Coulomb and spin blockade of two few-electron quantum dots in series in the cotunneling regime

We present Coulomb blockade measurements of two few-electron quantum dots in series configured such that the electrochemical potential of one of the two dots was aligned with spin-selective leads. Charge transfer through the system required cotunneling through the second dot (which was not in resonance with the leads). The amplitude modulation of the current was found to reflect spin blockade events occurring at either of the two dots. We also confirm that charge redistribution events that occurred in the off-resonance dot could be detected indirectly via changes in the electrochemical potential of the aligned dot.

Figure 1

(a) Scanning electron micrograph of the experimental device; (b) inverted gray scale of conductance through the left dot for four CB peaks in the vicinity of $\nu =2$. Dark (light) regions represent high (low) amplitude of the CB peak; and (c) typical modulation of the position and amplitude of a CB peak at the $\nu =2$ transition, shown for the case of the $12\leftrightarrow 13$ transition.

Figure 2

(a) The measured charging diagram of the DD. Inset shows how the amplitude of the CB peaks rapidly decreases away from the triple point (marked by an arrow). (b) Schematic representation of different energy configurations of a DD system close to and away from the triple points. Pair of numbers $(n_L,n_R)$ describe the occupation number for the left and right dots in each configuration. Open (closed) circles indicate empty (occupied) levels in the dot. (c) Inverted conductance gray scale showing the magnetic field dependence of four CB peaks, measured along the solid line marked in (a). Solid squares (triangles) mark peaks observed when only the left (right) dot is aligned with the leads. Arrows mark the $\nu =2$ transition as detected for each configuration.

Figure 3

Schematic representation of the configuration involved in the transfer of an electron through the DD system for the situation in which only the left dot is in resonance with the leads. Dashed arrows indicate cotunneling events through the virtual $(n_L+1,n_R-1)$ configuration.

Figure 4

Comparison of the DD data for the $(13,16)\leftrightarrow (14,16)$ peak with the SD data for the $L13\leftrightarrow 14$ and the $R15\leftrightarrow 16$ transitions on the left and right dots, respectively, in the vicinity of the $\nu =2$ transition. (a) Respective inverted gray scales of the DD (middle panel), the left dot (top), and the right dot (bottom) data. As a reference the peak related to the $(13,15)\leftrightarrow (13,16)$ transition is also shown. (b) Extracted amplitude of the respective CB peaks. Arrows and dashed lines mark transitions in the ground states of the respective dots, labeled by letters for reference.

Figure 5

(a) Peak positions of the data from Fig. 4a. Curves are offset vertically for clarity. The thick arrow indicates the direction for more negative plunger voltages. GS transitions on single dots are marked as in Fig. 4. Black triangles (diamond) mark features related to charge redistribution events occurring in the right (left) off-resonance dot and detected by the electrochemical potential of the in-resonance left (right) dot. (b) Schematic representations of the GS transitions that constitute charge redistributions. The horizontal dashed line marks the Fermi energy of the leads $\left(E_F\right)$. The vertical dotted line indicates the edge of both dots. Horizontal black (gray) bars indicate edge (center) orbitals of the 1LL (2LL).