-section’s Z3~4 dependence yields an over-response of +30~40%. This means that in nuclear medicine (99mTc, 140 keV), interventional radiology (pulsed fluoroscopy at 70~100 kVp), and similar settings, uncorrected TLD readings will be biased high.
There are three principal engineering countermeasures: (1) Multi-element detector with filter-window algorithm: The dosimeter holder incorporates metal filters of different thicknesses (e.g., Cu 0.5 mm + plastic window). The ratio of readings under different filters allows estimation of the incident radiation’s effective energy, which in turn permits an energy-dependent correction to the main detector reading. This is the standard design of the modern four-element TLD card (exemplified by the Harshaw 8800 series). (2) Dual-phosphor approach: Encapsulate low-Z and high-Z phosphors together and exploit their differing energy responses to back-calculate a spectral correction. (3) Use Li2B4O7:Mn or LiF:Mg,Cu,P: These phosphors have Zeff values even closer to tissue (~7.4 and ~7.5 respectively), yielding a flatter energy response, though with lower sensitivity or tighter annealing windows.
💡 Engineering Decision Guide — When to Abandon Simple Calibration and Deploy Spectral Algorithms
If the monitored population’s photon energy range is confined to 80 keV ~ 1.25 MeV (the typical scenario for routine nuclear power plant operations), a single-chip LiF:Mg,Ti dosimeter with a traceable 137Cs calibration factor from PTB or NIST is statistically sufficient to meet IEC 61066’s energy-response tolerance window of ±50%. But if the population includes interventional radiologists (X-ray spectra concentrated in the 30~100 keV range), nuclear medicine staff (multi-nuclide mixed fields from 99mTc at 140 keV to 131I at 364 keV), or security-screening equipment maintenance personnel, then a multi-element dosimeter with at least three filter windows and a matched energy-discrimination algorithm is mandatory — otherwise the systematic bias at low energies will be unacceptable.
If the monitored population’s photon energy range is confined to 80 keV ~ 1.25 MeV (the typical scenario for routine nuclear power plant operations), a single-chip LiF:Mg,Ti dosimeter with a traceable 137Cs calibration factor from PTB or NIST is statistically sufficient to meet IEC 61066’s energy-response tolerance window of ±50%. But if the population includes interventional radiologists (X-ray spectra concentrated in the 30~100 keV range), nuclear medicine staff (multi-nuclide mixed fields from 99mTc at 140 keV to 131I at 364 keV), or security-screening equipment maintenance personnel, then a multi-element dosimeter with at least three filter windows and a matched energy-discrimination algorithm is mandatory — otherwise the systematic bias at low energies will be unacceptable.
🔧 3. Engineering Practice and Pitfalls in Personal and Environmental Monitoring
3.1 Annealing: The Invisible Quality Gate in Dosimeter Zeroing
After irradiation and readout, every TLD chip must be annealed to completely empty all residual trapped electrons and restore the Mg defect dipole distribution to its pre-irradiation state before the chip can be re-deployed. The annealing process directly governs the residual signal (zero-dose reading) and sensitivity of the next wearing period. The standard annealing protocol for LiF:Mg,Ti (TLD-100) is:
- Standard anneal: 400 deg.C for 1 hour (high-temperature phase), followed by 100 deg.C for 2 hours (low-temperature phase). The high-temperature phase disperses Mg dipole clusters and resets them to a standard distribution; the low-temperature phase re-aggregates Mg into the dipoles/trimers that stabilise the Peak 5 trap distribution. Skipping the low-temperature anneal can reduce Peak 5 sensitivity by 20~30%.
- Reader anneal (post-read only): Some fully automated TLD systems incorporate an in-reader post-heat at 240 deg.C for 10 seconds at the end of each measurement cycle, intended to zero the chip directly and eliminate the need for an external annealing oven. The residual signal from this method is typically 5~10 muSv equivalent dose — acceptable for personal monitoring (where the wearing-period dose vastly exceeds this value), but unacceptable for environmental monitoring.
⚠️ Annealing Oven Drift — The Most Common Systematic Error in TLD Laboratories
Annealing oven temperature miscalibration is the most insidious source of systematic bias. If the actual oven temperature is 20 deg.C lower than the displayed setpoint (a common magnitude of thermocouple drift), a nominal 400 deg.C anneal is in reality only 380 deg.C, leading to incomplete Mg dipole disaggregation and elevated residual signal. In the 0.1 mSv low-dose region this residual signal can account for 20~50% of the total reading and be misinterpreted as abnormal occupational exposure. Worse still: nobody notices — because all control dosimeters processed in the same period go through the same biased oven and the calibration factor itself may already embed the error. The solution: perform a monthly temperature-distribution survey of the annealing oven using an independently calibrated furnace profiling system (with multiple Type-K thermocouple probes), and verify residual signal levels quarterly using unirradiated dosimeters.
Annealing oven temperature miscalibration is the most insidious source of systematic bias. If the actual oven temperature is 20 deg.C lower than the displayed setpoint (a common magnitude of thermocouple drift), a nominal 400 deg.C anneal is in reality only 380 deg.C, leading to incomplete Mg dipole disaggregation and elevated residual signal. In the 0.1 mSv low-dose region this residual signal can account for 20~50% of the total reading and be misinterpreted as abnormal occupational exposure. Worse still: nobody notices — because all control dosimeters processed in the same period go through the same biased oven and the calibration factor itself may already embed the error. The solution: perform a monthly temperature-distribution survey of the annealing oven using an independently calibrated furnace profiling system (with multiple Type-K thermocouple probes), and verify residual signal levels quarterly using unirradiated dosimeters.
3.2 Fading: The Uncertainty Time Bomb of Thermal Signal Loss
Fading refers to the thermally stimulated spontaneous escape of trapped charge carriers after irradiation, causing the measurable TL signal to decrease with time. IEC 61066 Clause 11.4 requires that after 30 days of storage at 30 deg.C/65%RH, the residual dose reading must remain within ±20% of the irradiated reference value. In operational reality, fading is governed by a complex intersection of factors:
- Trap depth: Peak 5 (~1.38 eV) has a room-temperature half-life of approximately 80 years — negligible fading. Peak 4 has a half-life of roughly 1 year, losing about 8% over a 30-day wearing period. Peak 3 has a half-life of only a few days and virtually vanishes within a month. This is why the pre-read preheat that eliminates Peaks 2 and 3 is decisive for fading control.
- Ambient temperature: The Arrhenius relationship means that every 10 deg.C rise in ambient temperature approximately doubles the fading rate. Dosimeters stored in an un-air-conditioned vehicle in a hot summer climate (cabin temperature can reach 50~60 deg.C) may experience 30-day fading of 30~40%, far exceeding the IEC 61066 limit of ±20%. The “non-wear period” of a dosimeter’s monitoring cycle — warehousing, transit, queuing in the laboratory awaiting processing — often affects data quality more than the wearing period itself.
- Phosphor composition differences: LiF:Mg,Cu,P (TLD-100H / GR-200A) is 25~30 times more sensitive than LiF:Mg,Ti, but its room-temperature fading rate is also 2~3 times higher — because the trap distribution corresponding to its main dosimetric peak (~210 deg.C) differs subtly from LiF:Mg,Ti. Choosing high sensitivity means accepting higher fading uncertainty; an engineering trade-off must be made between monitoring period length and fading correction effort.
✅ Operational Countermeasure — Control Dosimeters Are the Non-Negotiable Baseline for Fading Correction
In every batch of dosimeters issued to users, retain at least 5~10 identical-batch control dosimeters. Irradiate a subset of them to known doses on the issue date, store them alongside the batch through the same temperature history, and read them simultaneously when the batch is returned. The decay in the control readings is the true fading coefficient for that batch — no theoretical model is needed. IEC 61066 similarly requires the use of extra dosimeters for natural-background correction, following the same principle: physical controls lock down uncertainty.
In every batch of dosimeters issued to users, retain at least 5~10 identical-batch control dosimeters. Irradiate a subset of them to known doses on the issue date, store them alongside the batch through the same temperature history, and read them simultaneously when the batch is returned. The decay in the control readings is the true fading coefficient for that batch — no theoretical model is needed. IEC 61066 similarly requires the use of extra dosimeters for natural-background correction, following the same principle: physical controls lock down uncertainty.
3.3 Environmental Monitoring: Engineering Transitions from Personal to Area Dosimetry
When a TLD transitions from “personal dosimeter” to “area/environmental dosimeter” (measuring H*(10)), several engineering assumptions must be re-examined. Environmental TLDs are typically deployed outdoors for 3 months to 1 year, over a temperature range far wider than indoor personal monitoring periods. Key differences include:
- Natural background subtraction: The H*(10) measured by an environmental TLD includes contributions from cosmic rays and terrestrial natural radioactivity (40K, U/Th series). The global average natural background dose rate is approximately 2.4 mSv/a (0.27 muSv/h), but this varies by more than a factor of 20 across different regions in China (1~20 mSv/a). Environmental TLD data that lacks natural-background subtraction has zero cross-regional comparability.
- Waterproofing and sealing: IEC 61066 Clause 11.5 requires dosimeter sealing against moisture. In outdoor deployment, rainwater infiltration can cause phosphor hydrolysis (LiF is slightly water-soluble) or filter fogging, directly degrading optical transmission efficiency. The standard calls for ultrasonic-sealed pouches or O-ring-sealed aluminium capsules.
- Self-irradiation: Intrinsic radioactive impurities in the dosimeter itself (e.g., 40K from glass encapsulation materials, trace 87Rb or Th/U impurities) produce a cumulative background signal during long-term (more than 6 months) deployments. The self-dose rate of high-quality TLD chips themselves is typically below 1 muGy/a and is negligible, but the self-irradiation contribution of encapsulation materials must be verified.
