On-chip short-wave infrared multispectral detector



Figure 1 (a) Schematic diagram of the on-chip integrated FP microcavities array SWIR multispectral detector; (b) SWIR multispectral detector chip; (c) 16-channel response spectrum of the SWIR multispectral detector; (d) The gas detection response spectral channel of the SWIR multispectral detector chip.

Infrared spectrum detection is to detect and analyze the light wavelength information of matter through detectors, like Sun Wukong's piercing eyes, which can see the essence through appearance and accurately identify the material information that cannot be judged by humans eyes. The short-wave infrared (SWIR) band is rich in material spectral information, and has the ability to penetrate clouds and fog. It has a wide range of applications in food, agriculture, remote sensing imaging and other fields.

Conventional detectors do not have the capability of spectral identification and need to use gratings, interferometers or other spectroscopic devices to detect the spectral information of targets. For traditional multispectral detection systems, their detectors and spectroscopic devices are separated, requiring mechanical assembly and scanning, resulting in large volumes and complex systems, which are not suitable for portable and on-site detection scenarios.

In recent years, photonic crystals, metamaterials and quantum dots, which have received extensive attention and research, can reduce the spectral detection system to chip level, which has good application prospects. In the SWIR band, the photonic crystals and metamaterials need a high-precise fabrication process with high cost because of their sub-wavelength structures. The quantum dot filters array is restricted by its low repeatable synthetic preparation process and insufficient long-term stability. In contrast, pixel-level spectroscopic arrays based on Fabry-Perot (FP) microcavity can be fabricated by UV lithography require easier fabrication process. Its transmission spectrum is easy to be precisely tuned by cavity thickness, and its structure is stable and easy to be integrated with detectors.

Professor Wang Shaowei's team from Shanghai Institute of Technical Physics, Chinese Academy of Sciences, has reported an on-chip SWIR multispectral detector based on monolithically integrated Fabry–Perot microcavities array. By monolithically integrating the pixel-level FP microcavity structures on InGaAs focal plane array (FPA), a multispectral detection chip is realized in SWIR band. This research has been published in Chinese Optics Letters, Vol. 20, Issue 6, 2022 (Zhiyi Xuan, et al. On-chip short-wave infrared multispectral detector based on integrated Fabry–Perot microcavities array) and selected as the Cover paper.

The researchers monolithically integrated pixel-level FP microcavity arrays with different transmission spectra onto an InGaAs FPA by a combinatorial etching process, each FP microcavity structure corresponds to a multispectral detection channel, to achieve chip-level SWIR Multispectral detection.

In this work, a pixel-level FP microcavities array with 16 different spectral channels is monolithically integrated on a 64×64 pixel InGaAs FPA through 4 times combinatorial etching processes. For the first time, the multi-spectral detection chip in the 1400 ~ 1700 nm SWIR band has been demonstrated, which makes the chip itself have the ability of spectral identification, and can identify the SWIR spectral information of 16 different wavelengths, realizes the integration of spectroscopy and detection, and greatly simplifies the structure of SWIR multispectral detection system, minimized its size.

As shown in Fig. 1(a), when the incident light is irradiated on the chip, it first resonates in the FP microcavity and performs wavelength selection, and the light of the transmitted wavelength is received by the detector. By adjusting the thickness of the FP microcavity cavity layer at different positions, 16-channel on-chip multispectral detection is realized. Figure 1(b) shows the SWIR multispectral detector chip, which is only 2 mm2 in size, the same as the original InGaAs FPA. Figure 1(c) shows the response spectrum of the chip's 16 channels. The number of spectral channels of the multi-spectral detection chip can also be greatly increased according to needs. Only 7 combined etchings can achieve a wider range of spectral detection with 128 channels (the number of spectral channels increases exponentially by 2N). The multi-spectral detection chip turns the original detector that can only respond to light intensity information into a spectral detection chip, so that it has the ability to identify material components. Figure 1(d) shows the gas detection channel of the chip, which can detect CO2, C2H2, H2S, CH4 gases at the same time.

The bandwidth and center wavelength of the multi-channel response spectrum of the chip can be flexibly adjusted by the integrated FP microcavity structure, and the spectral resolution can be designed to be within 1 nm. In addition, each detection channel of the multispectral detector chip has a flexible arrangement, which can be set as one-dimensional array, two-dimensional array or super-pixel arrangement according to requirements. The unit size can be reduced to the size of the detector pixel. The SWIR multispectral detection chip has no moving parts, is especially suitable for portable micro-miniature spectroscopy instruments, and has broad application prospects in the fields of gas detection, food safety, agricultural production, biomedicine.



