Course Home

Elitzur–Vaidman Bomb Tester

Interaction-free measurement with a Mach-Zehnder interferometer

The Puzzle

Imagine a factory that produces bombs with ultra-sensitive triggers. Some bombs are live (functional) and some are duds (broken triggers). The trigger is so sensitive that even a single photon will detonate a live bomb.

The challenge: Can you identify live bombs without detonating them?

Classically, this is impossible. Any interaction with the trigger that gives you information also triggers the bomb. But quantum mechanics offers a loophole through interaction-free measurement.

The Setup: Mach-Zehnder Interferometer

A Mach-Zehnder interferometer splits a photon into two paths using beam splitters and mirrors, then recombines them:

  1. Beam Splitter 1 (BS1): A single photon enters and is put into a superposition of taking the upper path and the lower path.
  2. Mirrors: Each path is reflected by a mirror so the paths converge again.
  3. Beam Splitter 2 (BS2): The two paths recombine. Depending on the phase relationship, the photon exits toward Detector D0 or Detector D1.
  4. The bomb (if present) is placed on the lower path, between the mirror and BS2.

What Does a 50/50 Beam Splitter Do?

A balanced beam splitter transforms a photon state as follows:

|upper⟩1/√2 (|upper⟩ + i|lower⟩)

|lower⟩1/√2 (i|upper⟩ + |lower⟩)

The factor of i is a phase shift of π/2 that occurs upon reflection inside the beam splitter. This phase is what makes the interferometer work — it causes constructive and destructive interference at the outputs.

Case 1: No bomb (or dud) on the path

Both paths are open. The photon travels in superposition and interferes with itself at BS2. Destructive interference ensures the photon always goes to D0 and never to D1.

Step-by-step state evolution:

  1. Source: |upper⟩
  2. After BS1: 1/√2 (|upper⟩ + i|lower⟩)
  3. After mirrors (each reflection adds phase i): i/√2 (|upper⟩ − |lower⟩)
  4. After BS2: applying the beam-splitter transform and collecting terms: −|D0⟩

Result: the photon exits 100% toward D0. D1 never clicks.

Case 2: Live bomb on the lower path

The bomb acts as a measurement device. If the photon takes the lower path, the bomb detonates — collapsing the superposition. If it takes the upper path, the superposition is destroyed anyway (the bomb could have detected it), and we lose interference.

Step-by-step state evolution:

  1. Source: |upper⟩
  2. After BS1: 1/√2 (|upper⟩ + i|lower⟩)
  3. Bomb measures "which path":
    • 50% chance: photon is on the lower path → BOOM. Bomb explodes.
    • 50% chance: photon is on the upper path → photon survives, state collapses to |upper⟩.
  4. If survived, after mirror & BS2: a single path with no interference gives 50/50 at detectors.
    • 25% overall: D0 clicks — inconclusive (this also happens with no bomb)
    • 25% overall: D1 clicks — bomb detected without detonation!
Key insight: D1 never clicks when there is no bomb. So if D1 clicks, a live bomb must be present — and the photon never touched it! This is interaction-free measurement.

Interactive Simulation

Choose a scenario and fire a photon to see what happens. Run many trials to build up statistics.

Mach-Zehnder Interferometer

|
Try this

Set the scenario to Random and fire 50 photons. Does the fraction of D1 clicks match the expected 25% for live bombs? How close do the statistics get to the theoretical predictions?

Outcome Summary

Scenario D0 D1 Boom Interpretation
No bomb / Dud 100% 0% 0% Full interference — always D0
Live bomb 25% 25% 50% No interference — D1 click = bomb found safely!

Outcome Probability Tree

Here is the full decision tree when a photon meets a live bomb:

Each branch shows the probability and outcome. Only the green branch certifies a bomb without destroying it.

Quantum Circuit Analogy

The Mach-Zehnder interferometer maps directly onto a qubit circuit:

  • Each beam splitter is a Hadamard gate (H).
  • The two paths are the |0⟩ and |1⟩ basis states of a qubit.
  • The mirrors contribute phases, analogous to the Z gate.
  • No bomb: H · H = I, so you always measure |0⟩ (= D0).
  • Live bomb: The bomb is a projective measurement in the computational basis between the two H gates. It collapses the state, destroying the interference, and you get random outcomes — 50/50 for |0⟩ or |1⟩.

