We use cookies to improve your experience. By your continued use of this site you accept such use. To change your settings please see our Privacy Policy.
Close

Non-destructive Analysis (NDA)

Circuit Edit (CKT)

Materials Analysis (MA)

Application Forms

Software

MA-tek FTP

Sustainability report

A Discussion of the Evolution and Challenges of Quantum Computing Technology

2021/11/20

Preface

Quantum computing has become a hot topic in today’s global technology industry. Like the silicon technology of the last century, it is believed that quantum computing will be the next generation of technology to change the world. Although the development of quantum technology is still in its early stages, tech giants like the United States-based Google, IBM, Intel and others are already investing in related research and development in order to get a head start and take on a leading role in the coming era of quantum hegemony.

 

“The next three to five years present an opportunity for Taiwan to develop its quantum computing,” said Young Liu, chairman of Hong Hai. The global quantum industry is gradually taking shape as quantum computing moves rapidly from theoretical basis verification to practical application. As home to the world’s best semiconductor engineering talents, Taiwan’s existing industrial resources are a great advantage. They mean that Taiwan has an excellent chance to develop and manufacture the relevant core components, “Silicon (Germanium) based qubits”. Those in the industry, the government and academia should all pay close attention and grasp this opportunity to help bring about Taiwan’s next semiconductor miracle.

 

MA-tek has invited quantum technology expert Professor Pei-Wen Li to be a columnist for our “New Technology Channel” and introduce the core hardware of quantum computing, the “Qubit”, and its application principles, development and technical challenges, etc. Professor Li received his Ph.D. in Electrical Engineering from Columbia University in 1994. She proceeded to conduct extensive research in the field of quantum computing for both the industry and academia. Most notably, she has contributed to many major research breakthroughs in combining semiconductor elements with quantum structures. In this article, Professor Li will guide readers through the wonderful world of qubits.

 

 

Director of R&D Center & Marketing Division, Chris Chen, 2021/10/12

 

 

 

 

A Discussion of the Evolution and Challenges of Quantum Computing Technology

 

Professor Pei-Wen Li

Institute of Electronics, National Yang Ming Chiao Tong University, Taiwan

 

 

 

Low-dimensional material engineering has opened a window for traditional bulk material engineering. Nano and even atomic level sensors, imaging technology, electronics, optics, quantum information, quantum computing, and even energy conversion components have opened up a wide range of applications. However, many of these applied technologies cannot be practiced relying solely on bulk material science and engineering.

 

We can divide quantum materials into four main categories according to the spatial dimensions of free movement that charges being available in the materials.

 

  • Materials in which the charge can move freely in three dimensions.
  • Two-dimensional quantum wells (where a barrier dimension restricts the movement of electric charges).
  • One-dimensional quantum wires (where the charge is limited in two dimensions so that there is only one dimension in which movement is allowed).
  • Zero-dimension quantum dots (where charge movement is restricted in all three dimensions).

 

Interestingly, when a three-dimensional bulk material is reduced to a zero-dimensional quantum dot system, in addition to the movement limitation, the charge undergoes a miraculous quantization phenomenon. In other words, the state density of the charge in the material will change from a continuous energy band to discrete energy levels.

 

Take, for example, a spherical quantum dot. Due to the limitations of three-dimensional, radially symmetrical electrostatic potential, the charge in a quantum dot can only be stored within certain energy levels and no longer seem to be in bulk. In materials, charges can carry any energy and appear in any spatial position generally. According to Schrodinger’s Equation, the discrete amounts of energy levels in a quantum dot are mainly determined by its diameter, shape and stress. To put it simply, the smaller the diameter of the quantum dot, the greater the quantum effect. Therefore, the discreteness of energy levels becomes more pronounced, as shown in Figure 1.

 


Figure 1. The electronic structure of a quantum dot is highly dependent on its diameter. When the diameter of a quantum dot is smaller than its Bohr radius, its electronic structure gradually changes from a continuous energy band to discrete energy states, and the gap between split energy levels gets wider.

This is what we call the quantum size effect. According to the quantum size effect, we can adjust the diameter of a quantum dot to adjust its electronic structure, charge distribution, electricity, optics, and multiple other physical properties. The functional features of the quantum size effect and low-dimensional materials have already successfully opened up many areas of quantum engineering design. As long as we can accurately control the shape, diameter, crystal quality, deformation/stress and spatial placement of a quantum dot, we can greatly improve and expand sensing, image display such as the quantum-dot TV, quantum computing and other applications.

 

 

 

One of the most influential computing and data storage technologies in the coming digital era.

Among the many applications of quantum dots, there is one of particular note—quantum computing. The two great masters on physics Stephen Wiesner and Feynman took the lead in the 1980s by proposing the requirements and concepts of quantum computing. In recent years, responding to advancements in artificial intelligence, machine learning, security encryption that ushering in the era of Big Data, people around the world are racing to propose quantum computing solutions that can achieve supercomputing speeds in order to process huge amounts of data.

