-
InGaAs/InP APD (GD5522-SM-FC/UPC,重庆光电技术研究所)为单模光纤耦合入射光子。在线性模式下分析测试了InGaAs/InP APD的响应带宽。在光脉冲宽度50 ps的情况下,使用带宽4 GHz的DSA 70404 C系列示波器在APD线性模式测得结果如图1所示。上升时间
$ \tau $ 约为151.4 ps,由$ BW=0.35/\tau $ 计算可知APD的响应带宽约为2.3 GHz。考虑到盖革模式比线性模式增益高得多,根据增益与带宽的关系,盖革模式下的带宽有所下降。最终将门控频率设置在1.0~2.0 GHz。 -
量子探测器层析(Quantum Detector Tomography, QDT)[19]省去了传统“自下而上”的标定方法。将量子探测器视为“暗箱”,由已知的输入态和测量得到的输出结果来获得正定算子取值测度(Positive Operator-valued Measure, POVM)矩阵[20],可全面表征探测器。
由波恩定则[21],探测器输出结果为
${{s}}$ 的概率由以下公式表示:$$ {P}_{\rho ,s}=Tr\left(\widehat{\rho .} \widehat{\textit{Π }}_{s}\right) $$ (1) 式中:入射场密度为
$\; \widehat{\rho } $ ;$ {P}_{\rho ,s} $ 探测输出结果为${s} $ 的概率;${\widehat{\textit{Π }}_{s}}$ 为POVM矩阵。QDT实验装置图如图2所示,用脉冲激光器调节数字衰减器使得每脉冲光子数直至探测器输出饱和计数,根据输入平均光子数
$\; \mu $ ,探测器探测到光子输出计数$ {t}_{\mu } $ ,计算探测器输出概率:$$ {P}_{{\rm{on}}}\left(\mu \right)=\frac{{t}_{\mu }}{T}\times 100 \text{%} $$ $$ {P}_{{\rm{off}}}\left(\mu \right)=1-{P}_{{\rm{on}}}\left(\mu \right) $$ (2) 式中:T为激光器参考频率。
为了进一步研究探测器的量子性能,引入Wigner函数[22],Wigner函数是动量与位置联合的准概率分布,原点函数值负值表示探测器具有量子特性[23]。
对应于探测器POVM矩阵
$ {\mathrm{\Pi }}_{z} $ 的Wigner函数可以表示为:$$ {W}_{z}\left(x,p\right)={\sum }_{\mathrm{n}=z}^{M}{\mathrm{\theta }}_{\mathrm{z},n}{W}_{n}\left(x,p\right) $$ (3) 式中:
$M$ 为截止光子数;$ {\theta }_{z,n} $ 为当n个光子入射时探测器探测到z个光子的条件概率;$ {W}_{n}\left(x,p\right) $ 为每个光子数态对应的Wigner函数;$ {L}_{n} $ 为拉盖尔多项式。$$ {W}_{n}\left(x\text{,}p\right)=\left[\frac{{\left(-1\right)}^{n}}{2\pi }\right]{{\rm{e}}}^{-\tfrac{{p}^{2}+{x}^{2}}{2}}{L}_{n}\left({p}^{2}+{x}^{2}\right) $$ (4) -
InGaAs/InP APD的误计数主要包括暗计数和后脉冲。暗计数是指暗计数是 APD 在没有外界光子入射时,APD 由于自身及外界噪声影响引起的误计数。后脉冲概率
$ {P}_{ap} $ 定义为一段时间内由前一个光子探测而后产生的后脉冲计数的概率[24]:$$ {P}_{ap}=\frac{({I}_{NI}-{I}_{D})R}{{I}_{PH}-{I}_{NI}} $$ (5) 式中:
$ {P}_{ap} $ 为后脉冲概率;$ {I}_{PH} $ 为激光器打开时,光子照射的门脉冲中的计数;$ {I}_{NI} $ 为没有光子入射的门脉冲中的计数;$ {I}_{D} $ 为激光器关闭时,每个门脉冲中的计数;R为门脉冲重复频率与光脉冲重复频率的比值。采用时间相关光子计数器(Time Correlated Single-photon Counter, TCSPC)对单光子探测器的后脉冲进行测量。激光脉冲的同步信号(25 MHz) 送入“Start”端口开始计数,单光子探测器的输入信号送入其 “Stop”端口。根据公式(5)计算即可获得后脉冲概率。
在制冷温度设定−30.0 ℃,光脉冲重复频率为25.0 MHz,光在耦合APD之前衰减至0.1光子每脉冲。调节APD两端直流偏置电压至 10.0%的探测效率,测试了不同重复频率下,APD的误计数和后脉冲。如图4(a)所示,可以看到随着探测频率的提升,暗计数率和后脉冲率都在升高。后脉冲在超过1.50 GHz探测频率后变化明显。暗计数在1.50 GHz探测频率下分别为2.04、2.05、2.08×10−7/gate,暗计数变化很小维持在同一水平。暗计数超过1.50 GHz后最大只有2.5×10−7/gate, 没有数量级的改变,仍然处于很小的范围。低暗计数将给系统带来很小的噪声从而提升探测器性能。
图 4 (a) 在10%探测效率下的暗计数和后脉冲概率;(b) 在20%探测效率下不同探测频率的测试图
Figure 4. (a) Dark counting and afterpulse probability at 10.0% detection efficiency; (b) Test chart of different detection frequencies at 20% detection efficiency
鉴于1.00~1.5 GHz探测频率下,暗计数和后脉冲都比较小,随后又在此条件的基础上,提高探测效率,测试了不同探测频率下后脉冲的情况。如图4(b)所示,整体来看,随着探测效率的提升,不同探测频率下的后脉冲都在增加,探测效率越高增加的越明显。图中的虚线是在20.0%探测效率时,1.00、1.25、1.50 GHz探测频率的后脉冲分别为4.46%、6.62%、6.63%。在20.0%以内,后脉冲变化很小,探测效率超过20.0%后,分开的很明显。在10%探测效率这条虚线上,1.75 GHz和2.00 GHz后脉冲分别为10.4%和36.6%。高探测频率后脉冲很高的原因:是因为探测频率已经接近APD响应的极限带宽,无法快速响应信号。
-
