How a drive field rotates a qubit, and the rotating-wave approximation.
Shine a resonant field on a two-level system and its population sloshes back and forth at a rate set by the field's strength, not its frequency. Making that precise takes a change of frame and an approximation. This lesson examines both: when the rotating-wave approximation holds, what it misses, and how it becomes the pulse toolkit of quantum computers.
What you’ll learn
Derive the rotating-frame Hamiltonian H~=2ℏ(ΔZ+ΩcosϕX+ΩsinϕY) and explain how the drive phase selects the rotation axis.
Solve the Rabi problem within the RWA and derive the quadratic error scaling of π pulses under detuning and amplitude errors.
Derive the Bloch–Siegert shift Ω2/(4ω0) as the leading correction to the RWA.
Analyze Ramsey interferometry and its use for frequency calibration.
The rotating frame
DerivationDriven qubit in the frame of the drive
H(t)=−2ℏω0Z+ℏΩcos(ωt−ϕ)X
Qubit with ground state ∣0⟩, driven at frequency ω with phase ϕ and strength Ω.
Product-to-sum formulas split the drive into a static co-rotating part and a counter-rotating part at 2ω. Here Δ=ω−ω0.
H~RWA=2ℏ(ΔZ+ΩcosϕX+ΩsinϕY)
Rotating-wave approximation: drop the counter-rotating part. Amplitude sets the rotation rate and phase sets the axis in the equatorial plane. That is how Rx and Ry gates are made, and why a Z rotation can be done virtually, by shifting the phase of all later pulses (McKay et al., 2017).
The Rabi solution and its error budget
Averaged over time, P1=21Ω2/(Ω2+Δ2): a Lorentzian in Δ with full width 2Ω. Stronger driving broadens the line (power broadening). For gates the more useful question is how sensitive a calibrated pulse is to small errors.
DerivationErrors of a π pulse are quadratic
P1=1+δ21sin2(2π1+δ2),δ=ΩΔ,t=Ωπ
A π pulse calibrated on resonance, applied with a small detuning.
sin2(2π1+δ2)=cos2(2π(1+δ2−1))=1−O(δ4)
The over-rotation caused by W>Ω is only fourth order.
1−P1=δ2+O(δ4)
The tilt of the axis dominates. Δ=0.1Ω leaves about 1% in ∣0⟩.
Ω→(1+ε)Ω:1−P1=sin2(2πε)≈(2πε)2
An amplitude (or duration) error ε also enters quadratically. Composite pulses cancel these terms to higher order.
Beyond the RWA: the Bloch–Siegert shift
The counter-rotating term has strength Ω/2 and oscillates at 2ω. To first order it averages to zero; its leading effect is second order, and it is a frequency shift.
DerivationLeading high-frequency correction
H~(t)=H~RWA+he−2iωt+h†e2iωt,h=2ℏΩσ+,σ+=∣0⟩⟨1∣
Write the counter-rotating part (ϕ=0) with σ+=21(X+iY).
H~eff=H~RWA+2ℏω[h†,h]+…
Leading order of the high-frequency (Floquet–Magnus) expansion: for H0+he−iνt+h†eiνt with ν large compared with ∥H0∥/ℏ and ∥h∥/ℏ, the period-averaged dynamics is generated by H0+[h†,h]/(ℏν), with corrections of higher order in 1/ν.
[σ−,σ+]=−Z⟹H~eff=2ℏ(Δ−4ωΩ2)Z+2ℏΩX
Here [h†,h]=4ℏ2Ω2[σ−,σ+]=−4ℏ2Ω2Z.
ωres≈ω0+4ω0Ω2(Bloch–Siegert shift)
Resonance now needs Δ=Ω2/(4ω), so the line moves up by δBS≈Ω2/(4ω0) (Bloch & Siegert, 1940). Physically: virtual counter-rotating processes, detuned by about 2ω0, push the two levels apart.
Check your understanding
A qubit at ω0/2π=5 GHz is driven with Ω/2π=50 MHz. Roughly how large is the Bloch–Siegert shift δBS/2π?
500 kHz
12.5 MHz
125 kHz
785 kHz
Show the answer
C. 125 kHz
In frequency units (50 MHz)2/(4×5000 MHz)=0.125 MHz =125 kHz: tiny compared with Ω, but larger than the linewidth of a good qubit.
Ramsey interferometry
Two π/2 pulses separated by free evolution make an interferometer on the Bloch sphere. Between pulses the drive is off, and in the rotating frame a detuned qubit precesses about z at rate Δ.
