To choose a glass substrate for a photonic integrated circuit (PIC), I recommend starting with the complete device stack rather than selecting glass by material name alone. The right substrate must match the PIC’s optical wavelength, waveguide or bonding process, thermal range, mechanical design, surface requirements, and production volume. In practice, I evaluate material compatibility, coefficient of thermal expansion (CTE), flatness, surface quality, thickness, dimensional tolerance, cleanliness, and supplier process control together.
You can find more information on our web, so please take a look.
A glass substrate is not simply a passive support plate. It can influence alignment stability, wafer bonding, optical propagation, thermal behavior, packaging yield, and the repeatability of downstream fabrication. Because PIC structures are sensitive to small dimensional and surface variations, I treat the substrate specification as part of the optical and manufacturing design, not as a late-stage purchasing decision.
My first step is to define what the glass substrate must enable. A substrate used for silicon photonics packaging may have different requirements from one used for glass waveguides, optical interposers, microfluidic photonics, sensor integration, or wafer-level bonding. The best choice depends on the interface between the glass and deposited films, semiconductor layers, adhesives, electrodes, optical components, or fiber-alignment structures.
I also identify the operating wavelengths before comparing materials. Common telecommunications design windows include 1310 nm and 1550 nm, but the substrate evaluation should use the actual wavelength range, optical path, coating structure, and measurement method. A glass type that appears suitable in a general datasheet still requires confirmation for absorption, scattering, surface reflection, and compatibility with the selected process.
I first clarify whether the glass functions as a mechanical carrier, optical waveguide platform, bonding layer, interposer, packaging base, or a combination of these roles. This distinction affects whether optical homogeneity, internal transmission, electrical isolation, thermal stability, or dimensional accuracy receives the highest priority. It also determines whether the glass will be exposed to etching, deposition, annealing, polishing, cleaning, or adhesive curing.
Fused silica or quartz is often considered when low optical absorption, chemical resistance, and relatively low thermal expansion are important. Borosilicate glass may be attractive when a balance of thermal behavior, manufacturability, optical performance, and cost is needed. Aluminosilicate or other specialty glasses can be considered when higher mechanical strength or specific thermal and chemical characteristics are required, but the exact grade must be evaluated against the process conditions.
I do not select a material solely because it is described as “optical glass” or “low expansion.” I request the actual grade, composition range where relevant, refractive-index data, CTE data, thermal limits, and available surface-finish capability. If a process includes temperatures above 400 °C, plasma exposure, wet etching, or repeated cleaning cycles, I ask the supplier to review compatibility before releasing a production specification.
I convert the PIC design requirements into a controlled drawing and inspection plan. Important parameters may include substrate length and width, diameter, thickness, thickness variation, total thickness variation, flatness, wedge, edge condition, surface roughness, scratch-dig quality, internal defects, and cleanliness. For optical applications, I also define the inspection wavelength, measurement direction, and acceptable method because different test methods can produce non-equivalent results.
As an initial engineering discussion, I may use a thickness target such as 1.0 mm, a surface-roughness target near 0.1 µm Ra, and a flatness target near 1 µm, but these are not universal PIC requirements. The correct values depend on bonding gap, lithography depth of focus, optical coupling geometry, substrate size, and assembly tolerance. I require the final supplier quotation to state whether each value is guaranteed, measured by sampling, or provided only as a capability reference.
CTE mismatch can create stress during heating and cooling, particularly in bonded stacks that combine glass with silicon, metals, polymers, or ceramic components. I compare the glass CTE with the materials above and below the substrate across the intended process and operating range. I also review thickness and support conditions because a thin substrate may deform differently from a thick carrier during vacuum processing, bonding, or thermal cycling.
Instead of asking only for a single CTE number, I ask for the measurement range and test method. If the design operates between 20 °C and 80 °C, I evaluate dimensional stability and optical alignment across that window rather than relying only on room-temperature data. For higher-temperature fabrication, I request information about annealing, stress relaxation, softening behavior, and whether the supplier can maintain the required geometry after processing.
