UFZ Leipzig collaboration · Kolbe electrolysis · Max Pohl · draft 8 Oct 2026
Which parameters drive Kolbe performance, and by how much
Max’s 2026 paper (Pohl et al.) reports the main trends for mixed-acid Kolbe electrolysis in its abstract, which is the part MESSAI has read: the concentration threshold, the preference for longer chains, and the energy advantage of a one-chamber cell. MESSAI’s contribution starts there. We would map every operating parameter against every outcome, quantify each link from his runs and the wider Kolbe literature, and build a predictor that returns efficiency, product mix and kWh per litre, with an interval and the parameters that drive it, for conditions the data cover.
The question the predictor answers
“For this feed and these operating conditions, what CE to fuel, product mix and kWh per litre should I expect, how sure are we, and which parameters matter most for this prediction?”
Answers follow MESSAI’s prediction format. Each answer lists the parameters driving it and the evidence behind it. If an input is missing, the predictor says which one instead of filling in a default. Until the model is fitted, every field is pending.
query feed = { acetic 0.30, butyric 0.20, caproic 0.50 } mol/L (example inputs)
anode = Pt · j = 50 mA/cm² · one-chamber · 20 °C · pH = ?
answer CE_fuel { value: pending, unit: "%", ci_low: pending, ci_high: pending }
energy { value: pending, unit: "kWh/L", ci_low: pending, ci_high: pending }
drivers total concentration · composition · current density · electrode (ranked: pending)
evidence Pohl 2026 runs + Kolbe literature (extracted, review and harmonisation pending)
missing pH / degree of neutralisation → named as missing, no default filled in, answer flaggedThe plan on one page
What comes first, what each step needs, how we test the model, and what has gone wrong so far.
Agree the Kolbe parameter set
Not started434 extracted values link to no parameter yet (28 Oct 2026), so none can be compared.
- Needs
- Max’s check of the ten parameters and seven outcomes.
- Done when
- The set is agreed, and the extracted values link to it.
Build the evidence base
Needs Max54 Kolbe papers are registered, 9 have extracted values, and no value is reviewed yet.
- Needs
- His list of 20–50 reliable papers, and the run table behind the 2026 paper, failed runs included.
- Done when
- The evidence table shows what the field measured, with a source snippet for each value.
Reproduce his paper, then predict beyond it
Waiting on dataA model that cannot recover known trends should not be trusted on new ones.
- Needs
- The run table.
- Done when
- His reported trends come back, and predictor v1 is scored on held-out studies.
Full lists:To-dosDeliverablesWhat we ask
What the model should do, and how we test it
Each test copies something a researcher really does. Pass bars are proposals until the partner agrees them.
| The model should | How we test it | It passes when | Today |
|---|---|---|---|
| Recover what Max already published | Encode his runs and check the concentration threshold, the chain-length preference and the one-chamber energy advantage. | All three trends come back, or we write down what does not reconcile. | Needs partner data |
| Predict another lab’s Kolbe runs | Leave one study out and predict its CE to fuel and kWh per litre. | It beats the pooled baseline, and 90% intervals cover 85–95% of the values. | Needs partner data |
| Say what it does not know | Ask with pH missing, then ask about a real broth outside the data. | It names pH as missing and marks the broth answer as an extrapolation. | Not run yet |
Measured without partner data
Two checks we can run on published work alone. Both can be repeated, and the numbers are the same on every page.
| Check | Result | What it leaves open |
|---|---|---|
| Are the values we extract really in the paper? | 96% of 308 checked values from research papers are, across 20 papers (31 Oct 2026). | Checked by machine, not yet by a person. 101 values from review papers are other studies’ numbers. 41% of true values reach the model. |
| Does the model match published runs it never saw? | Lab simulator: 7 of 17 within 25%. API and chat: 4 of 17, and 8 got no answer (31 Oct 2026). | The simulator got 0 of 5 MFC values within 25%. An MFC request with no COD still answers from its materials alone. |
What has gone wrong so far
These rows come from our own sessions and checks, not from partner reports. New reports land here.
| What happened | What it changes | Found in |
|---|---|---|
| 434 Kolbe values sit in the corpus with no parameter link and no SI value. | Parameter set first, then harmonisation. No value enters a fit before review. | Our check |
| 45 of 54 Kolbe papers have no full text in MESSAI, and we did not record which are paywalled. | Record the reason per paper, and ask Max which papers matter most. | Our check |
| What we say about Max’s paper rests on its abstract, not on his data. | His run table replaces the abstract as the source for those cells. | Our check |
How a failure gets back into the model
- Report. Comment on this page, or send us the run.
