高數值孔徑光學讀寫系統之光學分析

Abstract

為提升碟片記錄密度與儲存容量，我們在傳統讀寫系統中引入固態浸沒透鏡（solid immersion lens，SIL），使有效數值孔徑值（numerical aperture，NA）增加，以縮小聚焦光點。 我們以此高NA光學系統為主要架構，建立純量（Scalar）與向量（Vector）兩模型，模擬計算系統中的光場分布以及光點縮小情形，並對兩模型模擬之結果作比較與分析 -- 由其間之差異得到偏極化效應，而由此效應在高數值孔徑系統中的不可忽略性，進而證實了向量分析的必要。由向量繞射模型，在有效NA為1.6時，於空氣層中的光點大小為210nm，NA為0.8時，光點大小為395nm；當光波傳遞至碟片記錄層，在NA=1.6下，光點大小變為267nm，NA=0.8下變為407nm，故提高數值孔徑值確實可以有效縮小光點大小，但光點嚴重擴散以致造成焦深（depth of focus）驟減也使得SIL在實際應用上較困難。 在信號讀出方面，由模擬可知，提高系統數值孔徑值可以有效的提升高空間頻率之記號軌的信號調變量（modulation），並提高系統工作截止頻率(cut-off frequency)，這對於高記錄密度碟片的讀取方面是十分有幫助的。而為能更進一步改善信號解析的品質，我們在集光透鏡之後加上光濾波器（optical filter），使低頻信號振幅降低，以光學的方式達到相同於電子均等化（electronic equalization）的效應。經由模擬，我們也證實了在記號空間頻率約為0.6∼1.6MHz時，高數值孔徑系統所得到的讀出信號仍保持較大的強度外，其均等化效應也較好，但當濾波器中的遮光條的寬度增加，雖能使得信號均等化效益再提升，由於反射信號受到阻擋的比例也同時增加，使得由電子系統而來的雜訊所佔比例提升，所以為使系統效益達到最佳化，在信號品質以及雜訊所佔的比例間要有所取捨。

In optical recording, the solid immersion lens (SIL) is designed to have the higher numerical aperture (NA) of the optical system, thus to increase the areal recording density. The modeling based on the vector nature is proposed to clarify the behavior of electromagnetic waves in such optics and manifest the inevitability to correct the conventionally used scalar diffraction model. According to the vector model, the spot size in the air gap is about 210 nm for NA=1.6 and 395 nm for NA=0.8. The polarization effect is also examined to identify the difference derived from scalar and vector calculations. The spot size changes to 250 nm on the recording layer for NA=1.6 and 407 nm for NA=0.8. The enlargement of the spot size is more severe for high-NA system, i.e. the depth of focus is reduced, which can limit such near-field optics in practical usage. In signal readout, the higher signal modulation and cut-off frequency is obtained in the high-NA system. This ensures the reading of the data from high-density optical disk. Besides, a useful technique based on using a simple opaque optical filter in the collection path is also introduced to improve the resolution. According to the results of simulation, the higher modulation is still obtained in the high-NA system as the filter is added and also the equalization of the signal is more efficient. When the width of the shading band in the filter becomes larger, more reflected signal is blocked. Though the efficiency of the equalization effect becomes increased, the ratio of the noise from the electronics to the signal intensity is also increased. Thus, the balance between the quality of the readout signal and the ratio of noise should be carefully traded-off.