Chapter 12 of 14
Qubits, Quantum Gates, and Decoherence in Real Devices
The abstract mathematics of two-level systems becomes very concrete when encoded in superconducting circuits, trapped ions, and spins. Trace how qubits are manipulated, entangled, and ultimately threatened by noise and decoherence in today’s quantum hardware.
From Two-Level Systems to Physical Qubits
Abstract Qubits vs Real Devices
You know abstract two-level systems: a state c0|0> + c1|1> with |c0|^2 + |c1|^2 = 1. In quantum computing, this is a qubit. Real hardware must physically realize these |0> and |1> states.
Common Physical Qubits (2026)
Superconducting transmons use the two lowest energy levels of a nonlinear LC circuit. Trapped ions use two internal states of an ion. Spin qubits use electron or hole spin-up and spin-down in a quantum dot.
Same Math, Different Physics
Despite very different hardware, all platforms engineer a two-level subspace that behaves like the same abstract qubit. Control pulses cause rotations within this subspace.
Guiding Questions
As we proceed, ask: What are |0> and |1> physically? How do we implement rotations and entangling gates? Which noise sources disturb this ideal two-level picture?
Bloch Sphere: Visualizing a Single Qubit
Bloch Sphere Coordinates
Any pure qubit state can be written as |ψ> = cos(θ/2)|0> + e^{iφ} sin(θ/2)|1>. This defines a point on the Bloch sphere with angles θ (polar) and φ (azimuthal).
Key Bloch Sphere Points
North pole: |0>. South pole: |1>. Equator: equal superpositions like (|0> + e^{iφ}|1>)/√2. The x-axis has |+> and |->; the y-axis has (|0> ± i|1>)/√2.
Control as Rotations
In hardware, microwave or laser pulses implement rotations of the Bloch vector. Gate design becomes pulse design that moves the state along the sphere.
Noise on the Bloch Sphere
Noise often appears as random rotations or as shrinkage of the Bloch vector toward the center, representing loss of phase information and mixed states.
Single-Qubit Gates as Rotations (X, Y, Z, H)
Single-Qubit Gates = Rotations
Single-qubit gates are unitaries that act as rotations on the Bloch sphere. They move the Bloch vector without changing its length (for ideal gates).
Pauli-X and Pauli-Z
X = [[0,1],[1,0]] flips |0> and |1>, a π rotation around x. Z = [[1,0],[0,-1]] leaves |0> and flips the phase of |1>, a π rotation around z.
Hadamard Gate
H = (1/√2)[[1,1],[1,-1]] maps |0> to |+> and |1> to |->. On the Bloch sphere, it rotates z-basis states to the x-basis on the equator.
Real Devices and Pulses
In superconducting and ion-trap qubits, X/Y gates come from resonant pulses; Z gates often come from phase shifts in the control frame, implemented in software or phase control.
Thought Exercise: Visualizing Gates on the Bloch Sphere
Work through these mentally (or sketch the Bloch sphere):
- Start in |0> (north pole). Apply H.
- Where is the state on the Bloch sphere now?
- Is it an eigenstate of X, Y, or Z?
- Start in |+> = (|0> + |1>)/√2 (on the +x axis). Apply Z.
- What state do you get in Dirac notation?
- Where does that sit on the Bloch sphere?
- Start in |0>. Apply H, then Z, then H again (HZH).
- Compute HZH as a matrix or track the state.
- Which standard gate does this sequence equal (X, Y, or Z)?
Pause and actually compute:
- For part 2, write Z|+> explicitly.
- For part 3, either multiply matrices or follow the state step by step.
After you finish, check yourself:
- Z maps |+> to |->, a flip along the x-axis.
- HZH = X, so combining gates can reproduce other rotations.
Two-Qubit Gates and Entanglement (CNOT Example)
Why Two-Qubit Gates?
Entanglement requires at least one two-qubit gate. Without it, circuits are just independent single-qubit rotations and cannot generate Bell states or perform universal quantum computing.
CNOT Action
CNOT flips the target qubit if the control is |1>. In basis |00>,|01>,|10>,|11>: |00>→|00>, |01>→|01>, |10>→|11>, |11>→|10>.
Creating a Bell State
From |00>, apply H on qubit 1 to get (|00>+|10>)/√2. Then apply CNOT (1→2) to get (|00>+|11>)/√2, a maximally entangled Bell state.
Real Hardware Entangling Gates
Superconducting qubits use microwave-enabled couplers or cross-resonance; trapped ions use shared motional modes and laser pulses. These gates are slower and noisier than single-qubit gates.
Simulating a Simple Quantum Circuit with Qiskit
Let’s simulate a 2-qubit Bell-state circuit using Python and Qiskit, which as of 2026 remains a widely used open-source framework for quantum circuits.
The circuit:
- Start in |00>.
- Apply H on qubit 0.
- Apply CNOT with control=0, target=1.
We will inspect the resulting statevector and measurement statistics.
Decoherence: T1, T2, and Noise Sources
Open Quantum Systems
Real qubits interact with their environment. This causes decoherence, gradually degrading quantum information and turning pure states into mixed states.
