PAM-XIAMEN is able to provide you with silicon wafers for optical studies, more please refer to https://www.powerwaywafer.com/silicon-wafer.

The global quantum internet needs a long-life, telecom band photonic material interface that can be manufactured on a large scale. The preliminary quantum network based on photon matter interfaces that meet these subsets of needs is encouraging efforts to find new high-performance alternatives. Silicon is the ideal subject for commercial scale solid-state quantum technology. It is already an advanced platform in the global integrated photonics and microelectronics industry, and has a record breaking long-life spin quantum bit. Although silicon quantum platforms have great potential, optical detection of the photon spin interface on silicon-based materials remains elusive.

1. Integration and Optical Coupling of T-Centers in Silicon

The T-center is the radiation damage center in silicon, consisting of two carbon atoms, one hydrogen atom, and one unpaired electron (Fig. 1a). At 935.1 meV (1326 nm), there is a zero phonon line (ZPL) optical transition, and the T-center is one of the known silicon radiation damage centers that emit light in the near-infrared communication band.

Measurement of the T center ensemble in isotopic enriched 28Si revealed an excited state lifetime of 940 ns and a transition linewidth as low as 33 MHz. The T-center ground state has non electron spin and hyperfine coupled hydrogen nuclear spin. The ground state electrons and hydrogen nuclear spin are both long-lived, with coherence times greater than 2.1 ms and 1.1 s in 28Si, respectively. In the optical excited state of the bound exciton, two electrons form a singlet state, and the decrease in defect symmetry splits the hole state into two spin double states, labeled TX0 and TX1, respectively (Fig. 1b). Under the action of a static magnetic field, TX0 ZPL splits into four spin related transitions.

This work first prepares T-centers in industrial standard SOI chips. As shown in Fig. 1c, T-center photoluminescence (PL) dominates the spectrum of the sample. To achieve spatial resolution of a single T-center, experiments were conducted in a self-made SOI device layer 4.3(3)K low-temperature confocal microscope. Fig. 1d shows the simulated field of the dipole emitter at the center of the microback, oriented in the device plane. This work estimates that in unmodeled SOI, the strength of ZPL is increased by up to 58 times compared to the center.

Fig. 1 Integration and optical coupling of T-center

2. Silicon Based Center

The confocal PLE reveals evidence that a single T-center can be addressed. This study selected a group of micropucks with a radius of 305 nm (Fig. 2a), and then measured the PLE spectrum for each micropucks within the 776μeV range around the bulk TX0 ZPL. Three examples of single push PLE spectra are shown in Fig. 2b. Each PLE spectrum contains a small amount (averaging 1.1) of narrow resonances sampled from larger non-uniform distributions.

Fig. 2c shows the position and linewidth of the T-center ZPL peak in the 144 microcracks in Fig. 2a. The distribution of ZPL peaks is caused by changes in the local isotopes and strain environment of each defect. This uneven distribution has a wider range and is slightly offset from the non graphical SOI ZPL (as shown Fig. 1c).

This work found that in severely damaged and unoptimized materials, the total spectral diffusion of the selection center is less than 400 MHz. Surface optimization, electrostatic control, and lower injection damage have all been proven to significantly reduce environmental noise and spectral diffusion of other color centers, and similar techniques can also be applied to this system.

Fig. 2 Silicon based center

3. Silicon Based Single Spin Optical Initialization and Readout

In Fig. 3a-c, this work presents the dual color PLE spectra of three T centers extracted from microbubbles with a radius of 305nm. The ZPL splitting of each TX0 is different, reflecting different orientations. T-center 1 has almost degenerate B and C transitions, with each laser independently driving continuous fluorescence.

In contrast, the B and C transitions of T-center 2 were well resolved under the splitting of 1 GHz. The brightest fluorescence is generated by a combination of two colors, in which the laser is detuned and resonates with the B and C transitions, respectively. The B-C splitting of T-center 3 is only 0.7 GHz, but the two A-D resonances are still well resolved.

After confirming that these are single spins, the next step of this work is to perform optical initialization and reading of the spin state, and measure the spin lifetime (T1). The optical pulse sequence shown in Fig. 3f processed the B and C transitions of T-center 3. When the waiting time approaches 1 ms, asymmetric transient lifetimes of 0.85 (6) and 1.2 (1) ms are generated for B (orange) and C (blue) reading pulses, respectively. In future work, more optical extinction will extend the measurable spin lifetime.

Fig. 3 Single spin optical initialization and readout

This work integrates individually addressable T-center photon spin qubits into silicon photonic structures and characterizes their spin related optical transitions in the telecommunications band. A device integrated T-center with sub 400 MHz long-term optical linewidth below 2.5 K was measured. Through process development, electrostatic engineering, and dynamic control, the long-term optical linewidth of many emitters was improved. This research achievement provides a direct opportunity for constructing a silicon integrated telecommunications band quantum information network.

For more information, please contact us email at victorchan@powerwaywafer.com and powerwaymaterial@gmail.com.