Kiesel Labs

Research

We invent new experimental primitives — qualitatively different ways to control, couple, and read out the motion of levitated nano- and microparticles — and use them to reach regimes that were previously out of range. 

What we explore

Non-Gaussian states with ever larger objects

Superpositions of ever larger, heavier and longer-lived objects are the sharpest test of how far quantum mechanics reaches. Making and certifying them is what this line of work is about. 

The obstacle is time. Delocalising a particle over a large distance normally means letting it evolve freely for a long while — and every moment of that free evolution lets decoherence and vibration accumulate. Our route is to make the protocol fast instead: the SUPER-MARIO protocol will reach interference at a mass, length and timescale previously out of range by rapidly switching linear and nonlinear optical potentials.

Colour map of the achievable coherence length versus particle radius and protocol duration, with the proposed protocol marked by a dot.
the authors (Neumeier et al., PNAS 121, e2306953121, 2024)

Achievable coherence length as a function of protocol duration and particle size. The marked point corresponds to the concrete protocol we propose.

Thermodynamics far from equilibrium

The same primitives turn levitated particles into a precise testbed for thermodynamics at its smallest scales: information erasure beyond Landauer's bound, the thermodynamics of feedback and memory, and entropy production in mesoscopic systems — in both the classical and the quantum regime. The experiments are done hand in hand with our theory collaborators.

Position of a nanoparticle over time during a bit-erasure protocol: a single trajectory drawn over the position density of many repetitions, moving from one well of the memory to the other.
the authors (Ciampini et al., Phys. Rev. Research 7, 043321, 2025)

A single trajectory together with the distribution of particle positions during a bit-erasure protocol. Starting from a non-equilibrium state allows the erasure to go below the equilibrium Landauer bound.

To top

What we build

Three primitives. Each one exists because it opens up one of the two directions above.

Deliver — clean particles into the trap

HCPCF · hollow-core photonic-crystal fibre loading

Every levitation experiment starts with getting a particle into the trap. For ultra-high vacuum and for multiparticle experiments, that has to happen cleanly and deterministically. We pioneered loading through hollow-core photonic-crystal fibres: delivery of nanoparticles from ambient pressure directly into ultra-high vacuum — an enabling technology the field has come to rely on.

Video frame of a hollow-core fibre delivering a bright nanoparticle into an optical trap, with the measured power spectral density of the trapped particle below.
the authors (Lindner et al., Appl. Phys. Lett. 124, 143501, 2024)

A frame from the video of a complete handover of a 100 nm nanoparticle from the hollow-core-fibre conveyor belt into the target trap, recorded at 8 mbar. Below: the power spectral density measured at the dark port of the Sagnac trap once the particle is captured.

Shape — engineering the potential landscape

Dynamic, nonlinear and pulsed optical potentials → SUPER-MARIO

A particle can only do what its potential allows. We go beyond the harmonic trap and shape the landscape itself: switching potentials on the timescale of the motion, cubic and other nonlinear potentials, pulsed interactions for matter-wave control, and inverted landscapes in which the particle sits where the light is darkest. This is the primitive that makes the SUPER-MARIO route to macroscopic superpositions possible — and it provides an interesting domain for thermodynamics experiments far from equilibrium.

Schematic of an optical tweezer with electrodes, three shapes of the optical potential, motional spectra of the particle and an example pulse sequence.
the authors (Ciampini et al., arXiv:2511.22079)

SLM-based control of the optical potential, and the analysis of the resulting process, with a nanoparticle in an optical tweezer.

Control — trapping where there is no trap

FLIP · feedback levitation on an inverted potential · cavity coupling

Shaped landscapes are often unstable — an intensity minimum is not a trap. We hold a particle there in two complementary ways. 

Actively, by measuring and correcting. FLIP stabilises a levitated particle at an intensity minimum using optimal feedback control, developed together with control engineers: the position is measured continuously and a corrective force applied faster than the particle can escape. Levitating in the dark suppresses photon recoil and absorption, and opens a path to longer coherence. 

Passively, by letting the light do the work. Placing the particle inside an optical cavity shapes its motion without any measurement at all — the line of work in which we pioneered the cavity cooling of levitated nanoparticles.

Schematic of an optical tweezer with electrodes, three shapes of the optical potential, motional spectra of the particle and an example pulse sequence.
the authors (Ciampini et al., arXiv:2511.22079)

SLM-based control of the optical potential, and the analysis of the resulting process, with a nanoparticle in an optical tweezer.

Collaborations

Many of our experiments are run jointly with the Aspelmeyer Group at the Faculty of Physics. Beyond that we work with Andreas Kugi and Andreas Deutschmann-Olek (TU Wien) on optimal feedback control, Andreas Dechant (Kyoto) and Eric Lutz (Stuttgart) on stochastic thermodynamics, Mauro Paternostro (Belfast / Palermo) on quantum thermodynamics, Uroš Delić and Philipp Haslinger (TU Wien), and Anton Zasedatelev (Helsinki).

To top