 

In 2017, the US House of Representatives Science Committee declared the need to ensure “America’s dominance in quantum technology”. In 2018, the EU implemented its “Quantum Flagship” program. The Chinese Academy of Science and Alibaba joined forces to establish the Dharma Cloud Quantum Lab. Japan announced that it would provide free quantum neural network services, and the Canadian and Australian governments have invested hundreds of millions of dollars into research and development, etc. The Boston Consulting Survey has also announced that companies from various countries (Google, IBM, Intel, Microsoft, Honeywell, D-Wave, QxBranch, QCWare, QuTech, and 1Qbit, etc.), research institutions (MIT, Oxford, the University of New South Wales, Keio University, Delft Tech, imec, and Leti, etc.), and even large traditional industries (BASF and DowDuPont, etc.) are investing heavily in quantum computing/software development. Quantum computing is very likely going to become one of the most influential computing and data storage technologies in the coming digital era.

 

To conduct quantum computing, we need to do more than “upgrade or improve” existing computers. Quantum computing uses completely new algorithms of a completely different nature, so we need brand new software and hardware. There are three key building blocks for constructing quantum computing hardware. These are:

  • basic arithmetic unit¾ the qubit

  • control circuit
  • reading circuit [1]

 

In essence, it is the use of qubit Quantum Superpositioning, Quantum Entanglement, and related principles to create multiple combinations of quantum states. It can break through the limitations of the two states of “0” or “1” of classical calculation.

 

However, though the basic concept of quantum computing seems simple at first glance, the actual hardware and software technology threshold is extremely high. So far, the world has been unable to reach a consensus on the best approach to quantum computing hardware. Physicists, semiconductor engineers and computer scientists from around the world are all developing different types of qubits. Representative examples include the Superconductor, Ion Trap, Diamond, Photonics, quantum dots (QDs) and Topologic qubits. Most of the qubit types that have been shown are still in the laboratory prototype stage.

 

The main factors to consider when assessing the quality of a qubit include its decoherence time, scalability, fidelity, connectivity, operation temperature and mass production potential (as shown in Table 1) [2]. The decoherence time and scalability are particularly the keys to evaluating the success or failure of qubits in the initial stage.

  • Decoherence Indicates whether the quantum state will be affected by disturbances in the external environment and is gradually lost over time.
  • Scalability refers to the maximum number of qubits that can be created through expansion.

 

Classical computers use semiconductor process technology to expand the number of bits at the cost of increased surface area.Furthermore, the process integration is complex and the circuit/System design are difficult, etc. In quantum computing, however, just using process technology to expand the number of qubits is not enough. The biggest challenge we face is how to accurately control the coherence and entanglement of multiple qubits because every additional qubit increases the difficulty of entanglement exponentially. If a quantum chip cannot accurately control the coherence of multiple qubits then quantum computing would have no practical value. Therefore, we need a low-temperature CMOS circuit capable of real time manipulation and effective reading of the quantum state.

 

Table 1. A comparison of the key indicators, strengths and weaknesses of various types of qubits [2]

 

 

 

Epoch-making progress in quantum computing hardware

Fortunately, through the efforts of people from around the world, quantum computing hardware has seen many exciting developments over the past five years. In 2017, IBM introduced the 50-bit superconducting IBM-Q. Then, in 2018, Google claimed to have erected a milestone in quantum supremacy when it launched its 72-bit superconducting Bristlecone, which achieved a reading accuracy of 99%. In China, also during 2018, Alibaba introduced its 10-bit superconducting quantum computer. What’s more, Intel has announced its 49-bit Tangle Lake superconducting quantum chip.

 

Nevertheless, there are still many improvements to be made in the fidelity, expansion, and error correction of superconducting qubits. In addition, interferences in the operating environment (such as temperature and electromagnetic fields) is a major constraint on the development of quantum computing. For example, superconducting qubits can only operate in an extremely cold (mk) environment close to absolute zero. The related control circuit must therefore be placed outside of a large refrigerator and use RF components to control the superconducting qubits. However, the packaging technology required to maintain the long-term stability of superconducting qubit chips in extreme cold is very complicated. Recently, photonic quantum computing and ion-trap qubits have also repeatedly produced good results. However, a large optical table and numerous optical components must be used to manipulate photons, and an ion trap must be manipulated in an ultra-high vacuum environment. In other words, the control environments for both are still very severe.

 


Figure 3. Self-assembled germanium quantum dots formed using thermally oxidized silicon-germanium structures, embedded in silicon dioxide, silicon nitride, and silicon.

