An interpreter in a crowded room tries to make out one person's speech but hears the murmur of hundreds of voices. A molecule storing quantum information in its spin finds itself in a similar situation: its spin is surrounded by the thermal jitter of atoms, and it is this jitter that gradually erases memory. The task is to figure out which vibrations are the ones "drowning out" the rest.
Copper-based molecular qubits are attractive because their properties can be tuned by changing the chemical environment. But calculations of spin lifetime stubbornly disagreed with experiment: theory insisted that slow distortions of the molecular shape were to blame, while experiments pointed to fast stretching of the metal–nitrogen bond. What was missing was a tool that could automatically pick out, from a warm, jittering molecule, exactly the motions that actually disturb the spin.
For three variants of copper porphyrin, molecules that differ only in their side appendages, a computer simulation was run at room temperature. Every five femtoseconds (millionths of a billionth of a second), how the spin "feels" the magnetic field was recorded. Then, using a regression that automatically discards the unimportant, these recordings were decomposed into individual vibrations. The result was unambiguous: the main culprits are fast metal–nitrogen bond stretches with frequencies around 100–300 inverse centimeters, not slow distortions as previously thought.
Plugging the found couplings into formulas that describe energy transfer from the spin to vibrations, the experimental order of relaxation times was reproduced. The molecule with intermediate appendages lost memory fastest; the other two were slower and almost identical. The temperature curve got the correct bend, reflecting the transition from one loss mechanism to another.
The interpreter has finally learned to hear, in the hubbub, exactly the voices that drown out the conversation. Now, by tweaking the chemical "settings" of the molecule, one can predict in advance which vibrations need to be damped to make quantum memory last longer.
The three porphyrins studied differ only in the substituents around the macrocycle periphery, yet their g-tensors fluctuate differently: the least distorted CuOEP has fluctuation amplitudes comparable to the most distorted CuTiPP, while the intermediate CuTPP shows noticeably smaller ones. Static geometry does not predict dynamics.