DerivationRamsey fringes
∣0⟩Rx(π/2)21(∣0⟩−i∣1⟩),r=−y^
First pulse, assumed short compared with 1/∣Δ∣.
r(τ)=(sinΔτ,−cosΔτ,0)
Free evolution under H~=2ℏΔZ: precession by Δτ about z.
Rx(2π):(x,y,z)↦(x,−z,y)⟹r=(sinΔτ,0,−cosΔτ)
Second pulse.
P1(τ)=21(1−rz)=cos22Δτ
The fringes oscillate at the detuning; at Δ=0 the two pulses add to a π pulse. With dephasing, the fringe contrast decays on the timescale T2∗. Fitting the fringe frequency measures Δ, which is how qubit frequencies are calibrated.
Predict before you look
In a Ramsey experiment the drive is detuned by Δ/2π=1 MHz. At what free-evolution time τ (in ns) does P1 first reach 0?
Show the answer
500 ns
P1=cos2(Δτ/2)=0 first at Δτ=π, so τ=π/(2π×106s−1)=0.5μs. The fringe period is 1/(1 MHz)=1μs. With N repetitions of length τ, the frequency is resolved to about 1/(τN): longer τ helps, up to the coherence time.
Key takeaways
In the frame rotating at the drive frequency, the fast ω0 precession disappears and the drive splits into a static co-rotating part and a counter-rotating part at 2ω.
Within the RWA, H~=2ℏ(ΔZ+ΩcosϕX+ΩsinϕY): amplitude sets the rate, phase sets the axis, detuning tilts it, and P1=W2Ω2sin22Wt.
Pulse errors are quadratic: 1−P1≈(Δ/Ω)2 for detuning and ≈(πε/2)2 for amplitude error.
The leading non-RWA effect is the Bloch–Siegert shift Ω2/(4ω0). Ramsey fringes cos2(Δτ/2) measure detunings and T2∗.
Check yourself
In the interactive lesson these are scored with your confidence, and what you miss comes back for review.
Question 1
With the same resonant drive, how do you implement Ry(π/2) instead of Rx(π/2)?
Double the pulse duration
Shift the drive phase by 90∘
Detune the drive by Ω
Use only the counter-rotating component
Show the answer
B. Shift the drive phase by 90∘
On resonance H~RWA=2ℏΩ(cosϕX+sinϕY), so ϕ=π/2 gives the Y axis. The same envelope with a shifted phase rotates about a different equatorial axis.
Question 2
Why are the populations P0,P1 the same in the lab frame and in the frame defined by V=e−iωtZ/2?
Because the RWA is exact for populations
Because the drive is weak
Because V is diagonal in the computational basis, so it changes only the phases of the amplitudes
They are not; populations must be transformed back
Show the answer
C. Because V is diagonal in the computational basis, so it changes only the phases of the amplitudes
V multiplies the ∣0⟩ and ∣1⟩ amplitudes by e∓iωt/2 and leaves their moduli unchanged. Only the azimuth of the Bloch vector differs between frames.
Question 3
A π pulse calibrated on resonance is applied with detuning Δ=0.1Ω. About how much population is left in ∣0⟩?
10−2
10−1
10−4
0: π pulses are insensitive to detuning
Show the answer
A.10−2
1−P1=δ2+O(δ4) with δ=Δ/Ω, so about 10−2 (numerically 0.0100).
Question 4
In a Ramsey sequence (Rx(π/2), free evolution for τ, Rx(π/2)) with detuning Δ and no decoherence, P1(τ) equals:
sin2(Ωτ/2)
e−τ/T2
21 for all τ
cos2(Δτ/2)
Show the answer
D.cos2(Δτ/2)
Between the pulses the Bloch vector precesses about z by Δτ, and the second pulse maps that azimuth onto rz=−cosΔτ, so P1=cos2(Δτ/2).
Further reading
I. I. Rabi, Space quantization in a gyrating magnetic field, Physical Review (1937).
F. Bloch & A. Siegert, Magnetic resonance for nonrotating fields, Physical Review (1940).
N. F. Ramsey, A molecular beam resonance method with separated oscillating fields, Physical Review (1950).
D. C. McKay, C. J. Wood, S. Sheldon, J. M. Chow & J. M. Gambetta, Efficient Z gates for quantum computing, Physical Review A (2017).
P. Krantz et al., A quantum engineer's guide to superconducting qubits, Applied Physics Reviews (2019).