InGaAs/InP APD是半导体器件,工作温度对其性能的影响非常大。通过上位机向FPGA发送制冷指令,分别设置了制冷温度为−50.0、−40.0、−30.0、−20.0、−10.0、0.0、10.0 ℃和室温25.0 ℃这几个比较有代表性的温度点。下面测试了1.50 GHz探测频率在20.0%探测效率时不同工作温度下直流偏置电压、后脉冲率、暗计数率的变化。
如图5(a)所示,加载在APD的直流偏置电压在−40.0~25.0 ℃呈现线性趋势,在−50.0 ℃有所下降。后脉冲整体随着温度提升而呈现下降趋势,后脉冲在−50.0 ℃到室温这个范围基本维持在6.6%。原因是APD 工作在低温时,雪崩的过程中倍增层载流子容易被陷阱能级捕获,而引发较高后脉冲。随着温度提升,倍增层的场强增大,载流子离化率大[25],被捕获的概率降低,从而后脉冲率也降低。
图 5 20.0%探测效率下不同温度下的性能测试图
Figure 5. Performance test chart at different temperatures with 20.0% detection efficiency
温度升高时APD 暗计数因热噪声激发的暗电流增大,误计数增大,而光生载流子会被暗电流湮没,这使得在常温下对单光子的探测极为困难。整体暗计数理论是随温度升高而升高的。但是暗计数在−50.0~−30.0 ℃中出现了反转情况。温度在−30.0 ℃时,暗计数为6.7×10−7/gate,但在此温度前后步进出现了数量级变化,而且出现了反转情况。推测其可能的原因,一是后脉冲在−30 ℃相对于热噪声占据了主导地位,从而引起反转;二是在确定的短死时间下,载流子被某一能级俘获后,在温度较高时才开始释放引发,温度较低时未释放。
-
饱和计数率标定的方法为光子标定法,激光器发出的连续光经过衰减器衰减送入单光子探测器[24-26]。在调节偏置电压使其工作在20.0%探测效率时,通过调节数字光学衰减器(EXFO-LTB-1)增加平均入射光子数,如图6所示,探测器的饱和计数率约为56.2 MHz。饱和计数率反映了探测器探测光子通量的能力,通常为死时间的倒数。提升饱和计数率死时间相应会减少,这样也会导致后脉冲概率升高。
采集到不同平均光子数对应的输出概率后,用最大似然估计算法对公式(1)线性反演重建POVM矩阵。
为了简化计算只考虑了探测效率和暗计数率,建立对应的POVM理论模型:
$$ \begin{split} {{\textit{Π } }}_{{\rm{off}}}=& {{\rm{e}}}^{-v}{\sum }_{n=1}^{M}{\left(1-\eta \right)}^{n}|n > < n\text{|}{} \\ & {\textit{Π } }_{{\rm{on}}}=1-{{\textit{Π } }}_{{\rm{off}}} \end{split} $$ (6) 式中:
$ \eta $ 为探测效率;$ \nu $ 为暗计数/gate。从图7(a)中发现实验数据点与QDT拟合的概率分布几乎重合,可以表明重新构建的POVM与实际量子探测器的量子特征有很好的一致性。QDT拟合和基于理论POVM拟合的概率分布也很接近,表明建立的理论模型是符合的。在光子数为0~5时,吻合度很高,三组数据具有相同的趋势。5~20光子实际与理论的曲线有所偏离。由于理论模型忽略了后脉冲概率,随着入射光子数的增加到一定数量,未被及时淬灭的光生载流子也随之增加到一定程度,导致后脉冲概率对探测器输出概率产生影响,而使理论模拟计算得到探测到光的概率相对偏高。
图 7 (a) QDT层析、拟合与理论图; (b) Wigner函数图
Figure 7. (a) QDT chromatography, fitting and theoretical diagram; (b) Wigner function diagram
然后将基于QDT重建的POVM矩阵带入Wigner函数,如图7(b)所示,Wigner函数在原点处为负值验证了在高输入平均光子数时候并未破坏探测器的量子相干特性,表明探测器具备量子探测能力。
-
对探测器的稳定性比较关心的一是探测效率的稳定性,二是探测器的工作温度稳定性。这些都是探测器实用化的必要条件。前者是整个探测器的计数稳定程度,后者主要是指APD工作温度的稳定性。计数的跳动太大或者温度变化太大都会降低探测器的实用性。这里,采用方差来评估这一参数。
在探测频率为1.50 GHz时,探测效率设置在20.0%,12 h稳定性测试如图8所示。探测效率整条曲线的最低与最高探测效率之差为0.6%。探测效率的方差为1.0%。温度曲线上下浮动基本维持在±0.1 ℃,曲线整体集中分布,性能十分稳定。
Low-noise GHz InGaAs/InP single-photon detector (invited)
-
摘要: InGaAs/InP雪崩光电二极管(APD)体积小、功耗低、响应速度快,被广泛应用于近红外单光子检测。文中分析了APD雪崩及噪声信号的频谱分布特征,提出了正弦门控结合低通滤波级联方案,噪声抑制比超过40 dB, 实现了1~2 GHz高性能探测。当工作速率为1.5 GHz,探测效率设置为20.0%时,后脉冲概率为6.6%,暗计数率仅为6.7×10−7/gate。此外,集成了APD高速门控产生及延时调节模块,温度反馈稳定控制模块,实现了单光子探测器12 h稳定运行,计数标准差仅为1.0%。最后,为了更完整的描述探测器的量子特征,引入量子探测器层析技术进行标定,重新构建了其正值算符测度矩阵以及对应的Wigner函数,为其在量子通信、量子计算等量子信息技术的应用中提供支撑。Abstract:
Objective With the development of quantum information science, laser radar and deep space detection, the traditional linear photoelectric detection technology has been unable to meet the needs of sensitive optical signal detection. The single-photon detection technology has gradually become an important research in the fields of weak light detection. InGaAs/InP avalanche photodiodes (APDs) are widely used in near-infrared