We've been evaporating 250nm thick layers of SiO2 (for antireflection coatings) using an ebeam evaporator on GaN and on sapphire. The samples appear fine when we unload them from the chamber but when we start unmounting the GaN sample, cracking of the oxide occurs. The sapphire sample remains fine. Any ideas? We tried a plasma ash before hand on both samples but still saw the cracking and also made sure the samples were in good contact (for thermal purposes) with the metal mount. Thanks --John, Boston University Photonics