In a new paper published in *Physical Review Letters*, physicists J. A. Sherman, et al., at the University of Oxford, have experimentally demonstrated a recovery method that can restore the state of a single qubit, in principle perfectly, after it has partially collapsed.

As the physicists explain, a full collapse of a qubit results from a measurement that reveals the qubit's state, while a partial collapse results from a measurement that can be thought of as a "peek" at the qubit because it doesn't reveal the qubit's state, but simply verifies that the qubit hasn't decayed. The problem is that, often the mere act of peeking alters the qubit's state. A recovery method would essentially reverse the effects of peeking on the qubit, thereby allowing peeking to serve as a sort of non-destructive quantum quality control technique.

The concept of a partial collapse can also be imagined in terms of Schrödinger's cat.

"To really torture the cat analogy, imagine the cat could be in three states: happy, sad, or dead," Sherman told *Phys.org*. "Then the technique is a way to measure just whether the cat is dead or not without learning anything at all about whether the cat is either happy or sad. The precise quantum mixture of happy and sad is actively recoverable even after verifying that the cat isn't dead.

The physicists explain that the qubit recovery method can be thought of as a generalization of a concept called spin echo, which can be considered as a way to "unwind" a spin rotation. The method was proposed in 2002 by L.-A.Wu, et al., and experimentally realized for the first time by N. Katz, et al., in 2008.

Here, the physicists have improved the accuracy of this method by reducing the infidelity by an order of magnitude. The improvements enabled them to achieve substantial recovery of a qubit's state for relatively large partial collapses. For instance, even with an 80% probability of decay, the information content of the qubit is preserved with an accuracy greater than 98%.

However, the recovery method is not perfect. The probability of recovering the qubit's state depends on how much it has collapsed, so that the more collapsed the qubit is, the less likely it is to recover. A fully collapsed qubit has zero probability of recovery.

Still, the recovery method could be very useful for overcoming one of the biggest challenges in developing quantum systems: decoherence, which results in the loss of a system's quantum properties.

"The source of great interest in quantum information is also its biggest weakness: quantum coherence is fragile since all quantum systems are greatly affected by noisy environments and spontaneous decay," Sherman said. "To make sophisticated use of quantum information, we need methods of detecting and correcting these random errors. The 'reversible peek' we describe is universally applicable, and is most helpful in cases where one qubit state decays much faster (or is more sensitive to noise) than the other. For photonic qubits, the 'reversible peek' can be implemented with birefringent optics and polarizers. For qubits formed from superconductors, it can be realized with microwave pulses. For atomic qubits like ours, we employed optical and radio-frequency pulses. The 'reversible peek' may be one of several techniques which promote any of these quantum computation architectures from a laboratory curiosity to a real, widely deployable, useful device."

In the future, the physicists plan to continue working on ways to use the qubit recovery method for different purposes.

"Ideally, the 'reversible peek' could become a standard and widely used part of any quantum mechanic's toolbox," Sherman said. "Consider the spin-echo. Viewed one way, a spin-echo is a correction procedure for an unwanted qubit phase shift. But far from being an academic novelty, the spin-echo finds use in nearly every quantum information experiment, to say nothing about its routine role increasing signal levels in every hospital's magnetic resonance imaging (MRI) machine. (I guess about one-third of the loud noises one hears in an MRI machine are caused by spin-echo pulses.) The 'reversible peek' we investigated has a structure very much like a spin-echo, but is designed to detect and correct for a different sort of qubit error: spontaneous decay."

In the future, the Oxford group will continue investigating many methods of making quantum computation with trapped atomic ions more scalable and robust. In this context, the 'reversible peek' quality-control technique may be among several innovations which together make a quantum computation system practical, which is the ultimate goal.

**Explore further:**
Quantum computing: Manipulating a single nuclear spin qubit of a laser cooled atom

**More information:**
J. A. Sherman, et al. "Experimental Recovery of a Qubit from Partial Collapse." *Physical Review Letters*. DOI: 10.1103/PhysRevLett.111.180501

## antialias_physorg

Not quite. The loud noise you hear in an MR stems from putting juice on individual coils and the coils consequently expanding.

Rest of the article is fascinating, though. Peeking at a state without getting any information..sounds weird but it's another demonstration that one can't trick the information content out of a quantum state.

## sigfpe

Fortunately it wouldn't. This isn't a new theory that contradicts the old theory of quantum mechanics that allowed quantum encryption. "Uncollapse" is predicted by the same theory. So the same argument for the security of quantum encryption still stands.

## the_walrus

## mogmich

In fact, I expect that something makes this impossible. Maybe the entanglement would be destroyed when you make a "weak measurement"?

## mogmich

## antialias_physorg

Problem is: to send a message you have to encode something on the particle (i.e. you have to SET a property into a defined state).

But since you only read what is already there you can't send a message (setting a defined state precludes entanglement).

If you try to set anything after entangling two entities you break entanglement (i.e. your particle may be set to the message you want to send - but the particle at the other end wouldn't react to that change)

No FTL information transmission that way, I'm afraid.

## bluehigh

Happy, sad, dead, alive or ...not there.

Measurement realises the existence of physical properties.

## MRBlizzard

About breaking the entanglement,

Please give a citation for this derivation/calculation. I want to understand this thoroughly.

Thanks

## antialias_physorg

Now you CAN use knowledge of two entangled entities for something at 'superluminal speeds': namely encryption. BTW: this is an excellent way of showing that encryption doesn't constitute information. I.e. that an encypted text does not carry more information than the same text in non-encrypted form.

## antialias_physorg

http://rationalwi...nglement

...which makes the same point almost verbatim (especially in the 'insanity?' section)

## DarkHorse66

'Insanity?' ? I think that might actually read: 'Instantly?' Was that a Freudian slip? Um, what else were you concentrating on, while reading the article? Woo? ;)

( http://rationalwi...wiki/Woo )

I hadn't come across that variation of wiki before. Nice find.

Cheers, DH66

## srikkanth_kn

Any reference for this?

## mohammadshafiq_khan_1

Nov 19, 2013## mogmich

If it is possible to only peek on the other set of three (entangled) qubits, and if this tells you which qubits are still in a superposition because they have not been measured, you get the code 101 out of it.

In other words: The measurements themselves are the message! The states of the particles are not changed in order to encode any message. Of course the state of a particle is changed if you measure it, but this change is irrelevant because it has nothing to do with the message.

## antialias_physorg

Since there is no correlation between Alice and Bob measuring/not measuring there's no information transfer because Bob has no information which bits he should measure and which ones he should disregard.

It's a subtle bit of business in information theory. In order for information to be transmitted you

a) need a priori knowledge of the message

b) need a posteriori knowledge of the sent bits

c) need to show correlation between the two

Using the entangled property as the information carrier does not allow for a). Using the measurement as information carrier does not allow for c)

Ergo: In either case no information is transmitted.