采用石英增强光声光谱检测系统，并引入空芯光子晶体光纤作为气体参考气室，实现对痕量氨气的高灵敏度检测。参考气室采用长5 m 的空芯光子晶体光纤，两端熔接单模光纤，内部填充标准氨气。通过分析空芯光子晶体光纤的模态干涉，获得低干涉噪声的透射谱。气体填充过程中，控制填充压强与时间，提高谱线分辨率，完成分布反馈式(DFB)激光器波长的精确锁定，提高检测精度。测量参考气体腔内氨气吸收谱线线宽，并与高分辨率光谱谱线(HITRAN)数据库数据对比验证实验结果。采用光声光谱检测系统，优化调制参数，获得氨气噪声等效浓度(即指体积分数)为6.74×10-6(3σ)。

A system based on quartz enhanced photo- acoustic spectroscopy technique is investigated for trace ammonia gas detection. And hollow- core photonic band- gap fiber is introduced as reference gas cavity for high accuracy and sensitivity. The reference gas cavity is consisted with a 5 m long hollow- core photonic bandgap fiber, which filles with ammonia gas and splices with single-mode optical fibers at both ends. Modes interference is analyzed to obtain transmission spectrum with low interference noise. For improving detection accuracy, gas filling time and pressure are controlled in filling procedure. Ammonia absorption line width in ref? erence gas cavity of hollow-core photonic bandgap fiber is measured, and compared with the high-resolution transmission (HITRAN) database data. By this method, wavelength of distributed feed back (DFB) laser is locked accurately. Quartz enhanced photo- acoustic spectroscopy system is used for ammonia detection with optimal modulation parameters, which yields that a noise equivalent concentration (namely volume fraction) of ammonia is 6.74×10-6 (3 s).