Bomb-Sorting Strategy

With the basic interferometer, we can use repeated testing to sort a stockpile of bombs. Each round, fire a photon at each untested bomb:

  • D1 clicks → certified live bomb (keep it!)
  • Boom → lost forever
  • D0 clicks → inconclusive — test again next round

How many live bombs can you save from a stockpile? Adjust the starting count and see:

100
Try this

Run the sorting simulation with 100 bombs, then try 500. Does the percentage saved stay roughly the same? Why does the theoretical yield converge to ~33%?

Going Deeper: The Quantum Zeno Approach

Can We Do Better Than 25%?

With the basic setup, only 25% of live bombs are successfully identified without detonation, while 50% explode. That's a terrible yield. But the success rate can be pushed arbitrarily close to 100% using the Quantum Zeno Effect.

The idea: Replace each 50/50 beam splitter with a very weak beam splitter that only nudges a tiny fraction of the amplitude into the lower path. Then chain N such stages in a loop. Each stage rotates the state by a small angle θ = π/2N.

If no bomb is present, after N stages the full rotation completes and the photon exits toward the detector — equivalent to a single 50/50 beam splitter.

If a live bomb is present, each stage acts as a measurement. Because each rotation is tiny, the probability of triggering the bomb at any single stage is only sin²(π/2N) — which shrinks as ~1/N². Over all N stages, the total explosion probability is approximately N × sin²(π/2N) ≈ π²/4N, which vanishes as N grows.

Result: Detection probability → cos2N(π/2N) → 1 as N → ∞. We can detect live bombs with near-certainty!

Quantum Zeno: N-Stage Detection

1 |

Basic (1 stage)

Detection25%
Explosion50%
Inconclusive25%
VerdictPoor

Zeno (N=1 stages)

Detection50.0%
Explosion50.0%
Inconclusive0.0%
VerdictPoor
Try this

Slide N from 1 to 50 and watch the explosion rate plummet in the chart above. At what value of N does the detection rate first exceed 90%? Run 1000 trials at that N to verify.

Key Takeaways

  1. Interaction-free measurement is real: quantum mechanics allows you to learn about an object without any particle ever touching it. The bomb tester is the clearest demonstration.
  2. Superposition enables the trick: the photon travels both paths simultaneously. The bomb's presence on one path collapses the superposition, altering the interference pattern at the output — even when the photon takes the other path.
  3. Basic efficiency is low: the standard Mach-Zehnder setup detects only 25% of live bombs safely, while 50% explode. D0 clicks are inconclusive.
  4. The Quantum Zeno Effect fixes this: by replacing one beam splitter with N weak stages, the detection rate approaches 100% as N → ∞. Frequent measurement freezes the state on the safe path.
  5. Circuit analogy: the interferometer maps to H–measure–H on a qubit. The bomb is a projective measurement between two Hadamard gates.

Check Your Understanding

Historical Context

The bomb-testing thought experiment was proposed by Avshalom Elitzur and Lev Vaidman in 1993. It provided one of the most vivid demonstrations that quantum measurement is not merely a passive reading of pre-existing values — the possibility of an interaction can change outcomes even when no interaction occurs.

1993 — Elitzur & Vaidman publish "Quantum mechanical interaction-free measurements", introducing the bomb-testing gedanken experiment.

1995 — Kwiat, Weinfurter, Herzog, Zeilinger & Kasevich demonstrate interaction-free measurement experimentally using a polarization interferometer, confirming the theoretical predictions.

1999 — Kwiat et al. implement the Quantum Zeno approach with multiple cycles, achieving detection efficiencies above 73% — far beyond the 25% basic limit.

These experiments established interaction-free measurement as a genuine quantum phenomenon with applications in quantum imaging, counterfactual computation, and ultra-sensitive non-destructive testing.