Supplimental Information for Flat-Band Enhanced Antiferromagnetic Fluctuations and Unconventional Superconductivity in Pressurized CsCr${}_{3}$Sb${}_{5}$

I Details of First-Principles Calculations

Our first-principles calculations are performed with density functional theory (DFT), as implemented in the Vienna Ab initio Simulation Package (VASP) [1]. The Kohn-Sham wave functions are represented under projected augmented wave (PAW) basis [2]. The crystal structure of CsCr${}_{3}$Sb${}_{5}$ is optimized at external pressure of 5 GPa, with a solid revised Perdew-Burke-Ernzerhof (PBEsol) exchange correlation functional [3]. For the self-consistent and density of state (DOS) calculations, we employed a 12$\times$12$\times$8 and a 24$\times$24$\times$12 $\Gamma$-centered $k$ mesh respectively. Our tight-binding Hamiltonian is then constructed via Wannier downfolding [4], where 30 atomic like projectors of 15 Cr-$d$ and 15 Sb-$p$ are adopted as initial seeds. The local axes are set that the three Cr sites are connected by successive $C_{3}$ rotations, as shown in Fig. 1(b).

II Band Structure Downfolding

In CsCr${}_{3}$Sb${}_{5}$, the Kagome CrSb plane serves as a symmetric mirror plane (denoted as $M_{z}$). This $M_{z}$ symmetry is preserved in the $k_{z}=0$ plane in reciprocal space. Therefore, we can divide the Wannier orbitals into $M_{z}$-even and $M_{z}$-odd sets and downfold the in-plane band structure into each subspace. The $M_{z}$-even orbital set consists of Cr-$d_{z^{2}}/d_{x^{2}-y^{2}}/d_{xy}$, Sb1-$p_{x}/p_{y}$, and $M_{z}$-even combinations of Sb2-$p$ orbitals; while the $M_{z}$-odd set consists of Cr-$d_{xz}/d_{yz}$, Sb1-$p_{z}$, and $M_{z}$-odd combinations of Sb2-$p$ orbitals.

Figure S1: (a) Comparison between original band structure and downfolded band structures for CsCr${}_{3}$Sb${}_{5}$. (b) Downfolded band structure for $M_{z}$-even orbital set. (c) Downfolded band structure for $M_{z}$-odd orbital set. The flat band, van Hove singularities (vHSs), and Dirac points are well distinguishable for each set of band structures.

Fig. S1 shows the downfolding result for both $M_{z}$-even and $M_{z}$-odd orbital sets. Clearly, the $k_{z}=0$ in-plane band structure could be well-decomposed. Notably, there is an extremely flat band at $\Gamma$-M line in Fig. S1(c). The eigenvector of this flat band is composed of the equal-weight superposition of Cr-$d_{xz}$ and $d_{yz}$ orbitals, which is quite similar to the $\Gamma$-M-L-A flat band in FeGe [5].

III Details for RPA Calculations

Our calculations begin from the model Hamiltonian that:

$$H=H_{0}+\sum_{\tau}\left[U\sum_{i}n_{\tau,i\uparrow}n_{\tau,i\downarrow}+U^{\prime}\sum_{i<j}n_{\tau,i}n_{\tau,j}+J\sum_{i<j,\sigma,\sigma^{\prime}}c_{\tau,i\sigma}^{\dagger}c_{\tau,j\sigma^{\prime}}^{\dagger}c_{\tau,i\sigma^{\prime}}c_{\tau,j\sigma}+J^{\prime}\sum_{i\neq j}c_{\tau,i\uparrow}^{\dagger}c_{\tau,i\downarrow}^{\dagger}c_{\tau,j\downarrow}c_{\tau,j\uparrow}\right],$$ (S1)

where $H_{0}$ is the bare-election tight-binding Hamiltonian obtained from Wannier downfolding; $\tau$ is the Cr site index, $i$, $j$ are $d$ orbital indices for each Cr. The Coulomb parameters follow the general symmetry relation that $J=J^{\prime}$, $U=U^{\prime}+2J$.

