| 摘要: | 本論文的目標為改善以前亭儒學長在真空中使用鹼金屬棒(dispenser) 取數據時有50 kHz 的誤差[1]。最終目標是未來測量6D5/2 的超精細結構與絕對頻率。
在我們的系統中，有兩組雷射系統，分別稱為主雷射和僕雷射。主雷射的頻率固定於6S1/2 F = 3, 4 → 6D3/2 躍遷譜線上。僕雷射則透過偏頻鎖相技術控制雷射頻率，獲得真空中的Doppler-free 雙光子躍遷譜線。我們改良麒翔學長所設計的真空腔體[2]，在真空中使用ampoule 提供Cs 並測量6D3/2 的絕對頻率和超精細結構。
我們使用新的真空設計產生的粒子濃度比dispenser 通電產生的Cs 蒸氣[1]高，訊號強度接近一般Pyrex cell 可獲取的訊號強度，訊噪比較高。在Cs 與雷射交互作用的區域我們使用荷姆霍茲線圈（Helmholtz coil）來進行隔磁，避免Zeeman effect 造成的頻率偏移；透過AOM 進行功率穩定來測量AC Stark shift。
在測量collision shift 的過程中，我們發現粒子濃度會受到gate valve 處的抽氣率影響，影響譜線訊號強度和線寬。此外，我們意外發現中華電信提供的Rb clock 頻率準確度會隨時間變化導致絕對頻率量測非實際數值。Rb clock 的頻率準確度在6 週內會由1 ∗ 10−11 變為5 ∗ 10−11，對我們的系統造成約17 kHz 的偏移。
最後，我們在測量6D3/2 四個能階的頻率間距時發現數據與正恩學長量測的數據[3] 有明顯差異，因此量測6S1/2 F = 3 → 6D3/2 數據確認6S1/2 F = 3, 4 是否符合clock frequency，並重新計算超精細耦合常數的值，與過去的數值進行比較。;The objective of this paper is to improve upon the 50 kHz error encountered by senior Ting-Ju Chen when using dispenser to collect data in a vacuum[1]. The ultimate goal is to measure the hyperfine structure and absolute frequency of the 6D5/2 state in the future.
Our system consists of two laser systems, referred to as the master laser and the slave laser. The master laser frequency is locked to the 6S1/2 F = 3, 4 → 6D3/2 transition line. The slave laser frequency is controlled via an offset frequency lock technique to obtain the Doppler-free two photon transition line in a vacuum. We improved upon the vacuum chamber designed by senior Chi-Hsiang Chu, using an ampoule to provide Cs in a vacuum and measuring the absolute frequency and hyperfine structure of the 6D3/2 state[2].
Our new vacuum design generates a higher particle concentration than the Cs vapor produced by the dispenser[1]. The signal strength is comparable to that obtained with a standard Pyrex cell, with a higher signal to-noise ratio. In the region where Cs interacts with the laser, we use a Helmholtz coil to mitigate magnetic interference and avoid frequency shifts caused by the Zeeman effect. Power stabilization via an AOM is employed to measure the AC Stark shift.
During the measurement of the collision shift, we found that particle concentration is affected by the pumping rate at the gate valve, impacting the signal strength and linewidth of the spectral lines. Additionally, we discovered that the frequency accuracy of the Rb clock provided by Chunghwa Telecom varies over time, leading to inaccurate absolute frequency measurements. The frequency accuracy of the Rb clock can degrade from 1 ∗ 10−11 to 5 ∗ 10−11 over six weeks, causing a shift of approximately 17 kHz in our system.
Finally, when measuring the frequency intervals of the four energy levels of 6D3/2, we found a significant discrepancy compared to the data measured by Senior Jeng En[3]. Consequently, we measured the data for the transition 6S1/2 F = 3 → 6D3/2 to confirm whether 6S1/2 F = 3, 4 conform to clock frequency, recalculated the value of the hyperfine coupling constant, and compared it with past values. |
