iiinova — where your digital self keeps evolving

Michael Marthaler 🇪🇺 @michaelmarthaler.bsky.social · 13h

How can we calculate NMR parameters to simulate an organic molecule’s spectrum? Propofol looks simple, but predicting its chemical shifts and J-couplings accurately is not. Quantum chemistry alone is only the starting point.

141

Michael Marthaler 🇪🇺 @michaelmarthaler.bsky.social · 30/09/2026

When an experimental ¹H NMR spectrum and a simulation disagree, is it a referencing offset—or the wrong structure? For Friedelin (60 MHz, 64 scans), we use a 3-step alignment workflow for a fair comparison.

121

Michael Marthaler 🇪🇺 @michaelmarthaler.bsky.social · 30/09/2026

000

Michael Marthaler 🇪🇺 @michaelmarthaler.bsky.social · 29/09/2026

Two takeaways on scaling accurate ¹H NMR simulations to larger molecules: align spectra transparently, then assess whether smaller spin clusters preserve coupling-driven fine structure. A friedelin comparison illustrates both points.

111

Michael Marthaler 🇪🇺 @michaelmarthaler.bsky.social · 28/09/2026

“Maybe nature is not really quantum mechanical.” A provocative riff on Feynman. Of course nature is quantum mechanical—but perhaps he meant that relevant real-world systems routinely occupy the full Hilbert space, making classical simulation exponentially hard.

131

Michael Marthaler 🇪🇺 @michaelmarthaler.bsky.social · 27/09/2026

USP’s Digital Standards initiative defines digital standards as machine-readable versions of established reference standards and test methods. It also introduces digital Reference Standards (dRS): digital data used for comparison in compendial tests. www.usp.org/digital-standards

121

Michael Marthaler 🇪🇺 @michaelmarthaler.bsky.social · 26/09/2026

What would make truly compact, affordable NMR practical outside specialized labs? A useful answer may come from integrating microfluidics with Overhauser dynamic nuclear polarization (DNP).

171

Michael Marthaler 🇪🇺 @michaelmarthaler.bsky.social · 25/09/2026

Mixtures are the default in the lab and a core challenge in process monitoring. You rarely measure pure substances—usually overlapping signatures from multiple components under changing conditions. NMR adds molecule-specific structural information, but peak overlap remains difficult.

121

Michael Marthaler 🇪🇺 @michaelmarthaler.bsky.social · 24/09/2026

NMR is one of the few techniques that lets us see deep into matter non-destructively while providing substantial structural information—and it can do so during ongoing reactions or electrochemical processes.

120

Michael Marthaler 🇪🇺 @michaelmarthaler.bsky.social · 23/09/2026

When we started HQS in 2018, I was often told low-field/benchtop NMR was a niche technology without practical applications. The field has come a long way since then.

111

Michael Marthaler 🇪🇺 @michaelmarthaler.bsky.social · 22/09/2026

UV/Vis spectroscopy reveals where electronic transitions occur and how strongly a molecule absorbs light. But in this region, electronic and vibrational transitions can occur simultaneously.

110

Michael Marthaler 🇪🇺 @michaelmarthaler.bsky.social · 21/09/2026

Spectra become actionable when models learn relationships between measurements and reference values. Partial least squares (PLS), related to PCA, connects Raman, IR, or UV–Vis spectra to concentrations and process variables.

110

Michael Marthaler 🇪🇺 @michaelmarthaler.bsky.social · 20/09/2026

Low-frequency spectra of salt solutions are mostly broad, featureless bands. Cation, anion, and water-network contributions overlap, making it hard to quantify what each contributes—or how those contributions change after excitation.

110

Michael Marthaler 🇪🇺 @michaelmarthaler.bsky.social · 19/09/2026

In UV/Vis spectroscopy, vibronic coupling turns a single vertical transition into broad, asymmetric bands. Nuclear motion affects peak position and width; at room temperature, fine vibronic structure is usually unresolved.

110

Michael Marthaler 🇪🇺 @michaelmarthaler.bsky.social · 18/09/2026

When people hear we simulate NMR spectra from first principles, they often assume the hard part is the exponential growth of the spin Hilbert space. That’s real — but in practice, symmetry, clustering, and approximations can tame a lot of it.

131

Michael Marthaler 🇪🇺 @michaelmarthaler.bsky.social · 17/09/2026

Coming from physics, I’m used to models/simulations first and fitting parameters second. In chemometrics, Multivariate Curve Resolution (MCR) almost flips that: start with a big spectral matrix, then decompose it into component spectra and concentration maps.

110

Michael Marthaler 🇪🇺 @michaelmarthaler.bsky.social · 17/09/2026

Great collaboration between HQS and Quantinuum

041

Reposted by Michael Marthaler 🇪🇺

arXiv bot (quant-ph) @krxiv-quant-ph.bsky.social · 17/09/2026

Large-scale NMR simulation on a trapped-ion quantum computer arxiv.org/pdf/2609.17102 Pascal Stadler, Alec Owens, Etienne Granet, David Muñoz Ramo, Michael Marthaler.

arxiv.org

https://arxiv.org/abs/2609.17102

arXiv abstract link

011

Michael Marthaler 🇪🇺 @michaelmarthaler.bsky.social · 16/09/2026

Quantum chemistry often benchmarks methods against each other, not measurable data. That’s understandable: most electronic-structure methods assume isolated molecules in vacuum or implicit solvent, while experiments probe complex condensed-phase environments.