single-photon detection due to the small size, low power consumption and fast response. The detection rate of most commercial InGaAs/InP detectors is at the level of 100 MHz, which cannot meet the application requirements for high counting rate. Meanwhile, low noise of the APD will bring smaller false counts to the system and further improve the performance. Therefore, a low-noise InGaAs/InP single-photon detector operating at the repetition frequency of GHz was demonstrated. Furthermore, the whole detector is evaluated with the quantum detector tomography technology, providing support for its application in quantum information technology such as quantum communication and quantum computation. Methods In order to determine the detection frequency of gating signals, the response bandwidth of the APD is analyzed in the linear mode, and the bandwidth range is calculated to be 1-2 GHz. The spectral distribution characteristics of APD avalanche and noise signals are analyzed in the Geiger mode. It could be figured out that the noise is mainly distributed in the gating frequency and its harmonic frequencies, while the avalanche signal is mainly distributed below 1 GHz. Therefore, a cascade scheme of sine wave gating combined with low-pass filtering is proposed (Fig.3). The detector comprises high-speed gate generation and delay regulation module, temperature feedback control module, etc. Sine wave gating could be precisely controlled from many parameters which include frequency, amplitude, delay in a wide range. Feedback is added in the temperature control module to improve the stability of the detector. In addition, quantum detector tomography (Fig.2) is introduced to calibrate the detector, which is regarded as a "dark box". The positive operator-value measuring matrix can fully characterize the detector, which is obtained from input states and output results. The Wigner function is employed to describe whether the detector has quantum properties at high input photons. Results and Discussions Sine wave gating combined with low-pass filtering is designed in the system, and signal-to-noise ratio is over 40 dB. The relationship between the detection efficiency and the afterpulse probability at the frequencies of 1-2 GHz is recorded. When the working rate is 1.5 GHz and the detection efficiency is set to be 20.0%, the afterpulse probability is 6.6% with the dark count rate of only 6.7×10−7 per gate (Fig.4). At constant detection efficiency of 20.0%, the DC bias voltage of the APD increases with temperature, showing a linear trend. While the afterpulse probability decreases, showing a contracting trend. The dark count rate degrades with the decrease of temperature and the trend is reversed at −30 ℃ (Fig.5), which might be related to high afterpulse or the intrinsic defection of APD. During