Controlled Dephasing of a Quantum Dot: From Coherent to Sequential Tunneling Academic Article uri icon


  • Resonant tunneling through two identical potential barriers renders them transparent, as particle trajectories interfere coherently. Here we realize resonant tunneling in a quantum dot (QD), and show that detection of electron trajectories renders the dot nearly insulating. Measurements were made in the integer quantum Hall regime, with the tunneling electrons in an inner edge channel coupled to detector electrons in a neighboring outer channel, which was partitioned. Quantitative analysis indicates that just a few detector electrons completely dephase the QD. The study of entanglement began in 1935 with the EPR and Schrodinger Cat paradoxes, but it languished until Bell's celebrated theorem and even thereafter. More re- cently, applications of entanglement to cryptography, ''tel- eportation,'' data compression, and computation (1) have given new impetus to the study of entanglement. Also ''dephasing'' (''decoherence'') is studied, both as a condi- tion for classical behavior to emerge and as an obstacle to applications of entanglement. Here we report controlled partial and full dephasing of electron interference in a mesoscopic Fabry-Perot type of interferometer —a quan- tum dot (QD)—entangled efficiently to a mesoscopic detector. Mesoscopic interferometers (2) include closed and open two-path interferometers, QDs and double-QDs, and elec- tronic Mach-Zehnder interferometers. Mesoscopic detec- tors (2) include quantum point contacts (QPCs) and partitioned currents. In our experiment, a QD serves as an interferometer of the Fabry-Perot type; the interference shows up as a resonant transmission peak in electron con- ductance through the dot. Figure 1 shows the device, including the QD. The dark gray strips represent charged gates. Rectangles along the edge of the device represent Ohmic contacts; tunneling electrons enter the device at the Ohmic contact marked VS. We worked at filling factors � � 2 and � � 3 in the integer quantum Hall effect re- gime, but nothing in our results depends essentially on edge channels or a magnetic field. In Fig. 1 we have � � 2, so electrons enter in two edge channels. The magnetic field turns electrons to their right but the gates repel them. As electrons from VS arrive at QPC 1, those in the outer quantum Hall edge channel (closest to the gates) make a right turn through QPC 1, while those in the innermost channel reflect towards the QD. Thus all electrons tunnel- ing through the QD come from the inner edge channel. For these electrons, the dot is an interferometer. (In general, electrons can ''relax'' from one channel to the other; such relaxation is measurable at QPC 4. We chose settings to eliminate relaxation.) To couple tunneling and ''detector'' electrons strongly, we chose them from neighboring edge channels, in close proximity. As electrons in the inner edge channel tunnel through the dot, they become entangled with electrons passing freely through the neighboring, outer (detector) edge channel. The source of these detector electrons is the Ohmic contact marked VD, and only electrons in the outer edge channel pass through QPC 1 on their way to the QD. These detector electrons couple Coulombically to the total charge Qtun tunneling through the dot, and their accumulated phase is proportional (via this Coulomb cou- pling) to the dwell time tdwell of the tunneling electrons: Q tunt dwell I tun , where I tun is the tunneling current.

publication date

  • January 1, 2007