- Reproduce. We turn it into a test with the real input and the expected result.
- Fix. We change the model or the pipeline at the cause.
- Re-test. The test stays in the set, and the result shows here.
We run these steps by hand. None of it is automated yet.
How this works
The path from a published paper to a number on this page. Each step feeds the next. The first three show what this site holds right now for the papers cited here.
Find the paper
DoneA paper enters the corpus as a record: DOI, title, authors, journal, date and abstract, taken from Crossref or the publisher.
- 12 of 12 cited papers with a DOI are in this corpus.
- Counted in this deployment on Nov 21, 2026.
- On 8 Oct 2026 the papers cited by the protocol chart were added to the staging corpus as records with an abstract only. On 31 Oct 2026 the open-access PDFs that a script could fetch were stored. The counts in these steps are read from the deployment you are on.
Get the full text
Partly doneWhere the publisher allows it, the PDF is stored and converted to text, tables and equations. Values can also be read from open-access full text without storing the PDF. A paper with neither stays at its abstract: only what the abstract states is available, without its tables or methods.
- 6 of 12 have a stored PDF.
- On 11 Oct 2026 the open-access full text of some of them was read. Paywalled papers, and open-access ones that a publisher does not serve to a script, still have an abstract only. Fetching the rest is part of the “Kolbe literature corpus” to-do, which needs Max’s list (see Our to-dos).
Extract values with their evidence
Partly doneA language model reads the text and tables and writes every value as one row: parameter, number, unit, the conditions it was measured under, the quote it came from, a confidence and the extractor version.
- 6 of 12 have extracted values (334 distinct facts in all; a value extracted twice counts once).
- On 11 Oct 2026 an AI model extracted values from that full text, each with the sentence it came from. A check confirms that the sentence is in the paper and contains the number. All values are pending review. Extraction with a Kolbe-specific specification is still part of the same to-do: the general extractor returns no values for a paper without microbes, so the stored PDFs that have no values yet wait for it.
Harmonise
Not done yetEach row is mapped to a canonical parameter where a rule matches, so “power density” means one thing across papers, and its unit is converted to SI. Conversion is not comparability: the species, the reference electrode and the area or volume basis must match too. Those checks are incomplete, so some rows that cannot be compared still reach the models.
- Canonical names and units for the Kolbe parameters are the first of our to-dos (see Our to-dos). Until then the extracted values keep each paper’s own names and units, so the predictor cannot use them.
Model and predict
Not done yetModels are fitted on harmonised rows only: first the effect of one parameter inside each study, then pooled across studies. A prediction returns a value, a unit, an interval, a confidence and its source. The aim is that a missing input is named, never filled with a default; some MESSAI predictors still fill one today, such as an applied voltage of 0.6 V for an MEC.
- Nothing is fitted for Kolbe yet: every field of the predictor’s answer is pending (see the question at the top of the page).
Check it
Links listedThe papers behind the protocol chart are listed below with a link to each publisher. The chart’s steps are steps of the shared protocol graph. Signed-in readers can comment on this page. A listed paper is a pointer for checking, not a check MESSAI has made.
Go to the papers
Parameter–outcome map
Rows are operating parameters, columns are outcomes. Each cell says how much we know about how that parameter moves that outcome. Most of the map is empty. The predictor’s value lies in filling those cells with effect sizes and intervals, starting with the ones that matter most for design decisions. Select a cell for the evidence and how we would quantify it.
| Parameter | CE to fuel | Kolbe vs Hofer–Moest | O₂ (OER) share | Chain shortening | Product chain length | Cell voltage | kWh per L fuel |
|---|---|---|---|---|---|---|---|
| Total acid concentrationmol/L | |||||||
| Chain-length compositionmole fractions | |||||||
| Cell designone / two chamber | |||||||
| Anode current densitymA/cm² | |||||||
| Anode materialPt, carbon, … | |||||||
| pH / neutralisationpH or % | |||||||
| Temperature°C | |||||||
| Supporting electrolytemol/L, type | |||||||
| Real broth vs syntheticimpurities | |||||||
| Electrode age / run order# runs |
“Reported” = stated in the abstract of Pohl et al., ChemSusChem 2026 (doi 10.1002/cssc.70753); page, table, number of runs and the meaning of ± are not recorded here yet. “Known direction” = general Kolbe electrochemistry, not quantified for these mixtures; to confirm with Max. “Physics” = follows from cell electrochemistry and is computable once the inputs are recorded.