the 12-hour test period, the detector performs perfectly stable and the variance of detection efficiency is 1% (Fig.8). Quantum detector tomography technology is employed to verify that high background noise does not affect the quantum properties (Fig.7). Conclusions A GHz low noise InGaAs/InP detector is designed, and its detection efficiency, false count, saturation count rate and stability are explored. Based on the analysis of the response bandwidth of APD, a cascade scheme of sine wave gating combined with low-pass filtering is determined, realizing a low noise single photon detection below 2 GHz. In addition, quantum detector tomography technology is employed to calibrate the detector and verify its quantum properties. The structure of the detection technology is simple and the detector can run stably in the long term, which provides strong support for the practical application of single photon detector in deep space communication, laser mapping, optical time domain reflection and other fields. -
-
[1] Wang S, Chen W, Yin Z Q, et al. Practical gigahertz quantum key distribution robust against channel disturbance [J]. Optics Letters, 2018, 43(9): 2030-2033. doi: 10.1364/OL.43.002030 [2] Gao L, Zhang X L, Ma J T, et al. Quantum enhanced Doppler LiDAR based on integrated quantum squeezed light source [J]. Infrared and Laser Engineering, 2021, 50(3): 20210031. (in Chinese) doi: 10.3788/IRLA20210031 [3] Liang Y, Xu B, Fei Q, et al. Low-timing-jitter GHz-gated InGaAs/InP single-photon avalanche photodiode for LiDAR [J]. IEEE Journal of Selected Topics in Quantum Electronics, 2021, 28(2): 1-7. [4] Shen G, Zheng T, Li Z, et al. High-speed airborne single-photon LiDAR with GHz-gated single-photon detector at 1550 nm [J]. Optics & Laser Technology, 2021, 141: 107109. [5] Qiu Jiawen, Wang Qiang, Ma Ji'nan. Deep space exploration technology [J]. Infrared and Laser Engineering, 2020, 49(5): 20201001. (in Chinese) doi: 10.3788/irla.1_invited-qiujiawen [6] Zhang J, Itzler M A, Zbinden H, et al. Advances in InGaAs/InP single-photon detector systems for quantum communication [J]. Science & Application, 2015, 4(5): 1-13. [7] Liu Junliang. Development and performance of single photon detector based on InGaAs(P)/InP APD [D]. Jinan: Shandong University, 2018. (in Chinese) [8] Jiang Wenhao. Research on high performance semiconductor single photon detector [D]. Hefei: University of Science and Technology of China, 2018. (in Chinese) [9] Yuan Z L, Kardynal B E, Sharpe A W, et al. High-speed single photon detection in the near infrared [J]. Applied Physics Letters, 2007, 91(4): 041114. doi: 10.1063/1.2760135 [10] Chen X, Wu E, Xu L, et al. Photon-number resolving performance of the InGaAs/InP avalanche photodiode with short gates [J]. Applied Physics Letters, 2009, 95(13): 131118. doi: 