In addition to actively expanding the number of superconducting qubits, Intel is striving to build a scalable silicon qubit semiconductor system. The advantage of silicon qubits is that they can be directly integrated with CMOS control circuits, which would improve the feasibility of practical quantum computing applications. At present, there are a few different ways to manufacture silicon quantum dots. Self-assembled quantum dots can be grown epitaxially (as shown in Figure 3)[3]. They can also be manufactured using lithographic patterning technology (as shown in Figure 4) [4,5].

 

 

 

Figure 4. In SiO2/Si, Si/SiGe, and Ge/Si, two-dimensional electron gas quantum wells [4] or Si nanowires [5]; Multiple electrodes are defined by lithography, silicon or germanium quantum dots formed by voltage induction and their confinement barriers.

 

However, in order to implement practical silicon quantum dot bits and the related charge sensing elements, the shape, diameter, and interface quality between the quantum dots and their shells, and the type and placement of shells, etc. need to be accurately controlled. The Bohr exciton radius of silicon materials is about 5nm. This means that the diameter of silicon quantum dots needs to be smaller than 5nm in order to exhibit a significant level of energy quantization. However, it is still very challenging to fabricate silicon quantum dots smaller than 5nm even with the most advanced of existing lithography techniques. Another challenge lies in how to make nanoelectrodes that can directly manipulate a specific quantum dot. Limited by the resolution of lithography machines and the accuracy of layer-to-layer alignment, nanoelectrodes inevitably affect or touch other quantum dots around them.

 

 

 

Germanium (Ge) qubit development offers broader options for silicon-based quantum computing

Although two silicon spin qubits have been introduced thus far, silicon has poor spin-charge conversion efficiency and a low signal-to-noise ratio, which is not easy to read and can only operate at extremely low temperatures (<<4K). In recent years, the research and development of germanium (Ge) qubits, which are also group IV semiconductors, has also attracted much attention. This has opened up more options for realizing silicon-based quantum computing. This is because the Bohr exciton radius of germanium is about 25nm, which is much greater than silicon’s 5nm. Moreover, the spin-orbit coupling effect of germanium is very strong, and it’s very likely that the spin can be directly driven by electricity, which would improve our control feasibility and reliability. International research institutions currently focusing on the development of silicon or germanium qubits include the University of Wisconsin in the United States, Princeton University, the University of California, Berkeley, the Hughes Laboratory, Sandia National Laboratories, imec Belgium, the French CEA-Leti Institute, the Netherlands Institute of Technology, the University of New South Wales in Australia and Tokyo University in Japan, etc.

 

In addition to the technical difficulties in the preparation or growth of silicon and germanium quantum dots, there is also the need to be able to accurately analyze the structural properties of quantum dots, such as shape, size, crystalline state, deformation stress and shell/quantum dot interface, in real-time. Since the diameter of quantum dots is mostly within tens or even a few nanometers, they are wrapped in shells and host materials of different crystalline states. Therefore, to use high-magnification Transmission Electron Microscopes (TEM), Micro‐Raman spectroscopy or wide-angles X-ray Diffraction Spectroscopy (XRD), Convergent Beam Electron Diffraction (CBD) or Nanobeam Electron Diffraction (NBD), etc. to test and analyze the structural properties of quantum dots, we will need more sophisticated and specialized of processing methods on specimen sampling, preparation and acquisition. You may even need to combine atomic electron mapping with Electron Energy Loss Spectroscopy (EELS) and Energy Dispersive Spectroscopy (EDS) with the imaging result by Scanning Transmission Electron Spectroscopy (STEM) to collect the localized chemical composition and structural information you need in order to construct a complete quantum dot structure analysis.

 

For example, we experimented with the selective oxidation of “silicon germanium” nanostructures, which can produce self-formed nanolayer heterostructures that “integrated” with spherical germanium quantum dots, silicon dioxide, and silicon germanium. By combining the use of a Transmission Electron Microscope, Nanobeam Electron Diffraction, and atomic element mapping by Energy Dispersive Spectroscopy with Scanning Transmission Electron Microscopy, etc., we can clearly see that the germanium quantum dots are covered by a silicon dioxide shell with a thickness of about 2nm and excellent fidelity. Moreover, SiGe nanolayers have been generated underneath the germanium quantum dot/silicon dioxide shell (as shown in the Transmission Electron Microscope and Energy Dispersive Spectroscopy images in Figure 5).

 

Figure 5. An observation of self-formed heterostructures of germanium quantum dots/silicon dioxide shells/silicon-germanium nanolayer via Transmission Electron Spectroscopy and Energy Dispersive Spectroscopy.

