Paulo Lima | Oxford Instruments Plasma Technology: Why do coupon-based etch tests mislead when scaling to full wafers, and how can you avoid the surprise?
13:03 - 14:04
Other snippets from this talk
Summary of the clip:
Why do coupon-based etch tests mislead when scaling to full wafers, and how can you avoid the surprise?
The question describes calibrating with coupons in different areas, then expecting full-wafer behavior. In practice edge coupons can differ, and that difference can bite badly. A process optimized on coupon may be loaded onto a full wafer with high expectations, only to produce disappointing results. This sets up why coupon data cannot simply be projected across the wafer.
The response says this is very true and names the cause: the chemical load of the wafer. That load must be taken into account. Therefore, the recommendation is to process a full wafer, or at least a wafer made from the same material being etched. The reason given is that etching is a chemical reaction, so the chemical load matters.
Coupons are used to speed up development, because using multiple wafers instead would increase cost. The stated trade-off is cost. The practical takeaway from the exchange is that coupon results can miss the chemical load, so a full wafer or at least the same material is needed to account for it during process development.
In this short video, you can learn:
* Coupon etch tests can differ at the edge, so optimizing only on coupons may disappoint when a full wafer is processed.
* The chemical load of the wafer must be taken into account because etching is a chemical reaction.
* To speed development, coupons replace multiple wafers, but a full wafer or at least the same material should be processed.
π **Clip Abstract** This clip addresses why coupon-based etch calibration can fail to predict full-wafer results. The answer points to the wafer's chemical load, notes that etching is a chemical reaction, and says to process a full wafer or at least the same material, while coupons are used only to speed development and avoid multiple-wafer cost.
About the speaker:
* Speaker: Paulo Lima
* Company: Oxford Instruments Plasma Technology
* Event: Eindhoven 2026
* Location: High Tech Campus, Eindhoven
#ChemicalLoad, #CouponEtch, #FullWaferProcessing, #EtchCalibration, #SemiconductorManufacturing, #ProcessDevelopment
This is a highlight of the presentation:
Broad Ion Beam Processing for Scalable Manufacturing of Blazed Gratings for AR Waveguides
More Highlights from the same talk.
01:49 - 03:52
Why do AR waveguides need blazed gratings, and what makes scaling them across a 200 mm wafer so difficult?
Why do AR waveguides need blazed gratings, and what makes scaling them across a 200 mm wafer so difficult?
Blazed gratings matter because they couple light efficiently and directionally into waveguides. That directionality demands precision: the blaze geometry, the etch depth and the surface quality must all be controlled tightly. Without that control, coupling efficiency and directionality suffer, so the grating profile itself becomes a critical process output rather than a simple surface feature.
The scale-up challenge is not making the centre of the wafer meet those characteristics, but making them hold across the wafer in uniformity and repeatability. Broad ion beam is one option because it combines wafer size with the ability to micro polish after creating the geometry. The work here used silicon dioxide wafers.
The equipment processes up to 200-millimetre wafers and uses a 30-centimetre broad ion beam. Its directional flux creates the gratings with very precise control. The substrate can be tilted against the angle of the beam, and process parameters are controlled to set the blaze angle, the etch depth and the resulting profile. A chamber schematic shows the arrangement.
In this short video, you can learn:
* Blazed gratings must couple light efficiently and directionally into waveguides, which requires tight control of blaze geometry, etch depth and surface quality.
* The scale-up problem is wafer-level uniformity and repeatability, not just performance at the wafer centre.
* A 30-centimetre broad ion beam processes 200-millimetre wafers, with substrate tilt and process parameters setting blaze angle, etch depth and profile.
π **Clip Abstract** Broad ion beam processing is presented as a route to blazed gratings for AR waveguides, where light must couple efficiently and directionally. The talk covers the precision needed in blaze geometry, etch depth and surface quality, and the wafer-level uniformity and repeatability challenge, using 200-millimetre silicon dioxide wafers and a 30-centimetre broad ion beam with substrate tilt.
About the speaker:
* Speaker: Paulo Lima
* Company: Oxford Instruments Plasma Technology
* Event: Eindhoven 2026
* Location: High Tech Campus, Eindhoven
#BlazedGratings, #BroadIonBeam, #SiliconDioxideWafers, #BlazeAngleControl, #AugmentedReality, #Nanofabrication
05:43 - 07:51
What blaze angle and etch depth uniformity can ion beam etching achieve across a wafer?
What blaze angle and etch depth uniformity can ion beam etching achieve across a wafer?
