Capacitance Matrix (C)¶
Capacitance outputs are the main Q3D bridge into transmon energy, coupling, detuning, and readout design.
# |
Parameter |
Symbol / Unit |
Extraction Method |
Typical Q3D Value |
Ideal / Optimal |
Good Range |
Worst Case |
Why It Matters |
Key Design Note |
|---|---|---|---|---|---|---|---|---|---|
15 |
Qubit Self-Capacitance (C_Sigma) |
C_Sigma / fF |
Q3D electrostatic solve; Maxwell capacitance matrix |
60 – 100 fF |
60 – 100 fF (transmon shunting) |
40 – 200 fF |
< 10 fF or > 500 fF |
Sets charging energy Ec = e²/2C_Σ; large C_Sigma → transmon regime → exponentially reduced charge noise |
C_Sigma = Σ|C_ij| from Maxwell matrix; target Ej/Ec = 50–80 for optimal transmon performance |
16 |
Readout Resonator Capacitance (C_r) |
C_r / fF |
Q3D capacitance matrix + HFSS eigenmode simulation |
200 – 500 fF |
200 – 500 fF (λ/4 CPW) |
100 – 600 fF |
< 50 or > 1 pF |
Resonator mode capacitance sets frequency: ω_r = 1/√(L_r C_r); must target 6.5–8 GHz window |
Combined with Q_ext sets readout bandwidth κ = ω_r/Q_ext; trade-off between speed and SNR |
17 |
Qubit–Resonator Coupling Cap (C_g) |
C_g / fF |
Q3D Maxwell matrix off-diagonal C_12 between qubit island and resonator |
1 – 10 fF |
1 – 10 fF (dispersive limit) |
0.5 – 15 fF |
< 0.1 or > 50 fF |
Sets coupling g = C_g/(2C_Σ)·√(ω_q ω_r/L_r C_r); must stay dispersive (g ≪ qubit–resonator detuning) |
g / 2pi target 50–150 MHz; too large → strong coupling regime; Purcell decay ∝ (g/Δ)² × κ |
18 |
Qubit–Qubit Coupling Cap (C_J) |
C_J / fF |
Q3D full capacitance matrix between qubit islands |
0.5 – 5 fF |
0.5 – 5 fF (tunable coupler) |
0.2 – 10 fF |
< 0.05 or > 30 fF |
Drives direct transverse coupling J; residual C_J causes always-on ZZ unless tunable coupler used |
Modern heavy-hex lattice uses tunable couplers to cancel residual ZZ to < 10 kHz |
19 |
Pad-to-Ground Parasitic Cap |
C_pad / fF |
Q3D with ground plane mesh |
< 5 fF per pad |
< 5 fF (small footprint) |
1 – 20 fF |
> 50 fF |
Unintended pad-to-ground capacitance shifts qubit frequency from design target |
Each 1 fF of parasitic shifts f_qubit by ~10–30 MHz; critical to include in Hamiltonian model |
20 |
Trace Mutual Capacitance (C_ij) |
C_ij / fF |
Q3D electrostatic solve off-diagonal extraction |
< 1 fF (separated lines) |
< 1 fF |
1 – 5 fF |
> 20 fF |
Capacitive coupling between control lines causes microwave crosstalk in drive and readout paths |
Overlapping traces on adjacent layers is the primary source; add ground shield layer to suppress |