Your runs
Drop a CSV of your own runs to see which cells of the map they could inform. The file is read in your browser and never uploaded or stored; nothing is fitted and no prediction changes.
Open puzzles
An unknown cell is a relationship this page has no measurement or literature estimate for yet, which may only mean nobody has written it down here. Below, one per parameter: what is unknown, the experiment that would fill it, and the bench steps of the Kolbe protocol it goes through.
Ordered by how many empty cells one experiment would fill. That is a count, not a measure of how much each run would narrow the predictions; that ranking needs the predictor (see “Next-measurement ranking” under “Our to-dos”).
pH / neutralisation
7 of 7 outcomes unknown
- CE to fuel
- Kolbe vs Hofer–Moest
- O₂ (OER) share
- Chain shortening
- Product chain length
- Cell voltage
- kWh per L fuel
Would be filled by a neutralisation series at fixed concentration.
Bench steps
Temperature
7 of 7 outcomes unknown
- CE to fuel
- Kolbe vs Hofer–Moest
- O₂ (OER) share
- Chain shortening
- Product chain length
- Cell voltage
- kWh per L fuel
Would be filled by a temperature series with cooling control.
Bench steps
Real broth vs synthetic
7 of 7 outcomes unknown
- CE to fuel
- Kolbe vs Hofer–Moest
- O₂ (OER) share
- Chain shortening
- Product chain length
- Cell voltage
- kWh per L fuel
Would be filled by paired runs: synthetic mixture against real broth at the same acid composition.
Bench steps
Electrode age / run order
7 of 7 outcomes unknown
- CE to fuel
- Kolbe vs Hofer–Moest
- O₂ (OER) share
- Chain shortening
- Product chain length
- Cell voltage
- kWh per L fuel
Would be filled by run order recorded on every run.
Bench steps
Show the other 6
Anode material
6 of 7 outcomes unknown
- CE to fuel
- O₂ (OER) share
- Chain shortening
- Product chain length
- Cell voltage
- kWh per L fuel
Would be filled by the same feed on two or more anode materials.
Bench steps
Supporting electrolyte
6 of 7 outcomes unknown
- CE to fuel
- Kolbe vs Hofer–Moest
- O₂ (OER) share
- Chain shortening
- Product chain length
- kWh per L fuel
Would be filled by runs with and without added salt.
Bench steps
Cell design
5 of 7 outcomes unknown
- CE to fuel
- Kolbe vs Hofer–Moest
- O₂ (OER) share
- Chain shortening
- Product chain length
Would be filled by paired runs in both cell types.
Bench steps
Chain-length composition
4 of 7 outcomes unknown
- Kolbe vs Hofer–Moest
- O₂ (OER) share
- Chain shortening
- Cell voltage
Would be filled by mixtures with different compositions at the same total concentration.
Bench steps
Anode current density
4 of 7 outcomes unknown
- CE to fuel
- Chain shortening
- Product chain length
- kWh per L fuel
Would be filled by a current-density series at fixed concentration and composition.
Bench steps
Total acid concentration
2 of 7 outcomes unknown
- Kolbe vs Hofer–Moest
- Product chain length
Would be filled by runs across concentration with everything else fixed (to confirm with Max that he has these).
Bench steps
How MESSAI turns this into predictions
MESSAI already runs this pipeline on the bioelectrochemistry corpus. For Kolbe it needs a parameter set, a literature corpus and his run data. The steps run in this order because each one feeds the next.
Within-study effects
Exists in MESSAIFor each study, estimate how one parameter moves an outcome while the others are held fixed. This is the cleanest evidence for a trend: differences between labs drop out, though confounders inside a study, such as run order or electrode age, remain. MESSAI fits these as population-level Bayesian effects and shows them in the lab’s sensitivity chart (for example, the within-paper HRT → COD removal effect in the artifact of 3 Sep 2026 rests on 8 papers and 60 observations; it has not converged and its interval includes zero). His paper would be the first Kolbe study encoded this way.
Cross-study pooling
Exists in MESSAIPool the same effect across Kolbe papers with a hierarchical meta-analysis (MESSAI’s
/meta-analysisforest plots). This answers whether his trends hold for other electrodes, solvents and labs, and how much they vary between studies.Correlation structure between parameters
Exists in MESSAISome parameters move together: concentration raises conductivity, which lowers cell voltage; chain length and hydrophobicity rise together. MESSAI’s learned dependency graph and knowledge graph record links like these (none yet for Kolbe), so that a predictor would not count one effect twice. Correlation across studies isn’t causation, so causal claims rest on step 1.
