Substrate & Dielectric Loss¶
Substrate and dielectric-loss outputs explain how material selection and surface participation affect T_1.
# |
Parameter |
Symbol / Unit |
Extraction Method |
Typical Q3D Value |
Ideal / Optimal |
Good Range |
Worst Case |
Why It Matters |
Key Design Note |
|---|---|---|---|---|---|---|---|---|---|
40 |
Substrate Bulk Loss Tangent |
tan delta_bulk |
Q3D dielectric loss tangent input; resonator Q fitting vs power |
< 10⁻⁶ (HR-Si, 4K) |
< 10⁻⁶ |
10⁻⁶ – 10⁻⁵ |
> 10⁻⁴ |
Bulk dielectric loss sets floor on 1/Q_int from substrate volume; sapphire < 5×10⁻⁷ |
tan δ improves by 10–100× on cooling from 300K to 4K due to reduced phonon and TLS population |
41 |
Metal-Air Interface Loss (tan delta_MA) |
tan delta_MA |
Surface participation ratio (SPR) from Q3D E-field + measured Q factor |
~10⁻³ |
< 10⁻³ (passivated Al₂O₃) |
10⁻³ – 5×10⁻³ |
> 10⁻² |
TLS loss at metal-air interface is the dominant T₁ source in planar transmon designs |
Etching native oxide before Al deposition reduces tan delta_MA by up to 10×; HF vapor clean |
42 |
Substrate-Air Interface Loss (tan delta_SA) |
tan delta_SA |
SPR analysis from Q3D E-field distribution |
~5×10⁻⁴ |
< 5×10⁻⁴ (HF-etched Si) |
5×10⁻⁴ – 5×10⁻³ |
> 10⁻² |
TLS at substrate exposed surface; addressed by passivation, UV ozone clean, or dry etching |
Hydrogen-passivated Si surface (HF dip) shows 5× lower tan delta_SA vs untreated Si |
43 |
Metal-Substrate Interface Loss (tan delta_MS) |
tan delta_MS |
EELS/TEM interface composition + Q3D SPR calculation |
~5×10⁻³ |
< 5×10⁻³ |
5×10⁻³ – 10⁻² |
> 5×10⁻² |
TLS at Al–Si or Nb–Si interface; reduced by HF dip substrate prep before metal deposition |
Amorphous interfacial SiOx layer of 1–2 nm is the primary TLS host; substrate HF clean removes it |
44 |
Surface Participation Ratio (SPR) |
p_MA / ppm |
Q3D E-field energy integral on metal-air interface: p = ∫_MA ε|E|²dV / ∫_all ε|E|²dV |
5 – 50 ppm |
< 5 ppm |
5 – 50 ppm |
> 200 ppm |
p × tan δ contributes directly to 1/Q; minimise by thick metal, wider gap, no sharp corners. For planar transmons, p_MA can reach 100–1000 ppm without geometry optimisation. |
1/Q_TLS = Σ p_i × tan δ_i; SPR is the design lever; tan δ is the material lever. <5 ppm ideal is achievable in optimised 3D cavity or large-gap planar designs; planar CPW without optimisation may be 100–1000 ppm. |