芯片级短波红外光谱探测器





图1 (a)片上集成FP微腔分光阵列的短波红外多光谱探测芯片结构示意图;(b)短波红外多光谱探测芯片;(c)该探测芯片的16通道响应光谱;(d)该探测芯片的气体探测响应光谱通道。

短红外波段拥有丰富的物质光谱信息,同时受大气环境中二氧化碳和水分子的影响小,具备穿云透雾的能力。因此,在食品、农业、遥感成像等领域具有广泛的应用前景。红外光谱探测通过探测器对物质的光波长信息进行探测分析,能够精准识别人眼无法判断的物质成分信息,就如孙悟空的火眼金睛一般,能够透过表象看本质。

然而,常规探测器本身不具备光谱识别能力,需要借助光栅、棱镜、干涉仪等分光装置来区分目标的波长信息。传统的多光谱探测系统,探测器和分光装置是分立的,需要机械组装并扫描,体积大、系统复杂,并不适用于现场检测和便携应用等场景。

近年来,光子晶体、超材料和量子点等滤光方式能够将光谱探测系统缩小至芯片量级,具有良好的应用前景。但在可见和近红外波段,光子晶体、超材料的亚波长结构显著增加了工艺难度和制作成本;而量子点滤光阵列在合成制备和光谱稳定性方面受到制约。相比之下,基于法布里-珀罗(FP)微腔的像素级分光阵列具有透射光谱可精确调控、工艺难度低、结构稳定和成本低等优点。

中科院上海技术物理研究所王少伟研究员团队提出了一种在铟镓砷(InGaAs)焦平面探测器上单片集成FP微腔的短波红外多光谱探测芯片。相关研究成果发表在Chinese Optics Letters 2022年第20卷第6期上(Zhiyi Xuan, Qingquan Liu, Zhuangzhuang Cui, Songlei Huang, Bo Yang, Chenlu Li, Shaowei Wang, Wei Lu. On-chip short-wave infrared multispectral detector based on integrated Fabry–Perot microcavities array[J]. Chinese Optics Letters, 2022, 20(6): 061302),并被选为当期封面。

InGaAs焦平面是目前最主流和最具应用前景的短波红外探测器,但自身无法识别光谱信息。FP微腔阵列制备与CMOS工艺兼容,非常易于与InGaAs焦平面探测器单片集成。研究团队通过早期自创的组合刻蚀的方法将具有不同窄带透射光谱的像素级 FP 微腔阵列单片集成到InGaAs焦平面探测器上,每个FP微腔结构对应一个光谱探测通道,以实现芯片级短波红外多光谱探测。通过4次组合刻蚀工艺,在64 × 64像元的InGaAs焦平面探测器上单片集成了16个不同光谱通道的像素级FP微腔结构,如图1(a)所示。当入射光照射到芯片上时,先在FP微腔内发生谐振并进行波长选择,透射波长的光被InGaAs焦平面探测器所接收。通过调控不同像元上对应的FP微腔腔层厚度,可以实现16通道片上多光谱探测。

此外,分光与探测一体化的实现,极大地简化了短波红外光谱探测仪的结构,使其体积最小化,其尺寸仅有2 mm2,与原本InGaAs探测器的尺寸相当,见图1(b)。

研究团队首次演示了在1400 ~ 1700 nm短波红外波段,多光谱探测芯片的光谱识别能力,成功识别了16种波长的光谱信息,如图1(c)所示。图1(d)为该芯片与相关气体特征吸收光谱对应的通道,可对CO2、C2H2、H2S、CH4等多种气体同时进行探测。该短波红外多光谱探测芯片将原本只能响应光强信息的探测器,升级为可以识别不同光谱信息的探测芯片,从而具备了识别物质成分的能力。

该多光谱探测芯片的光谱通道数量还可随加工次数呈指数型增加,7次组合刻蚀,即可实现128通道更宽范围的光谱探测。同时,各通道响应光谱的带宽、中心波长等可由所集成的FP微腔结构灵活调控,光谱分辨率可达1 nm。此外,该多光谱探测芯片各探测通道具有灵活的排布方式,可根据需求设置为一维线列、二维面阵或超像元排列,单元尺寸可缩小到单个探测器像元大小,无任何移动部件,特别适合于便携式微小型光谱仪器,在气体探测、食品安全、农业生产、生物医学等领域具有广泛的应用前景。