Flux-tunable qubit chip (chip_05)

A 7.5 mm die with seven Xmon qubits and eight quarter-wave resonators on one feedline. Two of the qubits (Q4 and Q5) have a SQUID instead of a single junction and a flux line from a small pad at the top or bottom edge, so their frequency can be tuned with a current. The chip also has flux-trap hole arrays, two HSQ junction test pads and HSQ corner markers. It is Quantcore’s own chip from the QWri legacy script chip_05_flux_TLQubits.py (May 2024), the one that tests flux tuning.

The rebuild draws it from Quantcore PDK parts: the flux lines are routed with the PDK router and the resonator tails run into the claws. The PDK cheese (5 µm squares with 20 µm gaps, at least 4 µm from the 3/0 keep-outs) also covers the whole die, and the holes drawn in the original layout stay. Build this one to see flux lines and SQUID qubits next to a readout feedline.

chip_05_flux_tlqubits as rebuilt on the Quantcore PDK

chip_05_flux_tlqubits as rebuilt on the Quantcore PDK

At a glance

Notebook

examples/designs/legacy/chip_05_flux_tlqubits.ipynb

Source script

components/python/chip_05_flux_TLQubits.py (QWri legacy, Quantcore internal)

Chip

7.5 × 7.5 mm die, 10/6 µm CPW

On the chip

7 Xmons (Q4, Q5 SQUIDs) · 8 λ/4 resonators · 2 flux lines (4/6 µm) · 8 hole arrays · 2 HSQ test pads · 4 HSQ corner markers

Builds today

yes, rebuilt from its JSON on 2026-11-19 in 2.5 s (the notebook itself was not re-run)

Pictures

chip_05_flux_tlqubits: the original legacy layout (render of the legacy GDS)

chip_05_flux_tlqubits: the original legacy layout (render of the legacy GDS)

chip_05_flux_tlqubits: the Quantcore rebuild

chip_05_flux_tlqubits: the Quantcore rebuild

Q4, a qubit with a flux line: the line arrives at the top, the resonator tail at the claw

Q4, a qubit with a flux line: the line arrives at the top, the resonator tail at the claw

Q1, a single-junction qubit, for comparison

Q1, a single-junction qubit, for comparison

An HSQ junction test pad

An HSQ junction test pad

LOM estimates

The notebook’s saved outputs hold no LOM run (its LOM cells were not executed), so there are no estimates to show here and none were computed for this page.

Build spec

The rest of this page is the detailed spec the notebook was built from: components, positions, parameters and what changed from the legacy script.

Source: components/python/chip_05_flux_TLQubits.py (create_test_chip(), May 2024; writes chip_05_flux_TLQubits_1.0.gds). Legacy QWri library, Quantcore internal. Notebook examples/designs/legacy/chip_05_flux_tlqubits.ipynb.

Changes made in the rebuild

  • Junction indexing fixed: bot[i] (Q1–Q5 js 5, Q6 js 8); the junction-layer reference uses the same sizes.

  • Flux-line corners: the largest radius that fits each line. FLUX4 r = 75 µm (its 155 µm pad lead cannot hold 100), FLUX5 the legacy 100 µm; no break or overlap (checked in a render).

  • R8: the PCell’s own open end (ground=True, as chip_10a does) instead of the resonatorGnd box: CUT8 and W_R8 removed.

  • Device ID: CHIP05 kept, placed by the frame rule. claw_tail_gap = 10 µm (legacy cap_spacing).

  • Junctions: legacy HSQ cells at the legacy fingers on the PDK junction rule (Nb trilayer, Δ = 1.4 meV, I_c·R_n = πΔ/2e = 2.20 mV, R_n 40–50 kΩ, qubits 4–8 GHz; PDK process get_pdk().junction: R_n·A 13.5 kΩ·µm², C_J 50 fF/µm², both assumed by the PDK): Q1–Q3 45 kΩ, Q6 28 kΩ, SQUIDs 22.5 kΩ parallel. LOM: f_q 5.8–6.1 GHz (SQUIDs 7.8 at zero flux); Q1/R1 and Q2/R2 hybridised.

  • The notebook’s JSON was rebuilt from its layout cell. The notebook has not been re-run, so its outputs are cleared and its PNGs may be out of date.

What it is: a 7.5 × 7.5 mm die centred on (0, 0) → Designer (x + 3750, y + 3750); PDK v2 die frame replaces the base-metal corner Markers and the layer-10 outline; the HSQ corner markers are kept. Feedline + two rounded pads (as chip_03), eight Coupled Resonators (ground=False: tails without an end cap, joining the claws), seven Xmons (Q4, Q5 SQUIDs with flux lines from two small pads at the top/bottom edge centres), flux-trap holes, two HSQ test pads, labels “500nm”, “800nm”, “QCG_Nb05_W”.

Builds: yes. Not reproduced: labels, the legacy flux-line end geometry (thin 1 µm section + ground boxes). Junction windows (CLEARER) as Xmon.junction_window; junctions from junction_creator_HSQ / squid_creator_HSQ (the scripts the source imports) as TrilayerJunctionHSQ / TrilayerSquidHSQ, XOR 0 on the junction layers.