Compact laws by symbolic regression
Exists in MESSAIMESSAI already uses symbolic regression (PySR) to search for short equations between parameters. For Kolbe the target is a law such as CE_fuel = f(concentration, mean chain length, current density) that a chemist can read and test, rather than a black box.
Predictor with intervals and drivers
To build for KolbeA hierarchical model with study as a random effect, using the effects from steps 1–4 as priors and his runs as the core data. His own study would be left out of those priors, so his runs are not counted twice. It would return each prediction with an interval and a ranked list of the parameters driving it. The intervals would still need calibrating on held-out studies: MESSAI’s served predictors do not return calibrated intervals yet.
outcome ~ f(concentration, composition, j, electrode, cell, pH, T, …) + (1 | study) + residual GP- To be validated by
- leaving one study out at a time. The real test is predicting another lab’s results, not interpolating within his own runs.
- Planned guardrail
- outside the conditions covered by data, the answer is flagged as an extrapolation; a wider interval alone would not make it reliable
Protocol
The Kolbe methodology as a graph of steps, from preparing the feed to calculating CE and energy per litre. Click a step for its setpoints, readouts, precautions and sources. Left to right is execution order, and every step links to the full /protocols graph. A selected step also shows which rows and columns of the parameter map it feeds.
Loading the protocol graph…
What this adds beyond the paper
- Generalisation. Do the threshold and the chain-length preference hold for other electrodes, electrolytes and labs? The pooled effects and their spread between studies would answer that.
- Parameters he didn’t vary. Current density, pH, temperature and electrode material would get literature-based estimates with wider intervals, until his runs narrow them.
- Interactions. Does the concentration threshold shift with current density or with composition? Single-factor trends can’t show this; a joint model can.
- A design tool (planned). A parameter combination inside the supported region in, a prediction with interval and drivers out. Run backwards, it would answer “which conditions reach 50 % CE_fuel at the lowest kWh/L for this broth?”
- Where to measure next. The empty cells in the map, ranked by how much each would narrow the predictions that matter, become the next experiments.
Illustration: one of his trends as a what-if (not fitted)
His finding that longer chains react first becomes a starting model of the product mix with one parameter, γ; whether one parameter is enough is open (see “Model form is open”). Each acid makes radicals in proportion to its concentration and a chain-length preference γ, and radicals pair at random. Fitting γ to his per-acid data would give a product distribution, conditional on coupling, for feeds like the ones he tested; it would not predict conversion, CE, yield, fuel volume or energy. It ignores oxygen evolution, Hofer–Moest, chain shortening and depletion during a batch; the full predictor adds those.
The model, exactly
- =radical weight of acid i [relative]
- =amount of acid i in the feed [relative]
- =carbons in acid i [C atoms]
- =chain-length preference [per C atom]
- =share of all radicals from acid i [0–1]
- =share of coupling events joining radicals i and j [0–1]
- =carbons in the dimer [C atoms]
Parameters: 1 (γ). Not fitted; γ is a slider.
An acid with n carbons loses CO₂ and leaves a radical with n − 1 carbons, so two radicals couple to a dimer with nᵢ + nⱼ − 2 carbons. The “per-acid weights” switch below replaces e^(γ(n−2)) by one free weight per acid.
Model form is open
One parameter may not be enough, and this page does not claim it is. These are the candidate forms; each is a different reactivity in the weight above.
- (a) Per-acid reactivity kᵢ: one free weight for each acid C2–C8, 7 parameters (only their ratios change the mix, so 6 are identifiable). The “per-acid weights” switch in the example shows this form.
- (b) Two parameters: weight = amount · exp(γ(n − 2) + δ), with δ an offset for acetate (C2) only, since a δ shared by all acids would cancel in the normalisation. Or a preference that saturates with chain length instead of growing without limit.
- (c) γ depends on current density: γ(j), so the same feed can give a different mix at a different current density. At least one more parameter.
Which data decide: his per-acid conversion per run, together with the feed and the current density of that run. Method: fit each form to those runs, compare the forms on held-out runs and by AIC, and report the parameter uncertainty (intervals, not point values). Until that data arrives, the one-γ form is an illustration only.
Example feeds
Example compositions, not measured broths.
Relative amount in feed
Model form
γ = 0: no chain-length preference. γ > 0: longer acids make radicals faster, the direction his paper reports. γ has not been fitted; whether his per-acid data can identify it is an open question.
- Gas, C2–C4
- 12.2 %
- Light liquid, C5–C7
- 30.6 %
- Jet carbon-number range, C8–C14
- 57.1 %
Molar shares of coupling events, not conversion, CE, yield or fuel volume. The classes group carbon numbers; they say nothing about fuel properties.
