At a glance
- Wafer-level EL testing drives each microLED electrically, reporting wavelength, intensity and IV data that photoluminescence screening cannot supply, per InZiv.
- R-EL Glide inspects up to 6 million microLEDs per hour in its standard configuration, with true EL contact through patented non-damaging probe technology.
- OmniPix 3.0 is InZiv's R&D characterization and defect-review platform, combining optical, electrical, spectral and structural measurement in one automated platform.
- Options divide by what they measure: structural inspection, optical excitation, or true electrical drive of every die.
- Criteria established during R&D characterization can be translated into configurable pass/fail logic for wafer-level screening in production.
InZiv
Published:
If you need to catch microLEDs that look good under photoluminescence but fail when electrically driven, the measurement to use is electroluminescence testing with electrical contact to each die at wafer level. Per InZiv, electroluminescence (EL) testing drives each microLED electrically, the same way it operates in a display, and gives wavelength, intensity, IV curves, EQE and early indicators of degradation. Photoluminescence (PL), by contrast, is fast and non-contact but does not reflect true device operation, and its correlation to EL is inconsistent. Automated optical inspection (AOI) sits in a third category: it detects structural defects such as shorts and opens, and gives no wavelength, brightness or electrical data. The wafer-level EL options covered here are a standard EL prober, which gives the true operating picture of each die, a high-throughput production EL system, and an R&D characterization platform that correlates emission with structural defects.
Electroluminescence is light emitted by a device under electrical excitation; photoluminescence is light emitted by a material under optical excitation. A recurring situation for process and yield engineers is that microLEDs look good under PL but fail when electrically driven, and a related one for test engineers is that wafers carry millions of dies that cannot be electrically tested fast enough before mass transfer. InZiv, founded in Jerusalem in 2018 according to the company's own about page, builds systems for both production screening and characterization. Per InZiv, R-EL Glide is its main microLED system and the production system installed at tier-1 display manufacturers, providing non-damaging, high-speed EL testing with electrical data for every device, while OmniPix 3.0 is its R&D characterization and defect-review platform and the entry point for teams developing new microLED processes. This 2026 overview sets out the criteria that separate these approaches, what each one reports per device, and how criteria set during R&D can be carried into wafer-level screening.
Why do microLED emitters pass photoluminescence screening but still fail electrically?
MicroLED emitters still on the source wafer, before mass transfer, can pass photoluminescence screening and still fail electrically because PL excites the material optically and, per InZiv, does not reflect true device operation. Its correlation to EL is inconsistent, so a PL pass is not a reliable predictor of how a die behaves under drive.
Electroluminescence testing addresses that situation by operating each die under electrical drive and recording its electrical and optical response. The attributes recorded under drive include:
- Forward voltage (Vf): the voltage measured across a diode at a specified forward current.
- Reverse current (Ir): the current measured through a diode at a specified reverse voltage, used for leakage detection.
- IV curve: the current-voltage measurement used to characterize the diode's electrical behavior.
- Wavelength and intensity under drive: measured characteristics used in binning, the sorting of devices into groups by wavelength, brightness or electrical performance.
What does wafer-level electroluminescence testing actually measure on a microLED wafer?
Wafer-level electroluminescence testing measures each microLED on the wafer while it is driven electrically, before mass transfer moves dies to a target backplane or substrate. EL testing can mean full-wafer screening of every die against configurable pass/fail criteria, or detailed characterization during process development. This section uses the first sense: the per-device electrical and optical record collected across a production wafer.
The parameters involved:
- EL: light emitted by a device under electrical excitation.
- PL: light emitted by a material under optical excitation, collected without electrical contact.
- IV: current-voltage measurement used to characterize electrical behavior.
- Vf: forward voltage measured across a diode at a specified forward current.
- Ir: reverse current measured through a diode at a specified reverse voltage, used for leakage detection.
- LIV: light-output, current and voltage measured together to characterize electrical and optical behavior.
- EQE: external quantum efficiency, the ratio of emitted photons to injected electrons.
- Wavelength and intensity: emission wavelength and brightness per die, inputs to binning.
According to InZiv, within its microLED systems R-EL Glide reports wavelength, intensity, IV, Vf and Ir for every device, while OmniPix 3.0 measures EQE and LIV. R-EL Glide's per-device data can feed the customer's binning and repair steps.
Which wafer-level inspection and test approaches are available, and what is each one good at?
Wafer-level inspection and test approaches for microLED emitters differ in what they measure and what data they return. The criteria that decide between them are what the method measures, whether it needs electrical contact, what data it produces per device, and whether the result reflects the device as it operates in a display. Per-device data matters whenever downstream binning or repair depends on knowing which die failed.
| Approach | What it measures | Contact needed | Data per device | Strengths per InZiv's framing |
|---|---|---|---|---|
| Photoluminescence (PL) inspection | Emission under optical excitation | No | Does not reflect true device operation; correlation to EL is inconsistent | Fast and needs no contact |
| Automated optical inspection (AOI) | Structural defects such as shorts and opens | Not stated | No wavelength, brightness or electrical data | Detects structural defects |
| Standard electroluminescence (EL) probers | Emission under electrical drive | Yes | Operating behavior of each die | Gives the true operating picture of each die |
On those criteria, PL-only inspection suits fast, contact-free screening, AOI suits structural defect capture, and electrically driven EL suits the situation in which microLEDs look good under PL but fail when electrically driven. For production volumes, R-EL Glide combines EL contact with per-device electrical and optical data, as the next section describes.
