As conventional semiconductor-based information technology is now reaching the ultimate limitation of nanoscale integration, the effort to realize devices that utilize quantum effects (e.g., quantum superposition, entanglement, etc.) is becoming an emerging research field. Among many plausible experimental implementations of quantum technology, the similarities between gated quantum dots and transistors in modern microelectronics – in fabrication methods, physical structures, and voltage scales for manipulation – have led to significant interest in the development of quantum bits (i.e., qubits) in semiconductor quantum dots. Primary efforts have been devoted to developing charge- or spin-based quantum dot qubits. Whereas spin-based quantum dot qubits have demonstrated long coherence times owing to their immunity to charge fluctuations, their manipulation is often slower than desired for important future applications, such as factoring. On the other hand, charge-based quantum dot qubits offer ultrafast electric field manipulation and simple architectures. However, they suffer from rapid dephasing because charge noise also couples strongly to this type of qubit. 

Figure 1. Schematic structure and two main types of qubits realized in semiconductor quantum dots. (A) Schematic and (B) realistic quantum dot structure [Nature 442, 766 (2006)]. (C) and (D) show a schematic representation of the logical basis and typical Rabi oscillation of a double-dot charge qubit (C) [Nature Nanotech. 10, 243 (2015)], and a single-dot spin qubit. (D) [Nature 442, 766 (2006)].

  With significant developments in material growth and nano-fabrication techniques, the field is now advancing to a point where quantum dot qubits can be realized in silicon. Silicon has weak spin-orbit coupling, and it is expected that the hyperfine interaction can be eliminated by the isotopic purification of 28Si with zero nuclear spin [Nature Nanotech. 9, 981 (2014)]. Many different types of qubits have been proposed and tested. Representatives are the single-dot spin qubit, the double-dot charge qubit, the double-dot single/triplet qubit, and the triple-dot exchange-only qubit [Nature Nanotech. 9, 666 (2014); Nature 481, 344 (2012)]. As of 2015, elementary single-qubit rotation has been experimentally demonstrated for all types. With the invention of surface code [Phys. Rev. A 86, 032304], which enables fault-tolerant quantum computing with a moderate threshold for the gate-error rate (roughly ~1%), great effort has been devoted to realizing qubit control fidelity that exceeds 99%. More recently, the three-electron double-dot spin-charge hybrid qubit has been theoretically proposed, and this new type of qubit has been experimentally demonstrated [Nature 511, 70 (2014)]. The quantum dot hybrid qubit offers a promising combination of characteristics that, in the past, have been found only separately in qubits based on either charge or spin degrees of freedom: a good ratio of manipulation time to coherence time, together with fast overall operation and the ability to fully control the qubit using a single control parameter. 


 In our labratory we actively pursue to develop integrated quantum coherent systems using semiconductor quantum dots. We mainly explore various types (charge, spin, singlet-priplet, and hybrid) of quantum dot qubits with an aim to increase control fidelity over current designs. Going beyond singlet qubit operations, experimental designs and measurements are ongoing to realize high-fidelity two qubit conditional gates and entanglement control in semiconductor quantum dot systems. 


Figure 2. Newly installed cryo-free dilution refrigerator with 5T magnet (Oxford Instruments Triton 500) equipped with bottom loading fast sample exchange mechanism.