Dimer products by carbon number, % of coupling events
Share of radicals by acid
Show as table
| Product | Class | % of dimers | Pairs |
|---|---|---|---|
| C2 ethane | Gas | 2.0 | C2+C2 |
| C3 propane | Gas | 4.1 | C2+C3 |
| C4 butane | Gas | 6.1 | C2+C4, C3+C3 |
| C5 pentane | Light liquid | 8.2 | C2+C5, C3+C4 |
| C6 hexane | Light liquid | 10.2 | C2+C6, C3+C5, C4+C4 |
| C7 heptane | Light liquid | 12.2 | C2+C7, C3+C6, C4+C5 |
| C8 octane | Jet carbon-number range | 14.3 | C2+C8, C3+C7, C4+C6, C5+C5 |
| C9 nonane | Jet carbon-number range | 12.2 | C3+C8, C4+C7, C5+C6 |
| C10 decane | Jet carbon-number range | 10.2 | C4+C8, C5+C7, C6+C6 |
| C11 undecane | Jet carbon-number range | 8.2 | C5+C8, C6+C7 |
| C12 dodecane | Jet carbon-number range | 6.1 | C6+C8, C7+C7 |
| C13 tridecane | Jet carbon-number range | 4.1 | C7+C8 |
| C14 tetradecane | Jet carbon-number range | 2.0 | C8+C8 |
An acid with n carbons gives a radical with n−1 carbons, so acids i and j give an alkane with nᵢ + nⱼ − 2 carbons. Acetic acid gives methyl radicals; acetate–acetate coupling gives ethane. Shares are molar, per coupling event.
Models
The electrolysis step the predictor is about, and a microbial step that could supply acids upstream. No coupling between the two is modelled, so the pair is an illustration, not a process design. Both are shared MESSAI 3D models; drag to orbit.
The electrolysis step
Kolbe electrolysis cell
Example one-chamber (undivided) cell: Pt anode, cathode with H₂, carboxylate electrolyte, fuel layer separating on top. The geometry is an example, not Max’s setup.
Kolbe electrolysis cell
For an abiotic electrochemical template, see the direct-air-capture (DAC) model in the lab.
Data the predictor needs
Two sources, same columns. His runs give clean within-study effects; the Kolbe literature gives spread across labs. Runs with low efficiency or that failed matter as much as good ones. For literature, MESSAI extracts the same columns with a source snippet for every value.
| Group | Column | Unit | His runs | Literature |
|---|---|---|---|---|
| Feed | concentration of each acid, C2–C8 | mol/L | Required | Extract |
| Feed | pH or degree of neutralisation; counter-ion | –, K⁺/Na⁺ | Required | Extract |
| Feed | synthetic mixture or real broth; impurities | text | Required | Extract |
| Cell | one- or two-chamber; membrane; electrode gap | text, mm | Required | Extract |
| Cell | anode material, surface and area | text, cm² | Required | Extract |
| Cell | temperature; solvent and supporting electrolyte | °C, text | Required | Extract |
| Operation | control mode; anode current density | mA/cm² | Required | Extract |
| Operation | cell voltage; anode potential with reference electrode | V, V vs ref | Required | Extract |
| Operation | charge passed; duration; run order on the same electrode | C, h, # | Required | If reported |
| Outcomes | products per compound (GC), incl. alkenes, alcohols, esters | mol | Required | Extract |
| Outcomes | residual acids over time (HPLC); O₂ in off-gas | mol/L, vol-% | Useful | If reported |
| Outcomes | CE per product class; energy per litre of fuel | %, kWh/L | We compute | Extract |
Papers behind this page
Every paper the protocol chart cites, with what it is cited for and where you can check it. Each row links the publisher and, if this site holds the paper, its corpus record.