❓ Frequently Asked Questions
- Q1: In operational service, how should I choose between LiF:Mg,Ti (TLD-100) and LiF:Mg,Cu,P (TLD-100H)?
- A: The choice involves a three-way sensitivity-versus-fading-versus-annealing-window trade-off. TLD-100H (GR-200A) offers 25~30 times the sensitivity of TLD-100, making it especially suitable for low-dose environmental monitoring (e.g., monthly accumulated dose at a nuclear plant perimeter below 0.05 mSv), where it can suppress statistical fluctuations of very low signals into an acceptable range. The costs are: 2~3x higher fading rate (monitoring period should not exceed 3 months), an extremely narrow annealing window (a 240 deg.C anneal shifted to 260 deg.C causes permanent large sensitivity loss, whereas TLD-100’s process window is much wider), and a lower maximum operating temperature (sensitivity degrades irreversibly above ~240 deg.C). TLD-100 remains the mainstream choice for personal monitoring, because its process maturity and wide annealing window are better suited to high-volume automated handling (hundreds to thousands per month). If both Hp(0.07) (beta/low-energy photons) and Hp(10) must be measured, TLD-100 can be thinned to 5~10 mg/cm2 ultra-thin chips, whereas TLD-100H thin chips are more brittle and fracture-prone.
- Q2: What is supralinearity and how is it corrected in the high-dose region?
- A: In LiF:Mg,Ti, above approximately 1~10 Gy, the TL yield per unit dose is no longer constant but increases with dose. This phenomenon is termed supralinearity. The physical mechanism: at high doses, irradiation produces an extremely high density of electron-hole pairs and traps become progressively filled; the competitive capture kinetics during carrier recombination become nonlinear. The result is that at 10 Gy, the light output per gray may be 1.2~1.4 times that at 0.1 Gy. IEC 61066 allows a response deviation not exceeding ±9% (factor 1.11) above H0 (the recording level). Correction is accomplished by establishing multi-point calibration curves with piecewise-linear or quadratic fitting, or by applying a vendor-supplied empirical supralinearity correction function (typically expressed as a supralinearity index curve). For TLD measurements in the accident-dose range (more than 1 Gy), the unreliability of supralinearity correction is the largest single contributor to dose uncertainty.
- Q3: How does TLD compare with OSL (Optically Stimulated Luminescence) dosimetry?
- A: Both are passive integrating dosimeters based on the “radiation-fills-traps, readout-stimulates-luminescence” principle. OSL (typified by Al2O3:C, the Landauer Luxel+ InLight system) uses laser or LED optical stimulation instead of heating to liberate trapped carriers. Its core advantages are: (1) re-readability — only approximately 0.05% of trap charge is consumed per read, making OSL nearly non-destructive, which is critical for legal archiving and dispute re-evaluation (TLD heating completely erases the signal after readout, precluding verification); (2) readout speed (<1 second vs. 10~20 seconds per chip for TLD); (3) Al2O3:C sensitivity is extremely high (~40~60x that of LiF:Mg,Ti), offering overwhelming statistical advantage at muSv-level environmental monitoring. TLD’s relative strengths are: (1) LiF’s tissue-equivalence is superior to Al2O3 (Zeff ~11.3 vs. 8.2); (2) cost — per-chip TLD cost is roughly 1/3 to 1/5 that of OSL elements; (3) in the very high-dose region (more than 10 Gy), TLD supralinearity behaviour has been extensively characterised, while OSL data in this region remain sparse. The global trend is that OSL is steadily displacing TLD for personal monitoring in developed-country markets, but TLD remains dominant in bulk environmental monitoring and in developing countries.
- Q4: How can one judge whether a TLD dosimetry service provider is technically competent?
- A: Evaluate at least five dimensions: (1) Does the laboratory hold ISO/IEC 17025 accreditation and has it participated in international dosimetry intercomparisons (e.g., IAEA/WHO postal dose audit, EURADOS intercomparisons) and achieved satisfactory results in all categories over the last 3 rounds? (2) Verify the laboratory’s control-dosimeter protocol — does it deploy at least 5 control dosimeters per batch for fading and background correction? If the answer is evasive, the practice is likely absent. (3) Request recent annealing-oven calibration and verification records — the temperature uniformity across the effective volume of the annealing oven should be within ±2 deg.C, supported by reports from the last 3 months. (4) Ask about dosimeter configuration strategy for clients with differing energy spectra — if the same single-chip dosimeter and single calibration factor are applied to all clients (nuclear power plant, hospital, industrial radiography), the energy-compensation strategy is likely crude. (5) Does the dose report include an uncertainty estimate for each monitored individual? Personal dose reports that omit uncertainty do not, strictly speaking, meet the requirements of ICRP Publication 75, nor do they align with IEC 61066’s conceptual framework (which frames a measurement result as a value plus an associated uncertainty, not a single point value). A technically confident dosimetry service will always provide an uncertainty estimate.
