Novel InGaAs/InAlAs travelling wave avalanche photodiode for ultra high speed photonic applications

互联网的爆炸性使用增加了对高速光电探测器的需求，以在2.5Gb/s、10 Gb/s或40 Gb/s光通信系统中将光信号转换为电信号。可以检测极低光级的高灵敏度光电探测器对于提高信号质量和增加传输距离并从而降低这些系统的成本非常重要。人类基因组测序、医学成像和未来的量子计算也需要它们。半导体雪崩光电二极管（APD）可以提供这些应用所需的高灵敏度，并且坚固、廉价、紧凑且高效。在APD中，电子-空穴对可以触发电子和空穴的雪崩（如雪崩效应）。该乘法过程提供了提高APD灵敏度的内部增益。在大多数半导体中，在倍增过程中存在显著的统计波动，从而引起不需要的过量噪声。低的多余噪声是很重要的，这可以通过使用电子比空穴更容易繁殖的材料来实现（反之亦然）。在光通信系统中，波长为1550 nm的红外光用于传输信息，以最大限度地减少光纤中的损耗。正因为如此，我们将不得不使用APD，该APD使用称为InGaAs的半导体作为吸收层来检测1550 nm的红外光，并使用另一种半导体InAlAs作为倍增层来产生雪崩效应。InAlAs在1550 nm处产生的过量噪声比目前可用的商业APD少，因为在这种材料中电子比空穴更容易繁殖。从我们的研究中我们知道，我们可以通过使用非常薄的亚微米倍增层（<头发直径的1/50）并仔细设计InAlAs倍增层中的电场分布来进一步降低过量噪声。我们已经表明，这些技术可以减少多余的噪声，从而提高灵敏度APD。在这个项目中，我们将增长，制造和测试一些不同的设计，以尽量减少我们的APD中的多余噪音。APD的另一个重要参数是带宽，因为它们以高达40 Gb/s的非常高的数据速率工作。为了实现高带宽，我们将结合以下创新：首先，我们将使用非常薄的亚微米吸收和倍增层，以减少电子和空穴的渡越时间。为了确保红外光被有效地吸收，我们将把光限制在一种称为光波导的特殊结构中。通过将APD与波导集成，红外光将被有效吸收，从而产生高速高灵敏度的波导APD。其次，我们将设计的波导APD的结构称为行波APD，它具有电力传输线的特性。这种结构将确保高速信号从我们的APD有效地传输到外部电路。我们将使用一个理论模型来预测行波的特性，以确保它们可以在高达40 GHz的频率下产生高灵敏度。我们将进行几个实验来了解我们需要生产高速高灵敏度光电探测器。将对APD进行测量，以监测倍增如何随温度变化（从室温到-250摄氏度），以及当信号频率增加到40 GHz时倍增如何变化。这将为我们提供生产用于光通信系统的非常高灵敏度的光电探测器所需的数据和理解，以及许多其他应用，如医疗成像，环境污染物监测，制造业缺陷监测和影响我们日常生活的许多其他领域。

The explosive use of internet has increased the demand for high speed photodetectors to convert optical signal to electrical signal at 2.5Gb/s, 10Gb/s or 40Gb/s optical communication systems. High sensitivity photodetectors that can detect very low light level are important to improve the signal quality and increase the transmission distance and hence lower the cost of these systems. They are also required in sequencing of the human genome, in medical imaging and in future quantum computing. Semiconductor avalanche photodiodes (APDs) can offer the high sensitivity required for these applications and are robust, cheap, compact and efficient. In APDs an electron-hole pair can trigger an avalanche of electrons and holes (like the snow avalanche effect). This multiplication process provides an internal gain which improves the sensitivity of APDs. In most semiconductors, there is significant statistical fluctuation in the multiplication process giving rise to unwanted excess noise. Low excess noise is important and this can be achieved by using a material in which electrons can multiply much easier than holes (or vice versa).In optical communication systems infrared light with a wavelength of 1550nm is used to transmit information to minimise loss in the optical fiber. Because of this we will have to use APDs fabricated using a semiconductor called InGaAs as an absorption layer to detect infrared light of 1550nm and another semiconductor, InAlAs, as the multiplication layer to produce the avalanche effect. InAlAs produces less excess noise compared to currently available commercial APDs at 1550nm because electrons can multiply much easier than holes in this material. From our research we know that we can further reduce the excess noise by using very thin sub-micron multiplication layer (< 1/50 of the diameter of our hair) and carefully engineer the electric field profile in the InAlAs multiplication layer. We have shown that these techniques can reduce the excess noise leading to higher sensitivity APD. In this project we will grow, fabricate and characterise a number of different designs to minimise the excess noise in our APDs. Another important parameter of APDs is the bandwidth since they operate at very high data rate up to 40Gb/s. To achieve high bandwidth we will incorporate the following innovations; Firstly, we will use very thin sub-micron absorption and multiplication layers to reduce the electron and hole transit times. To ensure that the infrared light is absorbed efficiently we will confine the light in a special structure called optical waveguide. By integrating the APD with a waveguide the infrared light will be efficiently absorbed to yield high speed high sensitivity waveguide-APD. Secondly, we are going to design the waveguide-APD into a structure called travelling wave-APD which has characteristics of an electrical transmission line. This structure will ensure that high speed signals are transmitted efficiently from our APD to the external circuit. We will use a theoretical model to predict the characteristics of the travelling-wave to make sure that they can produce high sensitivity at frequency up to 40GHz.There are several experiments that we will perform to give us the understanding we need to produce a high speed high sensitivity photodetector. Measurements on the APDs to monitor how the multiplication changes with temperature ranging from room temperature down to -250 degree Celcius as well as how the multiplication changes when the signal frequency is increased up to 40GHz will be carried out. This will provide us the data and understanding required to produce very high sensitivity photodetectors for optical communication systems as well as many other applications such as for medical imaging, environmental pollutant monitoring, defects monitoring in manufacturing and many other areas that affect our daily lives.