12 works cited · 12 with a DOI · 12 in this corpus
| Paper | Cited for | Steps | Check it | In the MESSAI corpus |
|---|---|---|---|---|
| Pohl et al. 2026, ChemSusChemThe Concentration Determines Reaction Efficiency of Hetero‐Coupling of Bio‐Based Medium Chain Carboxylic Acids by Kolbe Electrolysis |
| Prepare the carboxylic-acid feed · Assemble a two-chamber (divided) cell · Assemble a one-chamber (undivided) cell · +1 more | PublisherCorpus record |
|
| Neubert et al. 2022, ChemSusChemHetero‐Coupling of Bio‐Based Medium‐Chain Carboxylic Acids by Kolbe Electrolysis Enables High Fuel Yield and Efficiency |
| Prepare the carboxylic-acid feed · Pass a set charge · Close the balances; calculate CE and energy | PublisherCorpus record |
|
| Rózsenberszki et al. 2025Upgrading Potential of Waste Fermentation-Based C4-C5-C6 Carboxylic Acid Mixtures using Electrochemical Oxidative Decarboxylation |
| Prepare the carboxylic-acid feed · Run galvanostatic electrolysis | PublisherCorpus record |
|
| Stang & Harnisch 2015, ChemSusChemThe Dilemma of Supporting Electrolytes for Electroorganic Synthesis: A Case Study on Kolbe Electrolysis |
| Choose the supporting electrolyte | PublisherCorpus record |
|
| Urban & Harnisch 2017, ChemElectroChemDeterioration of Aqueous n‐ Octanoate Electrolysis with Electrolytic Conductivity Collapse Caused by the Formation of n ‐Octanoic Acid/ n ‐Octanoate Agglomerates |
| Choose the supporting electrolyte | PublisherCorpus record |
|
| Olde Nordkamp et al. 2022, ChemCatChemStudy on the Effect of Electrolyte pH during Kolbe Electrolysis of Acetic Acid on Pt Anodes |
| Set neutralisation (pH) and counter-ion | PublisherCorpus record |
|
| Ashraf et al. 2024, ChemElectroChemThe Influence of Alkali Cations on the Performance of Pt Electrodes in Kolbe Electrolysis of Acetic Acid |
| Set neutralisation (pH) and counter-ion | PublisherCorpus record |
|
| Neubert et al. 2021, ChemSusChemPlatinized Titanium as Alternative Cost‐Effective Anode for Efficient Kolbe Electrolysis in Aqueous Electrolyte Solutions |
| Prepare the Pt anode | PublisherCorpus record |
|
| Ranninger et al. 2022, ChemSusChemOn‐line Electrode Dissolution Monitoring during Organic Electrosynthesis: Direct Evidence of Electrode Dissolution during Kolbe Electrolysis |
| Prepare the Pt anode | PublisherCorpus record |
|
| Drögemüller et al. 2023, ChemSusChemClosing the Gap: Towards a Fully Continuous and Self‐Regulated Kolbe Electrosynthesis |
| Control pH during electrolysis | PublisherCorpus record |
|
| Baumgarten et al. 2024, Chem. Ing. Tech.Continuous Kolbe Electrolysis Using an Electrochemical Microreactor, Product Separation, and Electrolyte Recycling |
| Control pH during electrolysis | PublisherCorpus record |
|
| Klocke et al. 1993, Chem. Ber.Electroorganie Synthesis, 55 [1] . Influences on the Selectivity of the Kolbe versus the Non‐Kolbe Electrolysis in the Anodic Decarboxylation of Carboxylic Acids |
| Run galvanostatic electrolysis | PublisherCorpus record |
|
Corpus read from this deployment on Nov 21, 2026.
“Values” counts the distinct facts the extractor read from the paper, not only the ones cited here: a value written twice (same name, value, unit, quote and conditions) counts once. A paper whose full text was never read has none; values read from open-access text count even when no PDF is stored.
Where each Kolbe paper stands
One row per paper in the Kolbe cohort, measured in the staging corpus on Oct 31, 2026. The status is the furthest step the paper has reached; the figures count every paper that has reached at least that step.
