Take on a podcast episode from Reviewer 2 does geoengineering, originally published Thu, 16 Ap. Listen: https://podcasters.spotify.com/pod/show/reviewer2geoengineering/episodes/Avantium-Engineering-Sorbents-for-DAC---Wessels-e3i16mo

TL;DR

  • Avantium sells high-throughput sorbent testing rigs (€350k–€1M) to DAC developers including Climeworks and Shell; 150+ units deployed, mostly for catalysis.
  • Central claim: no standardized DAC sorbent test protocol exists — labs routinely publish one or two adsorption/desorption cycles as “results.” Plausible and underdiscussed.
  • Wessels says early cycles show large variation, so single-cycle data is essentially noise. Real deployment means thousands of cycles.
  • Avantium just joined the NIST Carbon Dioxide Removal consortium to push standardized breakthrough-curve protocols. Worth tracking.
  • Claimed 4x reduction in sorbent time-to-market via 16-channel parallel testing. Vendor number, but customer-testimonial-backed.

Andrew from Reviewer 2 does geoengineering interviews Rudolf Wessels, Director of Technology and Innovation at Avantium R&D Solutions, the Amsterdam-based Shell spin-out (founded 2000, publicly listed) better known for its plant-based PEF polymer. The DAC-relevant part: Avantium’s R&D Solutions unit builds parallel testing hardware that lets sorbent developers run identical adsorption/desorption experiments across multiple channels simultaneously. Episode here.

The load-bearing claim is about the state of sorbent science, not Avantium’s hardware. Wessels argues that DAC sorbent results presented at conferences are effectively incomparable: no shared test protocol, no agreed reporting baseline, and — most damning — researchers routinely showing a single adsorption/desorption cycle for a material that would need to survive thousands of cycles in a deployed system. Per Wessels, Avantium’s own testing shows “a lot of variation you already see in the first few cycles,” meaning cycle-one data isn’t just incomplete, it’s misleading. His minimum proposal is almost embarrassingly modest: report at least 10 cycles, regardless of cycle design. That this counts as an improvement tells you where sorbent benchmarking actually stands. The host pushes back reasonably — different sorbents (metal-organic frameworks vs. amines) use different beds, substrates, and desorption modes (temperature swing, steam, vacuum, moisture swing), so cross-comparison has real limits — and Wessels concedes standardization means shared guidelines and breakthrough-curve methodology, not identical rigs.

The institutional vehicle is the NIST Carbon Dioxide Removal consortium, which Avantium recently joined alongside US chemical majors, the Department of Energy, and research institutes. Wessels is candid that he doesn’t yet know the full member list — they’ve just signed on — so treat this as an early signal rather than a mature standards body. On the commercial side: rigs run €350k–€1M depending on parallelization, customers operate them in-house, sample volumes are teaspoon-scale (“a gram behaves much the same way as a ton” for intrinsic properties), and the headline efficiency claim is a 4x shortening of time-to-market with a 16-channel setup. The host correctly flags the incentive problem — a testing-equipment vendor advocating standardized testing is talking its book — but the underlying diagnosis stands independent of who profits from fixing it.

Context: this is the upstream, materials-level analog of the measurement, reporting, and verification (MRV) standardization fights happening downstream in credit markets. If sorbent data can’t be compared lab-to-lab, technoeconomic models feeding buyer decisions — including Climeworks’ cost-curve projections — inherit that uncertainty. Avantium’s DAC product page has the hardware specifics.

Useful for anyone evaluating sorbent-stage DAC startups: the “how many cycles did they actually run?” question is a cheap, high-signal diligence filter this episode hands you.