The bare electron susceptibility is calculated using:

$${\chi_{0}}^{pq}_{st}(\mathbf{q},\mathrm{i}\nu_{m})=-\frac{1}{N_{\mathbf{k}}\beta}\sum_{n,\mathbf{k}}G_{sp}(\mathrm{i}\omega_{n},\mathbf{k})G_{qt}(\mathrm{i}\omega_{n}+\mathrm{i}\nu_{m},\mathbf{k+q}),$$ (S2)

where $\mathrm{i}\omega_{n}$ and $\mathrm{i}\nu_{m}$ are fermionic and bosonic Matsubara frequencies respectively; $p$, $q$, $s$, $t$ are orbital indices; $G_{sp}(\mathrm{i}\omega_{n},\mathbf{k})=\sum_{\mu}\frac{\braket{s}{\mathbf{k},\mu}\braket{\mathbf{k},\mu}{p}}{\mathrm{i}\omega_{n}-E_{\mathbf{k},\mu}}$ is the bare electron propagator. We have adopted a 100$\times$100$\times$30 $Q$ mesh and $\frac{1}{\beta}=$ 0.001 eV for the calculations.

For spin and charge susceptibilities, we employ matrix random-phase approximation (RPA) formula that:

$$\chi_{\mathrm{S}}=\chi_{0}\left(1-U_{\mathrm{S}}\chi_{0}\right)^{-1},$$ (S3a)

$$\chi_{\mathrm{C}}=\chi_{0}\left(1+U_{\mathrm{C}}\chi_{0}\right)^{-1}.$$ (S3b)

The matrix elements for spin and charge vertices are:

$${U_{\mathrm{S}}}^{pq}_{st}=\begin{cases}U,&p=q=s=t,\\ J,&p=q\neq s=t,\\ J,&p=t\neq q=s,\\ U^{\prime},&p=s\neq q=t,\end{cases}\quad{U_{\mathrm{C}}}^{pq}_{st}=\begin{cases}U,&p=q=s=t,\\ 2U^{\prime}-J,&p=q\neq s=t,\\ J,&p=t\neq q=s,\\ 2J-U^{\prime},&p=s\neq q=t.\end{cases}$$ (S4)

The superconducting (SC) pairing vertex functions for spin-singlet and spin-triplet channels are then calculated by:

$$V^{\mathrm{singlet}}(\mathbf{k}^{\prime}q,-\mathbf{k}^{\prime}s;\mathbf{k}p,-\mathbf{k}t)=\frac{1}{4}(3U_{\mathrm{S}}+U_{\mathrm{C}})^{pq}_{st}+\Bigl{\{}\frac{1}{4}{\left[3U_{\mathrm{S}}\chi_{\mathrm{S}}(\mathbf{k}^{\prime}-\mathbf{k})U_{\mathrm{S}}-U_{\mathrm{C}}\chi_{\mathrm{C}}(\mathbf{k}^{\prime}-\mathbf{k})U_{\mathrm{C}}\right]}^{pq}_{st}+(\mathbf{k}p\leftrightarrow-\mathbf{k}t)\Bigr{\}},$$ (S5a)

$$V^{\mathrm{triplet}}(\mathbf{k}^{\prime}q,-\mathbf{k}^{\prime}s;\mathbf{k}p,-\mathbf{k}t)=\frac{1}{4}(-U_{\mathrm{S}}+U_{\mathrm{C}})^{pq}_{st}-\Bigl{\{}\frac{1}{4}{\left[U_{\mathrm{S}}\chi_{\mathrm{S}}(\mathbf{k}^{\prime}-\mathbf{k})U_{\mathrm{S}}+U_{\mathrm{C}}\chi_{\mathrm{C}}(\mathbf{k}^{\prime}-\mathbf{k})U_{\mathrm{C}}\right]}^{pq}_{st}-(\mathbf{k}p\leftrightarrow-\mathbf{k}t)\Bigr{\}}.$$ (S5b)

Here we only take the zero-frequency component of $\chi$. Note that in the SC pairing vertex calculations, only $pq$, $st$ $\in$ nonzero $U$ matrix indices contribute to the result. Thus we can restrict our calculations to the matrix block of these indices. Here we denote the matrix block by $\chi^{dd}$. It is straightforward to see that we can simplify the calculations by replacing all $\chi_{0},\chi_{\mathrm{S}}$, and $\chi_{\mathrm{C}}$ matrices with only $dd$ block in (S2)-(S5b) without any loss of the SC pairing vertex Information.

The weak-coupling SC gap equation is then constructed as [6]:

$$\lambda\Delta(\mathbf{k}^{\prime})=-\frac{1}{V_{\mathrm{BZ}}}\int_{FS}\frac{\mathrm{d}^{2}k_{\parallel}}{|v_{\mathbf{k}}^{\perp}|}V(\mathbf{k}^{\prime},\mathbf{k})\Delta(\mathbf{k}),$$ (S6)

where the integration is done over the Fermi surface (FS), $v_{\mathbf{k}}^{\perp}$ corresponds to the Fermi velocity of the FS $\mathbf{k}$ points.