- Registered
- 55
- PDF stored
- 13
- Text read
- 9
- Values extracted
- 9
- Usable by a model
- 0
| Year | First author | Title | Status | Links |
|---|---|---|---|---|
| 2026 | Dana M. Pinson | From Waste to Value: Kolbe Electrolysis for the Sustainable Production of Fuels and Chemicals | Registered | Corpus recordDOI |
| 2026 | Jian Yang | Modulating hydrocarbon coverage on Mo-doped RuO₂ anodes enables energy-efficient and selective Kolbe electrolysis | Registered | Corpus recordDOI |
| 2026 | Max Pohl | The Concentration Determines Reaction Efficiency of Hetero‐Coupling of Bio‐Based Medium Chain Carboxylic Acids by Kolbe Electrolysis | Extracted42 facts | Corpus recordDOI |
| 2026 | Zhenan Jin | Tuning Kolbe electrolysis selectivity by chain length of fatty acids: Mechanistic insights and a synergistic mixed-feed strategy in an aqueous electrolyte | Registered | Corpus recordDOI |
| 2026 | Chen Deng | Valorization of Biomass-Derived Carboxylic Acids into Electrofuels and Chemicals via Kolbe Electrosynthesis | Registered | Corpus recordDOI |
| 2025 | Xinyan Zhang | Anion intercalation enables efficient and stable carboxylate upgrading via aqueous non-Kolbe electrolysis | Extracted63 facts | Corpus recordDOI |
| 2025 | Wenhua Zhou | Dopant-Induced Electron Localization Drives Direct Current Kolbe Coupling of Biomass-Derived Carboxylic Acids | Registered | Corpus recordDOI |
| 2025 | Yunfei Zhang | Efficient electrosynthesis of valeric acid from levulinic acid using indium nanoparticles | Registered | Corpus recordDOI |
| 2025 | Tamás Rózsenberszki | Upgrading Potential of Waste Fermentation-Based C4-C5-C6 Carboxylic Acid Mixtures using Electrochemical Oxidative Decarboxylation | Extracted42 facts | Corpus recordDOI |
| 2024 | Nils Baumgarten | Continuous Kolbe Electrolysis Using an Electrochemical Microreactor, Product Separation, and Electrolyte Recycling | Registered | Corpus recordDOI |
| 2024 | Luis F.M. Rosa | Electrolysis of medium chain carboxylic acids to aviation fuel at technical scale | Registered | Corpus recordDOI |
| 2024 | Clemens Taube | Kolbe Electrolysis of Long‐Chain Fatty Acids for Efficient Production of Bio‐Based Hydrophobic Paraffin Waxes | Registered | Corpus recordDOI |
| 2024 | Talal Ashraf | The Influence of Alkali Cations on the Performance of Pt Electrodes in Kolbe Electrolysis of Acetic Acid | Extracted34 facts | Corpus recordDOI |
| 2023 | Chong Wang | Amorphous RuO₂ Catalyst for Medium Size Carboxylic Acid to Alkane Dimer Selective Kolbe Electrolysis in an Aqueous Environment** | Registered | Corpus recordDOI |
| 2023 | Patrick Drögemüller | Closing the Gap: Towards a Fully Continuous and Self‐Regulated Kolbe Electrosynthesis | Registered | Corpus recordDOI |
| 2022 | Nils Kurig | 3D Printed Microreactors for the Continuous Non‐Kolbe Electrolysis | Registered | Corpus recordDOI |
| 2022 | Tongying Yuan | Efficient Electrocatalytic Reduction of Levulinic Acid to Valeric Acid on a Nanocrystalline PbO‐In₂ O₃ Catalyst | Registered | Corpus recordDOI |
| 2022 | Lin Lang | Enhancing the selectivity of hydrocarbons during the Kolbe electrolysis of biomass-derived short-chain carboxylic acids by anionic surfactants | Registered | Corpus recordDOI |
| 2022 | Katharina Neubert | Hetero‐Coupling of Bio‐Based Medium‐Chain Carboxylic Acids by Kolbe Electrolysis Enables High Fuel Yield and Efficiency | Extracted68 facts | Corpus recordDOI |
| 2022 | Johanna Ranninger | On‐line Electrode Dissolution Monitoring during Organic Electrosynthesis: Direct Evidence of Electrode Dissolution during Kolbe Electrolysis | Extracted32 facts | Corpus recordDOI |
| 2022 | Nils Baumgarten | Scalable Microreactor Concept for the Continuous Kolbe Electrolysis of Carboxylic Acids Using Aqueous Electrolyte | Extracted67 facts | Corpus recordDOI |
| 2022 | Margot Olde Nordkamp | Study on the Effect of Electrolyte pH during Kolbe Electrolysis of Acetic Acid on Pt Anodes | Registered | Corpus recordDOI |
| 2022 | Sihang Liu | Understanding the reaction mechanism of Kolbe electrolysis on Pt anodes | Registered | Corpus recordDOI |
| 2022 | Niklas Teetz | Upgrading Kolbe Electrolysis—Highly Efficient Production of Green Fuels and Solvents by Coupling Biosynthesis and ElectrosynthesisHidden from the corpus listing · Found by title or abstract, outside the curated set | DOI | |
