Taming the noise: harnessing thermal fluctuations in nanomagnetic systems

At the nanometer length scale, thermal fluctuations are no longer averaged out the way we are used to in our macroscopic world, and instead provide an incredibly strong force. In nanoscale magnetic systems this causes thermal switching of the magnetization from one state to the next, often resulting in unwanted effects like the decay of the quality of data stored on magnetic recording media. In contrast, in our research we try to harness these thermal fluctuations and use them to our benefit. In this talk, we’ll explore two applications:

1. Characterization of magnetic nanoparticles for biomedical applications.

For nanoparticle-based biomedical applications to work safely and reliably, the particle properties should be well known. When injected in the body, the particles’ size plays an important role in where they eventually end up, and their magnetic properties determine their efficiency in several diagnostic and therapeutic applications. Typically, their magnetic properties are investigated by measuring the particles’ response to externally applied magnetic fields. We recently demonstrated the feasibility of a radically different approach based on the analysis of the thermal noise originating from the nanoparticles in the absence of any external excitation. With the help of SQUIDs in a magnetically shielded environment, we measured a picotesla amplitude noise signal, and the shape of the resulting noise spectrum was interpreted to estimate the properties of the nanoparticles.

2. Nanomagnetic logic gates capable of reversible computing

One limitation that our everyday computers have is that they cannot solve problems known as ‘inverse computations‘. This is because the logic gates (the building blocks of computers) can produce an outcome very quickly for a given input, but are unable to tell what the input was for a given outcome. This is the basis of e.g. cryptography, which relies on the fact that a product of two prime numbers is very difficult to factorize again. In order to solve this kind of problem, computers must be developed that work fundamentally differently from the ones we use today. We developed a logic gate consisting of nanomagnets in an arrangement such that the ground states correspond to logically correct states. If we fix the output states and let the system explore its energy landscape due to thermal switching, we thus end up most often in the logically correct input states, thereby solving this inverse computation.