T1: Relaxation
T1 is the relaxation time for |1>→|0>. It measures how fast excited-state population decays. On the Bloch sphere, it pulls states toward |0> at the north pole.
T2: Dephasing
T2 measures phase coherence loss between |0> and |1>. It shrinks the Bloch vector toward the z-axis, destroying superpositions even if populations stay fixed.
Typical Noise Sources Today
Superconducting qubits: dielectric loss, flux/charge noise, crosstalk. Trapped ions: laser noise, motional heating, magnetic fluctuations. Their T1/T2 and gate error rates differ significantly.
Estimating the Impact of Decoherence
Use these back-of-the-envelope estimates to build intuition.
- A superconducting qubit has T1 = 100 µs. A single-qubit gate takes 50 ns.
- Roughly how many such gates can you apply before significant T1 decay occurs?
- Hint: compare gate time to T1.
- A trapped-ion qubit has T2 = 1 s. A two-qubit gate takes 100 µs.
- How many two-qubit gates fit into one T2 time scale?
- Suppose a two-qubit gate has an error rate of 1×10^-2 (1%).
- For a simple algorithm requiring 100 such gates in sequence, what is a rough upper bound for the total success probability if errors were independent and no error correction is used?
- Hint: think (1 - errorrate)^numberof_gates.
Reflect:
- These quick estimates explain why error correction and noise mitigation are central topics in current (as of 2026) quantum computing research.
Check Understanding: Gates and Decoherence
Answer this question to check your understanding of gates and decoherence in real devices.
Which statement best describes T2 for a qubit in current quantum hardware (e.g., superconducting or trapped-ion devices)?
- T2 is the time scale for population decay from |1> to |0>, and it is always equal to T1.
- T2 characterizes loss of phase coherence between |0> and |1>, and it is often shorter than or equal to 2T1.
- T2 measures the speed of two-qubit gates like CNOT and is independent of environmental noise.
- T2 is only relevant for trapped-ion qubits and not for superconducting qubits.
Show Answer
Answer: B) T2 characterizes loss of phase coherence between |0> and |1>, and it is often shorter than or equal to 2T1.
T2 describes dephasing, the loss of phase coherence between |0> and |1>. In real devices, T2 is typically less than or equal to 2T1; if it is much shorter, strong pure dephasing is present. T1 describes energy relaxation, not T2, and both are relevant for all physical qubit platforms.
Review Key Terms
Flip through these flashcards to reinforce the main ideas from this module.
- Qubit
- A quantum two-level system with basis states |0> and |1>, described by a superposition c0|0> + c1|1> with |c0|^2 + |c1|^2 = 1.
- Bloch sphere
- A geometric representation of a single-qubit pure state as a point on the surface of a unit sphere, parameterized by angles θ and φ.
- Hadamard gate (H)
- A single-qubit gate that maps |0> to (|0> + |1>)/√2 and |1> to (|0> - |1>)/√2, rotating between z-basis and x-basis states.
- CNOT gate
- A two-qubit controlled-NOT gate that flips the target qubit if the control qubit is |1>, used to generate entanglement such as Bell states.
- T1 (relaxation time)
- The characteristic time scale for energy relaxation from |1> to |0>, causing population decay of the excited state.
- T2 (dephasing time)
- The characteristic time scale for loss of phase coherence between |0> and |1>, shrinking superpositions without necessarily changing populations.
- Decoherence
- The process by which a quantum system loses coherence due to interaction with its environment, turning pure states into mixed states.
- Superconducting qubit
- A qubit implemented using superconducting circuits (e.g., transmons) where |0> and |1> are the two lowest energy levels of a nonlinear resonator.
- Trapped-ion qubit
- A qubit encoded in two internal energy levels of a single trapped ion, controlled by laser pulses and shared motional modes.
Key Terms
- Qubit
- A quantum two-level system spanned by |0> and |1>, capable of existing in superpositions c0|0> + c1|1>.
- CNOT gate
- Controlled-NOT gate that flips a target qubit conditioned on the control qubit being in state |1>.
- Spin qubit
- A qubit encoded in the spin states of an electron or hole in a semiconductor or defect center.
- Decoherence
- Loss of quantum coherence due to interaction with an environment, leading to classical-like behavior.
- Bloch sphere
- A unit sphere representation where any pure single-qubit state corresponds to a point on the surface defined by angles θ and φ.
- Two-qubit gate
- A unitary operation acting on two qubits, capable of generating entanglement (e.g., CNOT, CZ).
- Hadamard gate (H)
- A gate that creates and manipulates equal superpositions, mapping computational basis states to x-basis states.
- Single-qubit gate
- A unitary operation acting on one qubit, typically corresponding to a rotation of the Bloch vector.
- Trapped-ion qubit
- A qubit realized using internal states of trapped ions, controlled by lasers and shared vibrational modes.
- T2 (dephasing time)
- Time constant characterizing loss of phase coherence between |0> and |1>, often limited by both relaxation and pure dephasing.
- T1 (relaxation time)
- Time constant characterizing energy relaxation from the excited state |1> to the ground state |0>.
- Superconducting qubit
- A qubit realized in superconducting circuits, controlled by microwave pulses and commonly used in current quantum processors.