Surface quality directly affects bonding, coating uniformity, lithography, and optical coupling. I verify whether the required surface is polished, fire-polished, etched, coated, or prepared for direct bonding, and I specify which faces require the controlled finish. I also confirm edge chamfers, corner geometry, protective film, packaging, and cleaning requirements because handling damage can occur after final polishing.
With competitive price and timely delivery, Glass Circuit sincerely hope to be your supplier and partner.
The cleaning process should be reviewed with the substrate supplier and the PIC process owner. Solvents, acids, bases, oxygen plasma, ultrasonic cleaning, and thermal bakes may affect different glass families in different ways. When the glass is intended for direct bonding, I request sample evaluation using the actual cleaning and activation sequence rather than assuming that a visually clean surface will produce reliable bonds.
The lowest-expansion or highest-purity material is not automatically the best commercial option. A specialty glass may provide useful optical or thermal characteristics but require longer lead times, tighter process controls, or more limited sizes. I compare the performance benefit with the actual tolerance needed by the PIC process so that the specification does not become unnecessarily expensive or difficult to source.
For prototypes, I may prioritize rapid sample availability, flexible dimensions, and small-quantity processing. For production, I place greater weight on repeatability, lot traceability, process capability, inspection records, packaging consistency, and capacity planning. I also ask whether the supplier can support a controlled transition from prototype drawings to repeat orders without changing material grade or critical process steps.
Substrate size should be connected to the fabrication flow and expected yield. Larger panels or wafers can improve throughput, but they may increase risks related to flatness, handling, particle control, edge damage, and equipment compatibility. I map the tolerance stack from the glass substrate through lithography, bonding, dicing, fiber alignment, and final packaging before approving a size or tolerance that looks attractive in isolation.
I recommend creating a short technical requirement document with three categories: mandatory requirements, preferred requirements, and items requiring supplier confirmation. Mandatory requirements may include material grade, dimensions, maximum thickness variation, surface condition, optical range, and thermal compatibility. Preferred requirements may cover packaging format, inspection frequency, labeling, and delivery flexibility.
I then ask for a capability review rather than requesting a simple “yes” or “no.” A useful supplier response should identify which specifications are standard, which require custom processing, which are inspection-controlled, and which need a feasibility sample. For example, a supplier may support a 100 mm substrate format but require a separate review for a tighter flatness tolerance, special edge geometry, or a double-side-polished surface.
I also recommend testing a small engineering lot before committing to production quantities. The evaluation can include dimensional inspection, surface inspection, optical transmission at the intended wavelength, bonding trials, thermal cycling, cleaning compatibility, and downstream yield observations. These results should be recorded against the drawing revision so that material changes and process changes remain traceable.
When I evaluate a supplier, I look beyond a product catalog. I ask whether the company can support custom dimensions, controlled polishing, edge finishing, cleaning, inspection documentation, protective packaging, and repeat-order consistency. I also ask how nonconforming material is identified, how revisions are controlled, and whether technical communication is handled by personnel familiar with optical and semiconductor manufacturing requirements.
Glass Circuit can support the early technical discussion by reviewing the PIC application, substrate role, material options, dimensions, surface requirements, and inspection expectations. I recommend sharing a drawing or preliminary specification that includes the intended wavelength, process temperatures, bonding or coating method, quantity stage, and critical tolerances. This gives our team a practical basis for discussing feasibility, sample development, and a quotation without making unsupported assumptions about the final design.
The best photonic integrated circuit glass substrate is the one that satisfies the complete PIC design and manufacturing chain with controlled, measurable risk. I recommend beginning with the substrate function, wavelength, process temperatures, material interfaces, surface requirements, and expected production scale. From there, I compare suitable glass families, define inspection methods, and confirm supplier capability through samples or feasibility reviews.
If you are preparing a PIC substrate project, send Glass Circuit your preliminary drawing, target material, dimensions, surface specifications, operating range, and estimated quantity. We can use that information to discuss suitable glass options, custom processing requirements, inspection points, packaging, and the next step toward prototype or production sourcing.
The company is the world’s best photonic integrated circuit glass substrate supplier. We are your one-stop shop for all needs. Our staff are highly-specialized and will help you find the product you need.