Sean Chang | Texas Instruments: How does digital masking let inkjet handle 150, 200, 300 or even 450 mm wafers without new tooling?
11:49 - 13:46
Other snippets from this talk
Summary of the clip:
How does digital masking let inkjet handle 150, 200, 300 or even 450 mm wafers without new tooling?
The core advantage of an additive process is what the talk calls the digital mask: the pattern is stored as data, not fixed tooling. The clip illustrates this with simple images that, pieced together, form the shape of a whole wafer populated with all of its dies. Editing those patterns per die becomes a very simple task, with no mask set change.
Because the mask is digital, wafer diameter stops being a constraint on the patterning approach. The talk lists 150, 200 and 300 millimetre wafers and looks forward to 450 millimetre: additive doesn't really care about that. The technique can accommodate any sort of device and any size of wafer, which is framed as a lot of great potential going forward.
The commercial backdrop is wafer-diameter economics. Chipmakers continually push to larger wafers to get as much die as possible out of a given process, and doing so forces higher CapEx into even more expensive fabs. Other major chipmakers are described pouring in billions and billions of dollars to increase the scale and capacity their fabs can accommodate.
In this short video, you can learn:
* Digital masks replace fixed tooling, so individual die patterns can be edited by modifying simple images rather than re-tooling.
* The additive approach is wafer-size agnostic, cited for 150, 200 and 300 mm today and 450 mm going forward.
* Larger wafer diameters buy more die per process but demand higher CapEx commitments to ever more expensive fabs.
📋 **Clip Abstract** In this segment the case for additive processing rests on digital masks: patterns held as data and edited per die, illustrated as simple images that assemble into a full wafer map. Because nothing is tied to wafer diameter, the same approach is presented as workable from 150 up to 450 mm, against a backdrop of chipmakers spending billions to scale fabs for larger wafers.
About the speaker:
* Speaker: Sean Chang
* Company: Texas Instruments
* Event: Printed, Hybrid, Structural & 3D Electronics 2021
* Location: Online
Join TechBlick - the home of printed electronics, additive electronics, flexible hybrid electronics and 3D electronics
https://www.techblick.com/
#DigitalMasks, #AdditivePatterning, #DieLevelPatterning, #WaferDiameterScaling, #PrintedElectronics, #SemiconductorManufacturing
This is a highlight of the presentation:
Semiconductor and Printed Electronics Industries: Overview of emerging applications
More Highlights from the same talk.
02:40 - 04:56
Can inkjet print directly onto a fabbed silicon wafer and locally vary thickness without a photomask?
Can inkjet print directly onto a fabbed silicon wafer and locally vary thickness without a photomask?
Chip fabrication is a photolithography loop. Photoresist is coated onto the wafer, exposed through a photomask to define the pattern features, then developed out. With the pattern defined, the deposition method of choice follows and the resist is stripped away. That simplified sequence is repeated over and over, building up the material stacks needed to make a functional working chip.
Alternatively, you can just print the thing. A printer sits at the bottom left; at bottom right, prominently, is a wafer received out of front-end fabrication. The green material on it is patterned in a very particular way, and viewers able to resolve micron-scale differences will notice that each of the three sections on that wafer is actually different in thickness. The alternative to the lithography loop is print.
The stated advantage is flexibility: additive manufacturing, and namely inkjet, offers a lot of flexibility that you don't really see in a lot of the semiconductor manufacturing methods used today. Inkjet is presented as the primary example of printed electronics, and the printed wafer, patterned with three sections of differing thickness, is what that flexibility looks like before the talk delves further.
In this short video, you can learn:
* Front-end chip fabrication runs resist coat, photomask exposure, develop, deposition and resist strip, repeating that loop until the material stack makes a working chip.
* A wafer received straight out of front-end fabrication can carry an inkjet-printed green material whose three sections differ in thickness at micron scale.
* Additive manufacturing, specifically inkjet, is positioned as offering flexibility that is not really seen in many semiconductor manufacturing methods used today.
📋 **Clip Abstract** The clip contrasts the photolithography loop, resist coat through photomask exposure, develop, deposit and strip, with simply printing the material onto a wafer straight out of front-end fabrication. A printed green layer, patterned and varying in thickness across three sections at micron scale, illustrates the flexibility claimed for inkjet additive manufacturing.
About the speaker:
* Speaker: Sean Chang
* Company: Texas Instruments
* Event: Printed, Hybrid, Structural & 3D Electronics 2021
* Location: Online
Join TechBlick - the home of printed electronics, additive electronics, flexible hybrid electronics and 3D electronics
https://www.techblick.com/
#InkJetOnSilicon, #PhotomaskLithography, #ThicknessVariationAcrossWafer, #FrontEndWaferPatterning, #PrintedElectronics, #SemiconductorManufacturing
17:58 - 19:42
What would change if one additive line could make multiple chip configurations instead of many fabs?
