In TGV glass via fabrication, batch-to-batch aperture drift and unstable dimension after laser-modified wet etching are common yield bottlenecks. The root cause of aperture fluctuation does not solely lie in laser drilling or etching alone. Instead, the process chain of laser modification parameters – etch rate – etch duration lacks standardized calibration, and variations in etching environment amplify dimensional deviation. This article sorts out four sensitive variables of etch rate and their control specifications, presents the joint laser-etch response surface calibration method, baseline calibration scheme for various glass substrates and a production coupon traceability system to stabilize aperture consistency.
The root cause of TGV aperture batch-to-batch drift resides neither in the laser process nor the etching process alone. It arises from the absence of standardized calibration within the chain of modification parameters – etch time – aperture, as well as insufficient stability control of the etching rate environment. The core target on the etching side is to stabilize the etch rate. Once four key variables including temperature, HF concentration, circulation flow and loading density are closed-loop controlled, the aperture growth curve becomes predictable.
The four variables are ranked by sensitivity to etch rate as follows:
Temperature (Highest sensitivity): The activation energy of fused silica / HF etching system is approximately 30~35 kJ/mol. A 10 °C temperature rise increases the etch rate by a factor of 1.7~2.1. Temperature exerts a stronger influence on etch rate than concentration. The equipment shall adopt PFA immersion heater with PID closed-loop control, dual-point temperature measurement inside the bath, ±0.5 °C control accuracy, and temperature records shall be kept per shift.
HF concentration: A ±1% absolute fluctuation of HF concentration corresponds to ±3%~5% variation of etch rate. Automatic chemical proportioning pumps are deployed together with daily titration verification, and compensated automatic replenishment maintains stable concentration.
Circulation flow: Insufficient flow deteriorates mass transfer inside vias and causes local rate non-uniformity. Variable-frequency magnetic drive pumps are recommended with bath turnover frequency of 2~4 cycles per hour, equipped with low-flow alarm for flow meters.
Loading density: Production panels and coupons share the same etchant in the same bath and consume HF together. An upper loading limit shall be defined based on the exposed glass etching area to avoid overloaded etchant and rate decline.
After the etching environment is stabilized, joint laser and etching calibration can be carried out. Keep the etching recipe fixed, set 5 levels of laser energy density × 5 levels of etch duration to form a 25-point aperture matrix. Fit the response surface among laser energy, etching margin, etch time and aperture, and generate two core process curves: aperture growth curve and selectivity curve (selectivity = etch rate of modified zone : etch rate of bulk glass, requirement ≥ 20:1). If the selectivity falls below this threshold, taper and lateral expansion become uncontrollable. Calibration specimens shall be retained as arbitration reference for future process disputes. In mass production, two-point re-verification is conducted on one lot weekly. Recalibration is mandatory once curve drift exceeds 5%.
The glass substrate itself changes the baseline of etch rate, which is an easily overlooked source of drift:
Fused silica: etch rate 1.5~2.5 μm/min with optimal sidewall cleanliness;
Borosilicate glass: etch rate 3~5 μm/min. While fast etching, dissolved boron and sodium tend to form a porous layer on sidewalls;
Alkali-free boroaluminosilicate glass (AF32, SGW3.0, mainstream substrate material for TGV): Al₂O₃ brings etching retardation effect, and residual passive film easily remains at via bottom. Enhanced vacuum degassing and endpoint management are required.
Before switching to a new glass grade or a new batch of the same grade, baseline rate calibration for the substrate must be performed. Prepare 6 coupons of identical material, measure initial thickness at five points. Under standard process temperature, apply three etching gradients of 10 min / 20 min / 30 min. Re-measure thickness after etching to fit the real rate curve of this batch, extrapolate the formal etching time and reserve 15%~20% etching margin.
A long-term coupon system must be implemented for mass production. Two coupons (same material, same thickness and identical pre-deposited mask) travel with each production lot. Thickness and aperture are inspected after etching, serving as release criteria for the lot and health indicator of the etchant bath. If the measured etch rate from coupons drops by more than 8% for two consecutive lots, bath replacement evaluation shall be triggered. All test data is logged into MES to form a lot-by-lot etch rate trend dashboard. Reliable traceability of aperture consistency is built upon cumulative coupon data from every batch.
Frequently Asked Questions
Q1: Laser energy is stable, why does aperture still drift batch by batch?
A1: Laser energy only determines the coverage and modification degree of the modified zone. The final aperture is jointly determined by modified zone selectivity + etch rate + etch duration. Even if laser parameters remain unchanged, aperture drift will occur if etching temperature, HF concentration or bath aging status changes and shifts the etch rate.
Q2: What is the significance of the selectivity ≥20:1 specification? Can we use 15:1?
A2: Selectivity reflects the etching difference between modified region and unmodified bulk glass. Higher selectivity confines etching within the laser-modified channel, with controllable lateral expansion and stable taper. Low selectivity means bulk glass is etched rapidly at the same time, making aperture extremely sensitive to etch time. Minor time deviation leads to large aperture shift and a narrow production process window.
Q3: A falling rate measured by coupons must be caused by low HF concentration?
A3: Not necessarily. Troubleshooting sequence for rate decline: bath temperature → insufficient circulation flow → impurity accumulation (enrichment of boron, aluminum, sodium ions inhibiting etching) → consumption of available HF concentration. Impurity buildup is a very common reason for rate decay for long-running baths with borosilicate and alkali-free boroaluminosilicate glass.
Q4: Is measuring etch rate sufficient for qualifying a new glass batch?
A4: No. Besides etch rate, sidewall morphology, via bottom residual film and taper variation shall be checked simultaneously. Minor composition fluctuation between batches of the same glass grade changes not only the etch rate, but also the formation behavior of retardation layers. A rate curve alone cannot predict residual film defects inside vias.
Q5: How to quantify the upper limit of loading density?
A5: Calculate based on total exposed etching glass area of all panels and coupons combined. Conduct limit verification according to bath volume and replenishment capacity. Confirm the safe loading range after verifying rate deviation across three consecutive batches under this loading condition. Counting substrate pieces alone is not recommended.
Q6: Is one-time response surface calibration valid permanently?
A6: No. The calibration is valid only when etchant system, mask process and glass substrate stay unchanged. Recalibration is required after bath renewal, mask revision, heater / circulation pump replacement or glass material switching. Weekly recheck captures slow long-term drift.