Across the whole wafer, ion beam processing after micro polishing delivered a blaze angle variation of only 1.5 degrees. Etch depth uniformity varied by 4%. Because high volume manufacturing is the focus, a mini marathon of 50 runs was executed, and the wafer-to-wafer vertical depth came in at 4.5%. These are the headline process outcomes for manufacturable blazed gratings.
Metrology confirms the surface quality. A zoomed scanning electron microscope image sits beside an atomic force microscope scan of the blazed facet. Ion beam already produces a very smooth surface, with roughness around 2.7 nanometres, and a micro polishing step can reduce the roughness on the facet even further. Both characterisation techniques were shown side by side.
Using a model-based calculation, the team inferred 24 less stray light from the blazed grating, which is very important for waveguide quality. Alongside the slanted grating knowhow, these are the key process outcomes presented. The combination of tight blaze angle control, etch depth control and low facet roughness points to ion beam processing as a route to scalable manufacturing.
In this short video, you can learn:
* Ion beam etching after micro polishing held blaze angle variation to 1.5 degrees across the whole wafer.
* Etch depth uniformity varied by 4%, while wafer-to-wafer vertical depth over a 50-run mini marathon reached 4.5%.
* Facet roughness sits around 2.7 nanometres, and model-based calculation points to 24 less stray light for waveguide quality.
π **Clip Abstract** Broad ion beam processing for blazed gratings in AR waveguides achieved 1.5 degrees of blaze angle variation across the wafer, 4% etch depth uniformity and 4.5% wafer-to-wafer vertical depth over 50 runs. Atomic force microscope and SEM data show around 2.7 nanometres of facet roughness, reducible by micro polishing, with model-based calculation inferring 24 less stray light.
About the speaker:
* Speaker: Paulo Lima
* Company: Oxford Instruments Plasma Technology
* Event: Eindhoven 2026
* Location: High Tech Campus, Eindhoven
#BlazedGratings, #IonBeamEtching, #StrayLightReduction, #EtchUniformity, #AROptics, #Nanofabrication
03:48 - 05:33
How does a dual-source ion beam chamber control etch depth, profile geometry, and center-to-edge uniformity?
How does a dual-source ion beam chamber control etch depth, profile geometry, and center-to-edge uniformity?
The chamber schematic shows an ion source that can run not only argon but also other assisted gases. The substrate can tilt relative to the beam direction, and a shutter sets the beam dwell time. A second source on the left is used for deposition, with four targets enabling a kind of assisted deposition. This dual-source arrangement combines ion beam processing and deposition in one chamber.
To map center-to-edge variation, coupons are placed at the center and at different positions on the wafer. The variation of etch depth and profile geometry is then measured. The process first forms the facets, after which micro polishing is done at a very grazing angle, an incident angle to the wafer. The quality is assessed and optimized using a scanning electron microscope and an atom-force microscope. This was done on 200-millimeter wafers.
Control comes from several knobs: the tilt of the substrate relative to the beam, the shutter dwell time, the choice of assisted gas, and the deposition source with its four targets. Coupons at the center and at different wafer positions quantify etch-depth and profile-geometry variation. After facet formation, a very grazing-angle micro polish refines the surface. Scanning electron microscopy and atom-force microscopy then guide optimization. The output is blazed gratings patterned across 200-millimeter wafers.
In this short video, you can learn:
* The chamber uses an ion source with argon or other assisted gases, a tiltable substrate, and a shutter for beam dwell time.
* A second source on the left provides deposition from four targets, enabling assisted deposition in the same chamber.
* Center-to-edge variation is mapped with coupons, measuring etch depth and profile geometry, then optimized by SEM and atom-force microscopy after facet formation and grazing-angle micro polishing.
π **Clip Abstract** A dual-source ion beam chamber controls processing by tilting the substrate relative to the beam, setting dwell time with a shutter, and using an ion source that can run argon or other assisted gases. A second deposition source with four targets enables assisted deposition, while coupons map center-to-edge etch depth and profile geometry on 200-millimeter wafers; blazed gratings are formed by facet generation and very grazing-angle micro polishing and optimized with SEM and atom-force microscopy.
About the speaker:
* Speaker: Paulo Lima
* Company: Oxford Instruments Plasma Technology
* Event: Eindhoven 2026
* Location: High Tech Campus, Eindhoven
#BroadIonBeam, #BlazedGratings, #GrazingAngleEtch, #AssistedDeposition, #ARWaveguides, #DiffractiveOptics