Three materials, three different answers to the same problem: how to seal a sensor module behind a surface that must also be optically clear.
Short answer
A ceramic (zirconia) case back is scratch-resistant but opaque — it needs a separate window insert bonded into an opening, which adds a bond line, a leak path, and a second cosmetic interface. A sapphire case back can be monolithic: one part, optically clear in the sensor zone, opaque where printed masking hides the module. Glass sits between them on cost and behind both on durability. For a premium device with an optical sensor, the monolithic sapphire back is the cleanest architecture, and the difference shows up in leak rate and part count rather than in the datasheet.
The architectural problem
An optical heart-rate sensor needs a window directly over the LED and photodiode, positioned against skin. The case back must simultaneously:
- Transmit at 525 / 660 / 940 nm in the sensor zone
- Block stray light everywhere else
- Seal against sweat and water (typically 5–10 ATM on sports watches)
- Survive skin contact, sunscreen, and cleaning
- Look like a premium surface
The material choice determines how many parts you need to satisfy all five.
|
Architecture |
Parts |
Sealing interfaces |
Cosmetic risk |
|
Ceramic back + window insert |
2+ |
2 (insert bond + case gasket) |
Bond line visible, colour match |
|
Glass back, printed mask |
1 |
1 |
Scratches in service |
|
Sapphire back, printed mask |
1 |
1 |
Higher material cost |
Every additional interface is a leak path and a cosmetic defect opportunity. That is the strongest argument for a monolithic sapphire back, and it is rarely stated in material comparisons.
Material comparison for a case back
|
Property |
Sapphire |
Zirconia ceramic |
Chemically strengthened glass |
|
Optical transmission |
≥84.5% uncoated; ≥95% double-sided AR |
Opaque |
~91% uncoated |
|
Mohs hardness |
9 |
~8–8.5 |
~6.5–7 |
|
Fracture behaviour |
Chips at thin edges |
Very tough, high impact resistance |
Breaks into fragments; designed to craze |
|
Thermal conductivity |
~30 W/m·K |
~2–3 W/m·K |
~1 W/m·K |
|
Density |
~3.98 g/cm³ |
~6.0 g/cm³ |
~2.45 g/cm³ |
|
Colour options |
Clear, or masked by printing |
Through-body colour, many finishes |
Clear only |
|
Sensor integration |
Monolithic with printed mask |
Requires window insert |
Monolithic with printed mask |
|
Relative cost |
Highest |
High |
Lowest |
Two rows deserve attention.
Optical transmission. Ceramic is not a window material. Any ceramic back with an optical sensor is really a ceramic-plus-insert assembly, with all the sealing and cosmetic consequences that follow.
Density. Zirconia is roughly 50% denser than sapphire. On a 40 mm back this is a few grams — which matters in a category where the product page lists weight to the tenth of a gram.
Why sapphire's thermal conductivity cuts both ways
At ~30 W/m·K, sapphire conducts heat about thirty times better than glass. That has two consequences for a sensor module:
- Good for skin-temperature sensing. A sapphire back equilibrates quickly with skin, reducing the lag that makes wrist temperature readings slow to stabilise.
- Bad for thermal isolation of the optical module. The same conductivity pulls body heat into the module, which shifts LED output and photodiode dark current — both of which affect the PPG baseline.
Neither is a reason to avoid sapphire. Both are reasons to model the thermal path rather than assume the window is optically passive.
Coating strategy for a case back
A case back differs from a small sensor window in that it is large, visible, and touched constantly. That changes the coating plan:
|
Surface |
Coating |
Reason |
|
Outer (skin side) |
AR + AF |
Transmission into tissue; easy cleaning; less sebum scatter |
|
Inner (module side) |
AR |
Second pass transmission |
|
Masked zone |
Opaque print, both sides ideally |
Crosstalk and module concealment |
|
Edge |
None |
Bond or gasket interface |
On large parts, coating uniformity across the aperture becomes a real specification: a 40 mm back with 3% transmission variation across the sensor zone produces a spatially uneven illumination profile at the skin.
Ask for a transmission map, not a single point value.
Flatness is a sealing specification
For a case back, flatness is not primarily an optical parameter — it is a seal parameter. A back that is not flat will not compress a gasket evenly, and the leak appears at the high point.
|
Parameter |
Typical requirement |
Consequence if missed |
|
Flatness (sealing land) |
Tight, defined over the gasket path |
Leak at high point |
|
Flatness (optical zone) |
Looser |
Minor illumination non-uniformity |
|
Thickness tolerance |
Defined |
Sensor standoff changes, focus shift |
|
Edge profile |
Matched to case |
Gasket extrusion or gap |
Specifying a single flatness value over the whole part often over-constrains the optical zone and under-constrains the seal. Specify them separately.
Failure modes by material
|
Mode |
Sapphire |
Ceramic |
Glass |
|
Surface scratching |
Very rare |
Rare |
Common |
|
Edge chipping |
Possible at thin edges |
Rare |
Possible |
|
Catastrophic break |
Uncommon, needs impact |
Very rare |
Possible on impact |
|
Haze over life |
Very low |
N/A (opaque) |
Moderate |
|
Seal failure |
Adhesive/gasket driven |
Insert bond line driven |
Adhesive driven |
|
Cosmetic wear |
Low |
Low, but gloss changes with cleaning |
Moderate |
The ceramic failure mode worth noting is not the material — it is the insert. A bonded window in a ceramic back is a two-material joint under thermal cycling, and delamination there is a leak, not a cosmetic issue.
Cost framing
Piece price ranks sapphire highest, ceramic next, glass lowest. But the comparison is not like for like:
- A ceramic back requires an insert, a bonding process, a leak test on an extra interface, and colour matching between two materials.
- A glass back is cheap per part but has the highest field-wear rate, which in a premium product is a service and review cost.
- A sapphire back costs the most per part and removes an assembly step.
For a device priced in the premium tier, the deciding factor is usually part count and leak rate, not the delta on the back itself.
FAQ
Can a ceramic case back transmit the sensor light? No. Zirconia is opaque at PPG wavelengths. A ceramic back with an optical sensor is an assembly with a bonded window insert — plan for the bond line and the additional leak path.
Is sapphire too heavy for a case back? Sapphire is denser than glass but about a third lighter than zirconia for the same volume. On a typical 40 mm back the difference versus glass is a few grams.
Does the case back need AR coating on both sides? For sensing, yes — light passes through twice. Omitting the inner coating costs roughly 4% per pass, which is significant at 940 nm.
How is the opaque zone produced? By printing on the sapphire, usually on the inner face, with a mask that defines the clear aperture. Ink choice affects compliance and outgassing, so specify it with the same rigour as the coating.
What flatness should I specify? Define the sealing land and the optical zone separately. A single value across the whole part usually raises cost without improving the seal.
Talk to the sapphire engineers
Sun Yin Crystal Industry Company Limited has produced sapphire watch components since 1994 — 20 million pieces per year, in-house from crystal growth through CNC, polishing, and AR/AF coating. Case backs, crystals, and sensor windows are supplied RoHS / REACH / CP65 compliant under ISO 9001 and ISO 14001.
https://www.sunyinsapphire.com/
SUN YIN CRYSTAL INDUSTRY COMPANY LTD