 

 

 

The last mile to achieve quantum computing hardware technology

From a practical perspective, effectively manipulating qubits and accurately reading qubit states is the final stretch we need to cover in order to make quantum computing hardware technology a reality. This is because the feebleness coupling effect of quantum dots is very weak, resulting in very small potential changes (about <0.1mV) and current changes (about pA-100pA) in the quantum state. Moreover, the state of the qubit is easily influenced by disturbances of the surrounding electric potential or electromagnetic field (mV). Therefore, control of the input voltage must reach mV or even sub-mV accuracy. The greatest technical challenge lies in reading the weak current signal of the qubit in a low-temperature operating environment. In addition, in contrast to quantum dots, the large number of control electrodes needed occupy a considerable amount of area, resulting in parasitic effects between electrodes, such as stray capacitance, crosstalk, and current leakages; all of which generate large amounts of background noise. This greatly increases the difficulty of manipulating and reading the quantum state.

 

At present, many research institutes build external test systems by themselves, designed and assembled ad hoc to measure the quantum state of qubits. However, when assembling multiple arbitrary waveform generators (AWG), microwave transmission lines, and lock-in amplifiers, the crosstalk between the various instruments will generate non-negligible background disturbances. In particular, the functional specifications of commercially available waveform generators and transconductance amplifiers are only marginally sufficient for use. Therefore, there is an urgent need for the development of both test instruments and techniques so that we can standardize testing methods for effectively reading the charge and spin states of semiconductor qubits.

 

Quantum computing hardware technology is an interdisciplinary area of research and development that requires the integration of physics, semiconductor engineering, nanomaterial testing, analogue/digital mixed-signal integrated circuits and microwave technology, etc. in order to make its way from proof of concept and system integration to the error correction, and eventually into the practical market. As a global leader in silicon semiconductor technology and an important center of the integrated circuit (IC) manufacturing industry, Taiwan should make an effort to develop silicon and germanium qubit dot technology.

 

 

Reference:

[1] D. Loss and D. P. DiVincenzo, “Quantum computation with quantum dots,” Phys. Rev. A, vol. 57, 120–126 (1998)

[2] T. Meunier et al., “Towards scalable quantum computing based on silicon spin,” VLSI Sym. T3-2(2019)

[3] I-Hsiang Wang, Po-Yu Hong, Kang-Ping Peng, Horng-Chih Lin, Thomas George, and Pei-Wen Li, (invited talk), “The Wonderful World of Designer Ge Quantum Dots” IEDM Tech. Dig. pp. 841-845, Dec. 2020, San Francisco, USA.

[4] W. I. L. Lawrie, H. G. J. Eenink, N. W. Hendrickx, J. M. Boter, L. Petit, S. V. Amitonov, M. Lodari, B. Paquelet Wuetz, C. Volk, S. G. J. Philips, G. Droulers, N. Kalhor, F. van Riggelen, D. Brousse, A. Sammak, L. M. K. Vandersypen, G. Scappucci, and M. Veldhorst, “Quantum dot arrays in silicon and germanium,” Appl. Phys. Lett., 116, 080501 (2020)

[5] K. Horibe, T. Kodera, S. Oda, “Lithographically defined few-electron silicon quantum dots based on a silicon-on-insulator substrate,”. Appl. Phys. Lett. 2015, 106, 083111

 

 

 

 

Postscript from the MA-tek Editorial Team

This year, MA-tek is honored to be joining hands with Professor Pei-wen Li to carry out an industry-academia collaboration project focused on developing the structural and electrical technologies behind quantum dots, the key components of quantum computing. The quantum dots in nanoelectrodes in particular are extremely small and very difficult to manufacture and analyze. As such, they call for the assistance of many advanced analysis techniques. MA-tek specializes in a wide range of structural and compositional analysis technologies such as positioning technology for component-level electrical property measurement, Focused Ion Beam (FIB) precision sample preparation, high-resolution Transmission Electron Microscopy (TEM), Energy Dispersive X-ray Spectroscopy  (EDS), Electron Backscatter Diffraction (EBSD) and more. Whether it is the superconductor, ion trap, Nitrogen–Vacancy (NV) or other types of qubit structure, MA-tek can provide all the needed analyses and supports to your quantum component research.

 

Quantum technology is currently in the midst of explosive development. Quantum computers in particular have already demonstrated that they can surpass the performance of regular computers in certain fields. It has caught the attention of many industries, and numerous companies and institutions in Taiwan so that many companies from different industries have begun investing in the establishment of their own quantum research and development departments. As the industry’s best research and development partner, MA-tek is proud to stand by our customers as we strive together to help Taiwan blaze a trail into the quantum era.

 

Plans for the next issue of the “New Technology Channel  |  Collaboration Column” are currently in progress, so stay tuned for more MA-tek technical articles. Become more competitive in the global supply chain by staying up to date on the most recent news on various cutting-edge technologies!

 

Read the Introduction to Structure and Composition Analysis Techniques: