 Hello everyone. I'm Zhen Yu Huang. This is an abstract of our paper synthesizing quantum circuits of AES with lower t-depth and less qubits. This work is joined with Su Wei Sun. Our work is motivated from quantum attacks to symmetric ciphers. To apply a quantum attack, the attacker needs an attack circuit based on quantum gates. And the quantum circuit for the encryption process is a part of the attack circuit. Our work is focused on reducing the cost of this encryption circuit, especially the cost of AES circuit. Our first contribution is a new general structure for constructing the whole encryption circuit. In previous works, all structures used out-of-place circuits for round transformations as sub-circuits. Here, out-of-place means the output we need is stored in some different qubits than the input. Due to the use of these out-of-place circuits, these structures have some redundant outputs. And the width, which is the number of qubits, will increase when the number of round increases. Here, we present a method to construct an in-place round circuit from two out-of-place sub-circuits. By this structure, the width of the circuit will not increase after each round. Here is in comparison of our new structure with our other ones. Our second contribution is some techniques for implementing linear and nonlinear transformations. For linear ones, it is well known that an invertible linear transformation can be implemented by an in-place C0 circuit. And we can achieve such C0 circuit by failure decomposition of some heuristic algorithms. But, the circuits advanced from these methods are not optimal. Here, we propose a new set-based method for finding a C0 circuit with a minimal number of qubits. The idea of our method is encoding the problem of finding a circuit with qubits into a set problem. Our experiments show that, by our method, problems with size less than 9 bits can be solved in a reasonable time. Our next problem is implementing nonlinear transformations in place. Even an out-of-place circuit of function f, we call it a C0 circuit if it works when the output wires are initialized to zero. While we call it a C-star circuit if it works when the output wires are initialized to nonzero. Then, we can classify the nonlinear transformations into two kinds. The first one is the vessel-like transformation, which can be implemented in place by a C-star circuit. The second one is the substitution-like transformation, which can be implemented in place by two C0 circuits with a structure we introduced before. Furthermore, we propose some criteria for efficiently designing C-star circuits. Under this criteria, a C-star circuit can be constructed from a special C0 circuit by adding a C0 sub-circuit. By this method, we construct a more compact C-star circuit of AES from the one proposed than the one proposed in AC 2020. Our next problem is constructing low-depth circuits for nonlinear transformations. Firstly, from an n-depth k classical circuit, one cannot always abandon a t-depth k quantum circuit by simple replacement. The AES S-box quantum circuit proposed in EuroCrip 2020 is an example. Here, we present a method for constructing a t-depth k quantum circuit from such classical circuit. As applications, we construct a t-depth 4 and a t-depth 3 quantum circuits for AES S-box. Since AES S-box has algebraic degree 7, t-depth 3 is optimal if we follow the way from classical to quantum. Based on all these techniques introduced before, we constructed some low-width circuits for AES. It is easy to see that our circuits are more compact than the one proposed in AC 2020. Furthermore, we construct some low-depth circuits for AES. Comparing to the one proposed in EuroCrip, our circuit has lower t-depth and lower full-depth. At last, we abandon the trade-off curve shown in this speaker. That's all. Thank you for your attention.