130

Michael Marthaler 🇪🇺 @michaelmarthaler.bsky.social · 15/09/2026

Happy to share our paper on the Active Space Finder (ASF): an open-source tool for automatic active space selection for excitation energies.

111

Reposted by Michael Marthaler 🇪🇺

Magnetic Resonance News @nmr900.bsky.social · 14/09/2026

Preprint, Can a Quantum Computer Simulate Nuclear Magnetic Resonance Spectra Better than a Classical One? Keith R. Fratus, Nicklas Enenkel, Sebastian Zanker, Jan-Michael Reiner, Michael Marthaler, Peter Schmitteckert arxiv.org/abs/2508.06448 #NMRchat #NMR 🧲

arxiv.org

Can a Quantum Computer Simulate Nuclear Magnetic Resonance Spectra Better than a Classical One?

The simulation of the spectra measured in nuclear magnetic resonance (NMR) spectroscopy experiments is a computationally non-trivial problem which, due to its natural interpretation as a quantum spin ...

011

Michael Marthaler 🇪🇺 @michaelmarthaler.bsky.social · 14/09/2026

Can a quantum computer simulate NMR spectra better than a classical one? Our preprint asks whether quantum computers can outperform strong classical methods for standard low- and high-field 1D NMR in practically relevant cases.

132

Michael Marthaler 🇪🇺 @michaelmarthaler.bsky.social · 13/09/2026

Struqture is what I reach for when I’m building model Hamiltonians. In a video I compare spin Hamiltonian definitions in Struqture, Qiskit, and QuTiP — and the key difference is how you express the model and what happens to memory as you scale.

120

Michael Marthaler 🇪🇺 @michaelmarthaler.bsky.social · 12/09/2026

Simulating NMR spectra sounds simple—until you do it on real molecules. Sub-Hz line widths, strong coupling, and dense J-networks make time-domain + FFT approaches expensive and delicate.

130

Michael Marthaler 🇪🇺 @michaelmarthaler.bsky.social · 11/09/2026

From quantum chemistry to usable spin models for magnets: can we really make the jump? Designing magnetic materials needs exchange couplings, anisotropies, and g-factors — but the bridge from ab-initio calculations to compact spin Hamiltonians is still the hard part.

130

Michael Marthaler 🇪🇺 @michaelmarthaler.bsky.social · 10/09/2026

End-to-end quantum linear-equation solvers are still promising: HHL, VQLS, SQLS, and related methods could turn O(N) or O(N log N) work into poly-log(N) under the right conditions. But the real question is not the kernel — it’s the input pipeline.

110

Michael Marthaler 🇪🇺 @michaelmarthaler.bsky.social · 09/09/2026

Vehicle Routing Problem (VRP) was often held up as a perfect quantum-optimization demo: NP-hard, easy to explain, and very relevant in logistics. But the latest QOBLIB benchmark tells a different story.

120

Michael Marthaler 🇪🇺 @michaelmarthaler.bsky.social · 08/09/2026

How will we prove quantum advantage in quantum chemistry? One path is to go beyond ground-state energies and target excited states, where classical methods often get harder first.

110

Michael Marthaler 🇪🇺 @michaelmarthaler.bsky.social · 07/09/2026

Quantum advantage for a real problem in NMR… still this year? Maybe — and the discussion feels much less hypothetical than it did a year ago. A good example is the recent work by Leviatan et al.:

130

Michael Marthaler 🇪🇺 @michaelmarthaler.bsky.social · 06/09/2026

QSCI promises a clean quantum-classical split: quantum device finds the important determinants, classical computer does the rest. Sounds perfect for near-term hardware. But does the quantum step actually carry the hard part? A 2026 paper gives us real data to work with. 🧵

100

Michael Marthaler 🇪🇺 @michaelmarthaler.bsky.social · 02/09/2026

QOBLIB’s 2025 preprint The Intractable Decathlon gave quantum optimisation a valuable benchmark set. Its clearest lesson still stands: modelling is destiny. A problem can gain or lose quantum viability in the Transform-to-Quantum step.

110

Michael Marthaler 🇪🇺 @michaelmarthaler.bsky.social · 31/08/2026

RPA isn’t just a classical approximation—it may be a smart way to bring dynamic correlation onto quantum hardware. By compressing virtual orbitals into bosonic screening modes, it keeps quantum models compact, structured, and chemically meaningful. quantumsimulations.de/paper-random-……

010

Reposted by Michael Marthaler 🇪🇺

Juan Carlos Sancho-García @quantumalicante.bsky.social · 30/08/2026

Glad to be part of the Scientific Committee of the #DFT conferences, which met last week at @dipcehu.bsky.social The next edition will be organized in @uni-wuppertal.bsky.social, hosted by Prof. Hilke Bahmann, who will serve as Chairperson. Excellent news!! #rheochem #compchem #quantumchem

0175

Michael Marthaler 🇪🇺 @michaelmarthaler.bsky.social · 30/08/2026

Long-term quantum roadmaps only matter if they connect real industry needs with realistic progress in hardware and algorithms. At HQS, we see the path to impact as use-case driven: start small, prove value, then scale.