10.1063/1.3242380 [11] Namekata N, Sasamori S, Inoue S. 800 MHz single-photon detection at 1550-nm using an InGaAs/InP avalanche photodiode operated with a sine wave gating [J]. Optics Express, 2006, 14(21): 10043-10049. doi: 10.1364/OE.14.010043 [12] Tada A, Namekata N, Inoue S. Saturated detection efficiency of single-photon detector based on an InGaAs/InP single photon avalanche diode gated with a large amplitude sinusoidal voltage [J]. Japanese Journal of Applied Physics, 2020, 59(7): 72004. doi: 10.35848/1347-4065/ab9625 [13] Wang Tianye, Fei Qilai, Xu Bo, et al. InGaAs/InP single photon detector with adjustable GHz repetition rate [J]. Journal of Infrared and Millimeter Waves, 2021, 40(6): 840-846. (in Chinese) [14] Gao Jiali. Design of high speed near-infrared InGaAs/InP single photon detector [J]. Journal of Optoelectronics · Laser, 2016, 27(10): 1042-1046. (in Chinese) [15] Jiang W H, Liu J H, Liu Y, et al. 1.25 GHz sine wave gating InGaAs/InP single-photon detector with a monolithically integrated readout circuit [J]. Optics Letters, 2017, 42(24): 5090-5093. doi: 10.1364/OL.42.005090 [16] Fang Y Q, Chen W, Ao T H, et al. InGaAs/InP single-photon detectors with 60% detection efficiency at 1550 nm [J]. Review of Scientific Instruments, 2020, 91(8): 083102. doi: 10.1063/5.0014123 [17] Liu C, Ye H F, Shi Y L. Advances in near-infrared avalanche diode single-photon detectors [J]. Chip, 2022, 1(1): 100005. [18] Liang Y, Fei Q, Liu Z, et al. Low-noise InGaAs/InP single-photon detector with widely tunable repetition rates [J]. Photonics Research, 2019, 7(3): A1-A6. doi: 10.1364/PRJ.7.0000A1 [19] Lobino M, Korystov D, Kupchak C, et al. Complete characterization of quantum-optical processes [J]. Science, 2008, 322(5901): 563-566. doi: 10.1126/science.1162086 [20] Chen Yu. Quantum detector chromatography calibration based on multi-pixel photon counter and its application [D]. Shanghai: East China Normal University, 2021. (in Chinese) [21] Banaszek K, D’ariano G M, Paris M G A, et al. Maximum-likelihood estimation of the density matrix [J]. Physical Review A, 1999, 61(1): 010304. doi: 10.1103/PhysRevA.61.010304 [22] Leonhardt U. Measuring the Quantum State of Light[M]. Cambridge: Cambridge University Press, 1997. [23] Schleich W P. Quantum Optics in Phase Space[M]. New York: Wiley, 2011. [24] Liang Yan. High speed single photon detection Method based on InGaAs/InP APD and its application [D]. Shanghai: East China Normal University, 2014. (in Chinese) [25] Tosi A, Scarcella C, Boso G, et al. Gate-free InGaAs/InP single-photon detector working at up to 100 Mcount/s [J]. IEEE Photon J, 2013, 5(4): 6801308. doi: 10.1109/JPHOT.2013.2278526 [26] Wang Yong. Single-photon avalanche photodiode calibration system based on FPGA [D]. Shanghai: East China Normal University, 2018. (in Chinese)