What would change if one additive line could make multiple chip configurations instead of many fabs?
The clip frames current chipmaking as hitting a cap: improving existing applications and tools is increasingly painful, and tooling and fabs keep getting more expensive. Additive manufacturing is positioned as the route to redefine chipmaking, with a single manufacturing line able to make different product configurations. That would cut prototyping and time-to-market, while adding flexibility and reducing associated materials and cost.
An inflection point is described for materials and tools: they are receiving much more attention and improving rapidly, and the outlook is bullish. Yet additive chipmaking still depends on an ecosystem that must improve. Readily available materials, such as inkjet materials covering a wide range of properties, are needed, alongside available design rules and demonstrated reliability and cost.
The clip stresses urgency: these ecosystem requirements are things the industry cannot wait for. If readily available materials, design rules, reliability and cost are not in place, manufacturers will keep going with an easily manufacturable solution. That sets the bar for additive approaches: they must compete on manufacturability, not only on the flexibility of a single line making multiple configurations.
In this short video, you can learn:
* Current chipmaking is described as hitting a tooling and cost cap, with fabs and tools becoming more expensive.
* Additive chipmaking could use one manufacturing line for different product configurations, reducing prototyping time, time to market and materials cost.
* The ecosystem needs readily available inkjet materials with broad properties, design rules, reliability and cost, or industry will stay with easily manufacturable solutions.
📋 **Clip Abstract** The talk describes additive semiconductor manufacturing as a way past rising tooling and fab costs, using a single line for multiple chip configurations to cut prototyping time and time to market. It says materials and tools are at an inflection point, but additive adoption still needs readily available inkjet materials, design rules, reliability and cost, or manufacturers will keep choosing easily manufacturable solutions.
About the speaker:
* Speaker: Sean Chang
* Company: Texas Instruments
* Event: Printed, Hybrid, Structural & 3D Electronics 2021
* Location: Online
Join TechBlick - the home of printed electronics, additive electronics, flexible hybrid electronics and 3D electronics
https://www.techblick.com/
#AdditiveChipmaking, #SingleLineManufacturing, #InkjetMaterials, #DesignRules, #PrintedElectronics, #SemiconductorManufacturing
06:02 - 08:02
Why do fabs treat new inkjet materials as contamination risks when they already run tens of billions of chips?
Why do fabs treat new inkjet materials as contamination risks when they already run tens of billions of chips?
The immediate risk is contamination of fab processes, which can trigger an excursion event: chip yields suddenly plummet and production engineers must diagnose the cause. Existing fab materials, processes and tools are well understood because chips have been made similarly for decades, with iterative improvements rather than major change. That stability makes any new process introduction sensitive to compatibility and manufacturability at scale.
Volume matters: tens of billions of chips are made each year, supported by tools and processes designed for extremely high yield. A new material therefore needs compatibility, manufacturability at scale, and bulk material properties plus subsequent performance comparable to currently qualified production materials. Without that equivalence, the fab cannot treat the material as just another drop-in input; yield and process stability are the constraints.
Inkjet materials are already at a disadvantage. Their nature generally requires lower surface tension and viscosity, and that severely limits the materials you can use. The challenge is formulation compatibility: those fluid properties must still satisfy fab compatibility, scale and performance comparisons against currently qualified materials. That is why introducing a new inkjet process raises contamination and manufacturability concerns alongside the material-property limits.
In this short video, you can learn:
* A contamination-driven excursion can make chip yields plummet and force production engineers to diagnose the cause.
* New fab materials must be compatible, manufacturable at scale, and comparable in bulk properties and performance to qualified production materials.
* Inkjet materials generally need lower surface tension and viscosity, severely limiting usable materials and raising compatibility challenges.
📋 **Clip Abstract** Introducing new materials into a semiconductor fab is constrained by contamination risk, excursion events and the need for high yield across tens of billions of chips. Inkjet materials face an additional disadvantage because lower surface tension and viscosity severely limit usable formulations, while compatibility, scale and performance must match qualified production materials.
About the speaker:
* Speaker: Sean Chang
* Company: Texas Instruments
* Event: Printed, Hybrid, Structural & 3D Electronics 2021
* Location: Online
Join TechBlick - the home of printed electronics, additive electronics, flexible hybrid electronics and 3D electronics
https://www.techblick.com/
#ContaminationExcursion, #InkjetMaterialViscosity, #FabYieldPlummet, #BulkMaterialProperties, #PrintedElectronics, #SemiconductorManufacturing
About this talk
→ Semiconductor and Printed Electronics Industries: Overview of emerging applications · Sean Chang · Printed, Hybrid, Structural, & 3D Electronics 2021