110

Reposted by Michael Marthaler 🇪🇺

Jesse Singal @jessesingal.com · 28/08/2026

*Jason Arday* approached a Guardian journalist to try to get him to cover claims like him receiving a severed pig's head in the mail. The journalist looked into those claims, found them to not be credible, and published an investigation. Now idiots are accusing The Guardian of murder. It's sick.

24110

Reposted by Michael Marthaler 🇪🇺

Matthew Yglesias @mattyglesias.bsky.social · 29/08/2026

Country music conquered America by becoming a more expansive version of itself. www.slowboring.com/p/everything...

slowboring.com

Everything is country

How a once-loathed genre became the sound of America

7202

Michael Marthaler 🇪🇺 @michaelmarthaler.bsky.social · 29/08/2026

Hot take: the next big opportunity for NMR may not be medicine — it may be taking work away from chromatography. As benchtop and low-field systems improve, NMR starts looking much more practical for real industrial analysis.

110

Michael Marthaler 🇪🇺 @michaelmarthaler.bsky.social · 28/08/2026

In quantum computing for chemistry, the default target is often the electronic ground-state energy at a fixed nuclear geometry. Useful, yes—but too narrow for predicting what experiments actually measure, especially in UV/Vis spectroscopy.

120

Michael Marthaler 🇪🇺 @michaelmarthaler.bsky.social · 27/08/2026

Active spaces are a great start for quantum chemistry—but not the finish line. They capture static correlation: the big multiconfigurational effects behind bond breaking, near-degeneracies, and open-shell behavior.

100

Michael Marthaler 🇪🇺 @michaelmarthaler.bsky.social · 26/08/2026

When we talk about identifying a relevant industry problem in quantum computing, NMR spectroscopy stands out as a prime candidate. This is not just academic curiosity; NMR touches major real-world applications—drug development, chemical analysis and more.

111

Reposted by Michael Marthaler 🇪🇺

Michael Marthaler 🇪🇺 @michaelmarthaler.bsky.social · 24/08/2026

only sub-problems that are really challenging for today’s classical toolchains belong on qubits. • Show quantum advantage: the benchmark is no longer a two-decade-old CPLEX default; it is the best combination of modern solvers, heuristics and hardware available right now.

041

Michael Marthaler 🇪🇺 @michaelmarthaler.bsky.social · 25/08/2026

How do we prove quantum advantage in chemistry? Not by comparing numerics to numerics. The clearest path is excited-state predictions benchmarked against high-resolution gas-phase spectroscopy—where classical methods struggle and experiment gives an independent gold standard. #Spectroscopy

020

Michael Marthaler 🇪🇺 @michaelmarthaler.bsky.social · 24/08/2026

Optimization is everywhere—from vehicle routing to staff scheduling to supply-chain design—and quantum computing is often pitched as the next big accelerator. At HQS we vet every candidate use case against four checkpoints:

120

Reposted by Michael Marthaler 🇪🇺

Jesse Singal @jessesingal.com · 22/08/2026

1/ Getting mad at Bluesky for being crazy is like getting mad at a toddler for not knowing which sppon to use at a fancy dinner party, but to be clear: This is really crazy. Cofnas is obviously being investigated in relation to the Arday thing, and that is obviously bad, because it's obviously good

6282

Reposted by Michael Marthaler 🇪🇺

Theory Department - FHI der MPG @thdept-fhi-mpg.bsky.social · 20/08/2026

👏 Congrats to Federico, Christoph, and collabs from NYU, whose paper “Multimodal Dynamics in Ionic Liquids Revealed by Molecular-Dynamics-Guided Multinuclear NMR Relaxation Analysis” is now published in J. Phys. Chem. Lett. @pubs.acs.org! 🎉 🔗 Paper: pubs.acs.org/jpclcd/artic... 📝 Paper digest 👇

pubs.acs.org

Multimodal Dynamics in Ionic Liquids Revealed by Molecular-Dynamics-Guided Multinuclear NMR Relaxation Analysis

Abstract. Ionic liquids are complex liquids characterized by high viscosity and high polarity. Their dynamics are of interest due to their use in many fiel

152

Michael Marthaler 🇪🇺 @michaelmarthaler.bsky.social · 19/08/2026

More end-to-end applications are needed for quantum computers. I wanted to share this insightful slides and notes from a presentation by Scott Aaronson at STOC 2025 in Prague. You can find the full slide deck here (ppt) www.scottaaronson.com/talks/status.…

130

Michael Marthaler 🇪🇺 @michaelmarthaler.bsky.social · 19/08/2026

We need a lot more work on how we map molecules to quantum computers. Most current efforts stop at “active-space” models—keep a handful of valence orbitals on the quantum device and treat everything else classically. Sensible as a first step, but chemists know ...

120