Components (legacy frame, µm)

instance

legacy → catalog ref

params

at

rot / mirror

FL

Coplanar Waveguide → CpwSegment

5460 → 5360 (pads snapped, same etch), a/b 10/6

(0, 0)

0

PL, PR

Connector Pad → ConnectorPad

600 × 200, taper 500, 10/6, round, sim False

(∓2680, 0)

270° / 90°

PFT, PFB

Connector Pad (cpadB / cpadT) → ConnectorPad

400 × 150, taper 100, trace 4/6, round

(0, ±3155)

180°; PFB mirrored

R1 … R8

Coupled Resonator → CoupledResonator

common gap 10, clength 200, cgap 7, width 300, neck 400, ground False; cpw_len 4740 / 4920 / 5590 / 6200 / 5200 / 4250 / 5840 / 4500; flip R2, R3, R7, R8; clength 400 (R3, R7, R8), 250 (R5); width 400 (R4)

x = −2000, −1430, −850, −280, 280, 850, 1430, 2000; y = −38, 33, −33, 33, −38, 33, −38, 33

mirror True at +33

Q1 … Q7

Xmon Cross → Xmon

as chip_03 (width 400; Q6 600)

(−1950, −2218), (−1480, 2391), (−800, −2237), (−180, 2492), (305, −2219), (900, 2203), (1480, −2503)

0° / 180° alternating

junctions

junction_creator_HSQ.create_junction(bot[i−1], top[i−1]) / squid_creator_HSQ.create_squid (Q4, Q5) → junction_type="Trilayer Junction HSQ" / "Trilayer Squid HSQ", junction_params={"finger_width": 0.1·bot, "second_finger_width": 0.6}

as built (bot[i−1] indexing bug): Q1 0.8 µm, Q2–Q6 0.5 µm straight; 0.6 µm second finger; Q7 none

(x, y ± 210) (Q6 ± 310)

Trans(0, flag), flag toggling per qubit: Q2, Q4, Q6 mirrored about x → mirror_squid on these 180° Xmons (junction-layer XOR 0)

CLEARER

30 × 27 µm boxes cut from 1/0 and 3/0 at (x, y ± 207) (Q6 ± 307): from 6.5 µm inside the arm to 0.5 µm into the ground

Xmon(junction_window=[30, 27, -3]) (metal on base_metal_addition, removed from M1 at export; dropped in the FEM export)

FLUX4, FLUX5

raw path polygons (4 µm centre, 6 µm gaps, corners rounded at 100 µm) + 50 µm thin tip (1 µm) + ground boxes

Designer routes PFT.a → Q4.flux, PFB.a → Q5.flux (nodes (0, 3000), (−175, 3000) / (0, −2795), (300, −2795); a/b 4/6, r 50); Xmon flux_port, flux_x −5, flux_tip_depth 7.5 → tip at (−175, 2719.5) / (300, −2446.5) as drawn

H0 … H7

flux_holes 30 µm squares, 3 × 15 at 100 µm

FluxHoleArray(30, 100, 100, 3, 15)

(x − 50, y): x = −2050, −1430, −850, −280, 360, 850, 1430, 2000; y = 300 / −1800 alternating

0

CUT8

resonatorGnd box (2035, 1660)–(2055, 1670): etches R8’s centre conductor over its last 10 µm (the tail, uncapped, ends at y = 1669.9)

CpweOpenEnd(length=10, a=10, b=5) + d.couple("W_R8", "R8.end", "CUT8.a", wire=True)

(2045, 1660)

90°

TP1, TP2

testPad_Nb_TL_HSQ.create_testPad(bf / 0.2, 0.6) (inserted 6× by an indentation bug: identical copies)

JunctionTestPads(**legacy_test_pad_params("testPad_Nb_TL_HSQ", bf / 0.2, 0.6)), bf 0.5 / 0.8

(−2850, 2000), (2850, −2000)

0

HSQ0 … HSQ3

Marker on 500/0 (HSQ), width 150

CornerMarker(hsq=True, thickness=50, width=150)

corners

1/2/3/0

Resonator tail ↔ claw

Tails without end caps run 0.9 / 2.9 / 17.9 / 22.9 / 0.9 / 29.9 / 6.9 µm into the claw tails of Q1 … Q7: the strip and the claw are one conductor (checked on the rebuilt metal). LOM: d.couple("W_Qi", "Qi.claw_n", "Ri.end", wire=True).

LOM node structure

FL.line (50 Ω at each pad) ─ HANGER (K_Ri) ─ Ri.res (λ/4, shorted at the coupler) = Qi.claw_n ─ C_g ─ Qi island (C, JJ or SQUID to ground; Q7 no junction). FLUX4/FLUX5 → control:flux of Q4/Q5. R8: the uncapped tail would run into the ground (λ/2), but CUT8 opens it, so R8 is an open λ/4 (W_R8 → the CUT8 termination): 7.085 GHz bare, 7.048 GHz on the feedline (target 7.0). A simulation selection of R8 must include CUT8.

Targets

Source comments: R1 6.5, R2 6.25, R3 5.5, R4 5, R5 6, R6 7.2, R7 5.25, R8 7 GHz.

Keep-outs / frame

Pads sit at all four edge centres, so the frame’s device ID (CHIP05) slides along the bottom edge to the first free spot, (4600, 100)–(5300, 400), clear of PFB. HSQ markers in the corner marker areas (legacy positions kept).