top of page
Search

Inline 100% Surface Inspection of VCR-Sealing Surfaces for UHP Semiconductor appliances

Updated: Mar 20

In the era of “ultra-high purity” standards in the semiconductor industry, ensuring leak-tightness of VCR sealing interfaces is of critical importance.


Due to the increasing storage density required for AI applications and modern high-bandwidth memory technologies, ever deeper structures must be created on wafers.

To achieve this, major memory manufacturers are increasingly using more aggressive and corrosive process gases—and doing so more frequently.


Scratches on the sealing surface can serve as entry points for corrosion.


This can not only lead to equipment downtime (worst case), but even very minor damage can result in yield losses due to the resulting corrosion (particles in the process).


This explains why extremely high surface quality is required in the semiconductor industry (SEMI F19 UHP grade and, in some cases, even beyond).


Most VCR sealing surfaces, across different standards, are typically manufactured from

316L / 1.4404 stainless steel and are finished by precision turning and/or electropolishing, featuring toroidal or conical geometries.


Image. 1: VCR Components with toroidal Sealing surface, with electropolished head-area (Main Sealing Area)
Image. 1: VCR Components with toroidal Sealing surface, with electropolished head-area (Main Sealing Area)

A non-destructive, inline-capable inspection of the sealing surface quality is the solution to this problem. However, the boundary conditions for such inspection — such as:

  • highly reflective and glossy surfaces

  • complex geometries (conical-concave, conical-convex, toroidal)

  • very small components (<10 mm diameter)

  • surface defects <10 µm


—pose more than significant challenges for vision-based systems.

Several 3D imaging technologies currently available on the market cannot be used, such as laser triangulation or pattern projection, as they do not generate usable data on reflective surfaces.


Conventional machine vision techniques, such as dark-field imaging and polarization, are also not feasible due to the component geometry.


To overcome these limitations, we present the solution:


MacroTrax - Dome

Image 2: MacroTrax 535-0,8 with Sample-part at feasibility Study at Client
Image 2: MacroTrax 535-0,8 with Sample-part at feasibility Study at Client

The MacroTrax Dome utilizes a hybrid technology combining shape-from-shading and deflectometry to enable surface inspection across the entire relevant component geometry.


This allows for the inspection of surfaces with inclinations of up to 40° relative to the horizontal; under suitable conditions, steeper angles are also possible.


With a diameter of 150 mm and a height of 230 mm, and without the need for additional ambient light shielding, the system is compact and can be easily integrated into and retrofitted onto production and inspection systems.


Three different image sensors are available, currently offering two magnification levels (0.8× and 1.0×).



MacroTrax-535

MacroTrax-550

MacroTrax-2025

(in Development)

Imagesensor

2432x 2040 @ 3,45µm

2432 x 2040 @ 5µm

5104 x 4092 @ 2,5µm

Cycletime

<200ms

<200ms

<350ms

Variant:

0,8x

1,0x

0,8x

1,0x

0,8x

1,0x

Field of View (mm)

10,5x8,80

8,40x7,04

15,20x12,8

12,1x10,2

15,9x12,8

12,8x10,2

Depth of Field

(mm)

≈0,6

≈0,5

≈0,8

≈0,7

≈0,5

≈0,4

Resolution

( µm/px )

4,3125

3,45

6,25

5,0

3,125

2,5


With this turnkey solution, the inspection of complex components is no longer a challenge. The system generates at least three different images for evaluation, enabling independent inspection of different surface areas.


The three primary image types are:

  • coaxial

  • diffuse

  • hybrid deflectometry

Additional image types can also be generated if required.


Image 3: Raw Image taken with MacroTrax-535-0,8 with nearly invisible Scratch
Image 3: Raw Image taken with MacroTrax-535-0,8 with nearly invisible Scratch

The diffuse image, despite its high quality, is often insufficient for detecting scratches. To make these defects visible and analyzable, the hybrid deflectometry image is used.


Image 4: Hybrid-Deflectrometry Image from the same part with clear extraction of defects
Image 4: Hybrid-Deflectrometry Image from the same part with clear extraction of defects

Image 5: Zoom in Area of Scratch
Image 5: Zoom in Area of Scratch
Image 6: Extraction of Scratch
Image 6: Extraction of Scratch











The raw images of this sample part are available here for review and evaluation.


The system is already successfully in use with renowned german manufacturers of UHP-VCR components.

Defect evaluation and correct categorization are performed without the use of AI, although AI can be optionally employed if requested.



 
 
 

Comments


How can we help?

bottom of page