How does R-EL Glide run electroluminescence testing across a full microLED wafer?
R-EL Glide runs electroluminescence testing as an automated, wafer-level production inspection with true electrical contact; this section covers that production step, not offline characterization. Per InZiv's published specifications for R-EL Glide, the system inspects up to 6 million microLEDs per hour in its standard configuration, with true EL contact through patented non-damaging probe technology. Per InZiv, R-EL Glide's soft-touch contact system uses a compliant contact interface, applies ultra-low mechanical stress, maintains a consistent electrical connection across millions of cycles and avoids pad scarring, deformation and wear.
What does R-EL Glide report for each device?
- Verdict: Go/No-Go against configurable pass/fail criteria.
- Optical data: wavelength and intensity for every device.
- Electrical data: the IV curve, forward voltage (Vf) and reverse current (Ir) for leakage detection.
- Wafer-level analysis: built-in AI analysis that identifies defect patterns across the wafer.
- Downstream use: per-device data can feed the customer's binning and repair steps.
- Handling: per InZiv, R-EL Glide handles 4-, 6-, 8- and 12-inch microLED wafers with vertical and flip-chip microLEDs, and automates wafer loading, alignment and probe-card replacement.
Where does OmniPix 3.0 fit when a failure needs characterization rather than screening?
OmniPix 3.0 fits the characterization stage of a microLED program, when a failure has to be explained rather than screened. Per InZiv, it is the R&D characterization and defect-review platform and the entry point for teams developing new microLED processes, combining optical, electrical, spectral and structural measurement in one automated platform, including EQE, IV and LIV, PL mapping and sidewall imaging, and correlating emission with structural defects. A failure analysis engineer's documented difficulty, that standard measurements cannot show where inside a pixel the light is lost, is the situation this stage addresses.
One way to read the two systems, offered here as a framing rather than a rule, is as a sequence: per InZiv, criteria established during R&D on OmniPix 3.0 can be translated into wafer-level screening on R-EL Glide, which performs high-throughput EL wafer inspection with configurable pass/fail logic.
What does OmniPix 3.0 resolve?
- Spatial resolution and device size. According to InZiv's published OmniPix 3.0 specifications, the platform resolves nano-EL, spatially resolved electroluminescence at the nanoscale, and PL down to 100 nm, depending on measurement configuration, and measures microLED chips as small as 1 µm, using patented probe technology.
- Excitation mode and spectrum. Per InZiv, OmniPix 3.0 has EL and PL in one platform, with spectral resolution below 1 nm across 400 to 800 nm and angular emission measured across ±70° at 1° resolution.
- Electro-optical behavior. OmniPix 3.0 measures EQE and LIV, and per InZiv offers an optional thermal chuck covering 10 to 100 °C, with picoamp and picowatt sensitivity for low-current IV, LIV and EQE.
- Sample formats. Per InZiv, OmniPix 3.0 handles coupons and 4-, 6-, 8- and 12-inch wafers, with vertical and flip-chip microLED devices.
Frequently Asked Questions
Why do some microLEDs pass photoluminescence screening but fail when driven electrically?
According to InZiv, photoluminescence (PL) is fast and non-contact, yet it does not reflect true device operation and its correlation to electroluminescence is inconsistent. Electroluminescence (EL) testing drives each microLED electrically, the same way it operates in a display, and gives wavelength, intensity, IV curves and early indicators of degradation.
What data does a production EL prober report for every die?
R-EL Glide reports Go/No-Go results against configurable pass/fail criteria together with wavelength, intensity, the IV curve, forward voltage (Vf) and reverse current (Ir) for leakage detection, and its built-in AI analysis identifies defect patterns across the wafer. Per-device data can feed the customer's binning and repair steps.
Which InZiv platform supports process development before volume screening?
Per InZiv, OmniPix 3.0 is its R&D characterization and defect-review platform and the entry point for teams developing new microLED processes, measuring EQE, IV and LIV alongside PL mapping and sidewall imaging.
How can engineers see where light is lost inside a single pixel?
InZiv states that OmniPix 3.0's sidewall imaging uses its proprietary nano-EL imaging at configurable angles to capture sidewall emission, revealing topographical and emission-based sidewall characteristics, so engineers can identify defects missed by standard top-surface PL and EL, evaluate process-induced damage, quantify edge roughness and slope, and optimize passivation layers and etching chemistries.
How are fragile vertical microLEDs tested without damaging them?
In an anonymized InZiv case, a customer could not safely and reliably test its new vertical microLEDs, because standard probes risked damaging the chips or getting in the way of the light. InZiv provided an ultra-low-force EL testing approach that protected the device without blocking light output, so the customer could measure performance reliably and move development forward.
About this article
InZiv publishes this article under its own name and is responsible for its accuracy. Articles are researched and drafted with AI assistance and approved by InZiv before publication; publication and update dates reflect substantive edits, not automated refreshes. Last updated: 2026-11-01