Acquisition of Instrument for the Study of Artificial Atoms

[700818]卡斯特纳计算机的能力随着时间呈指数级增长，很大程度上是因为设备的尺寸随着时间呈指数级下降，导致芯片上晶体管的密度更高。不幸的是，生产这些设备的设备成本也呈指数级增长，尽管设备市场增长迅速，但却没有呈指数级增长。因此，我们正在迅速接近新一代传统设备的设备可能过于昂贵而无法建造的时代。过去，新奇的设备似乎无法与传统设备竞争。然而，未来制造设备的经济限制为研究可能具有成本优势的设备类别提供了新的动力。单电子器件可能是这类器件之一。在过去，单电子器件只能在非常低的温度下工作，而最近关于在室温下工作的报道使有用的应用更接近现实。这个项目建立在麻省理工学院过去的工作基础上，在过去的工作中，人们发现当晶体管被制造得非常小的时候，它会以一种不寻常的方式表现出来。传统的晶体管在加入电子时只打开一次，而这样的晶体管需要加入大约103个电子才能从关断状态变为导通状态，而这种新型晶体管每次加入一个电子就会再次打开和关闭。单电子晶体管是通过隔离一个非常小的电子液滴制成的，这样电子只能通过量子力学隧道才能进出液滴。麻省理工学院的工作表明，这种晶体管的行为就像一个人造原子，因为液滴中的电子数量和能量都是量子化的。到目前为止，制造出来的晶体管使用的是表面带有金属电极的半导体砷化镓，用电子束光刻技术来限制电子液滴。由于这种砷化镓结构中的电子通常在表面以下100nm处，因此电子液滴的半径通常为~100nm。在液滴中添加额外电子的能量为r-1，这限制了单个电子在温度低于~1K时的操作，r=100nm。该项目的目标是探索提高单电子晶体管工作温度的方法，并更好地了解其物理特性。特别是，麻省理工学院的研究小组最近发现，在高磁场下，被限制在人造原子中的电子液滴发生了一系列相变。为了进一步研究这些相变，本文提出了获取高场磁体和专用制冷机的方法。新的人造原子将用Si代替GaAs制成。由于这些装置中的电子距离表面可以接近5纳米，因此可以产生更小的液滴，从而产生更高的能量和温度尺度。该装置将与麻省理工学院林肯实验室合作制造，使用在空军赞助下开发的杰出电子束和光学光刻能力。这样，国家科学基金的作用将被强烈地放大。除了更高的工作温度外，Si还提供了一个优势，即单电子器件可以很容易地与传统器件集成。期望硅中的单电子器件能在高于液氮温度的温度下工作。研究这些小结构中的电子状态需要更高的磁场。还提出了在砷化镓中制造电子液滴更小的器件。这将通过生长电子更接近表面的结构来实现。已经与以色列的魏茨曼研究所建立了合作关系，在那里已经开发了制造这种结构的专业知识。就像与林肯实验室的合作一样，这放大了国家科学基金会资助的影响。单电子器件的优点是它们可以电容耦合。有人提出，正因为如此，一组这样的设备可能起到联想记忆的作用。也就是说，任何一组输入电压接近，但不完全等于，一个特殊的集合将给出相同的输出。提出了这种阵列的制备和研究。虽然第一个这样的记忆将是只读的，但理解它们的物理性质可能也会导致书写的方式。* * *

9700818 Kastner The power of computers has increased exponentially with time in large part because the size of devices has decreased exponentially with time, resulting in a higher density of transistors on a chip. Unfortunately, the cost of facilities for producing these devices has also increased exponentially, and although the market for devices has grown rapidly, it has not grown exponentially. Thus we are rapidly approaching the time when facilities for new generations of conventional devices may be too expensive to build. In the past it seemed that novel devices could not compete with conventional ones. However, the economic limitation to future fabrication facilities provides a new motivation for research into classes of devices that may have cost advantages. Single- electron devices may be one of these classes. Whereas single- electron devices have, in the past, operated only at very low temperatures, recent reports of operation near room temperature bring useful applications much closer to reality. This project builds upon past work at MIT in which it was discovered that when a transistor is made very small it behaves in an unusual way. Whereas a conventional transistor turns on only once as electrons are added to it, and whereas such a transistor requires the addition of about 103 electrons to change from the off-state to the on-state, the new transistor turns on and off again every time an electron is added to it. The single electron transistor is made by isolating a very small droplet of electrons in such a way that electrons can move into and out of the droplet only by quantum mechanical tunneling. The MIT work has shown that such a transistor behaves like an artificial atom, in that the number of electrons in the droplet and the energy are both quantized. The transistors made so far have used the semiconductor GaAs with metal electrodes on the surface, patterned using electron-beam lithography, to confine the droplet of electrons. Beca use the electrons in such GaAs structures are typically 100nm below the surface, the droplet of electrons has typically been ~100nm in radius r. The energy for adding an extra electron to the droplet scales like r-1, and this limits single electron operation to temperatures below ~1K for r=100nm. The goal of this project is to explore ways of increasing the operating temperature of single electron transistors and to better understand their physics. In particular, the MIT group has recently discovered a series of phase transitions in the droplet of electrons confined in the artificial atom at high magnetic fields. The acquisition of a higher-field magnet and dedicated refrigerator are proposed here to study these phase transitions further. %%% New artificial atoms will be fabricated using Si instead of GaAs. Since the electrons in these devices can be as close as 5nm from the surface, much smaller droplets and, consequently, much higher energy and temperature scales are expected. The devices will be fabricated in collaboration with MIT Lincoln Laboratory using the outstanding electron- beam and optical lithography capability developed there under Air Force sponsorship. In this way, the effect of the NSF funds will be strongly amplified. In addition to higher operating temperatures, Si offers the advantage that single electron devices can be easily integrated with conventional ones. It is expected that single electron devices in Si can be made to operate well above liquid nitrogen temperature. Higher magnetic fields are required for the study of the state of electrons in these smaller structures. It is also proposed to fabricate devices in GaAs in which the droplets of electrons will be smaller. This will be accomplished by growing structures in which the electrons are closer to the surface. A collaboration has been established with the Weizmann Institute in Israel where the expertise in making such structures has already been developed. Like the col laboration with Lincoln Laboratory, this amplifies the impact of NSF funding. Single electron devices have the advantage that they can be capacitively coupled. It has been proposed that, because of this, an array of such devices may function as an associative memory. That is, any set of input voltages close to, but not exactly equal to, a special set will give the same output. It is proposed that such an array be fabricated and studied. Although the first such memories will be read-only, understanding their physics may lead to ways of writing as well. ***