| 2021 | Daniel Klüh | Kolbe Electrolysis for the Conversion of Carboxylic Acids to Valuable Products—A Process Design Study | Corpus recordDOI | |
| 2021 | Najoua Sbei | Organic synthesis via Kolbe and related non-Kolbe electrolysis: an enabling electro-strategy | Registered | Corpus recordDOI |
| 2021 | Katharina Neubert | Platinized Titanium as Alternative Cost‐Effective Anode for Efficient Kolbe Electrolysis in Aqueous Electrolyte Solutions | Extracted61 facts | Corpus recordDOI |
| 2020 | Nils Kurig | (Non-)Kolbe Chemistry Going with the Flow: The Continuous Electrolysis of Biogenic Acids | Registered | Corpus recordDOI |
| 2020 | F. Joschka Holzhäuser | (Non-)Kolbe electrolysis in biomass valorization – a discussion of potential applications | Registered | Corpus recordDOI |
| 2020 | Jérôme Meyers | Intramolecular Biradical Recombination of Dicarboxylic Acids to Unsaturated Compounds: A New Approach to an Old Kolbe Reaction | Registered | Corpus recordDOI |
| 2020 | Moritz C. Rehbein | Liquid–Liquid Equilibrium Data and Continuous Process Concept for the Electrosynthesis of Valeric Acid from Levulinic Acid | Extracted25 facts | Corpus recordDOI |
| 2020 | Yan Du | Synthesis of Valeric Acid by Selective Electrocatalytic Hydrogenation of Biomass-Derived Levulinic Acid | Registered | Corpus recordDOI |
| 2019 | Gang Yuan | Core–Shell Pt@Ir Nanothorns on Carbon Fiber Paper Electrodes for Carboxylic Acid Valorization via Kolbe Electrolysis | Registered | Corpus recordDOI |
| 2019 | Jérôme Meyers | Electrochemical conversion of a bio-derivable hydroxy acid to a drop-in oxygenate diesel fuel | Registered | Corpus recordDOI |
| 2019 | F. Joschka Holzhäuser | Electrochemical cross-coupling of biogenic di-acids for sustainable fuel production | Registered | Corpus recordDOI |
| 2019 | Zhu Xixi | Electrolyzing lactic acid in situ in fermentation broth to produce pyruvic acid in electrolysis cell | Registered | Corpus recordDOI |
| 2019 | Roel J. M. Bisselink | Identification of More Benign Cathode Materials for the Electrochemical Reduction of Levulinic Acid to Valeric Acid | Registered | Corpus recordDOI |
| 2019 | Gang Yuan | Kolbe Electrolysis of Biomass‐Derived Fatty Acids Over Pt Nanocrystals in an Electrochemical Cell | Registered | Corpus recordDOI |
| 2019 | Stefan Palkovits | The Role of Electrochemistry in Future Dynamic Bio‐Refineries: A Focus on (Non‐)Kolbe Electrolysis | Registered | Corpus recordDOI |
| 2018 | Yimeng Zhang | Electrochemical conversion of palmitic acid via Kolbe electrolysis for synthesis of n-triacontane | Registered | Corpus recordDOI |
| 2017 | Carolin Urban | Deterioration of Aqueous n‐ Octanoate Electrolysis with Electrolytic Conductivity Collapse Caused by the Formation of n ‐Octanoic Acid/ n ‐Octanoate Agglomerates | Registered | Corpus recordDOI |
| 2015 | Tatiane R. dos Santos | Electrochemistry for Biofuel Generation: Transformation of Fatty Acids and Triglycerides to Diesel‐Like Olefin/Ether Mixtures and Olefins | Registered | Corpus recordDOI |
| 2015 | Jiajie Xu | In-line and selective phase separation of medium-chain carboxylic acids using membrane electrolysis | Registered | Corpus recordDOI |
| 2015 | Carolin Stang | The Dilemma of Supporting Electrolytes for Electroorganic Synthesis: A Case Study on Kolbe Electrolysis | Registered | Corpus recordDOI |
| 2014 | Yang Qiu | Integrated electrocatalytic processing of levulinic acid and formic acid to produce biofuel intermediate valeric acid | Registered | Corpus recordDOI |
| 2013 | Le Xin | Electricity Storage in Biofuels: Selective Electrocatalytic Reduction of Levulinic Acid to Valeric Acid or γ‐Valerolactone | Registered | Corpus recordDOI |
| 2007 | Toshiki Tajima | Development of an Electrolytic System for Non-Kolbe Electrolysis Based on the Acid−Base Reaction between Carboxylic Acids as a Substrate and Solid-Supported Bases | Registered | Corpus recordDOI |
| 2006 | Hitoshi KURIHARA | Mixed-Kolbe Electrolysis Using Solid-Supported Bases | Corpus recordDOI | |
| 1999 | Yasushi ONO | Kolbe Electrolysis of Carboxylates on a Hydrophobic Platinum Electrode Composite-plated with PTFE Particles | Corpus recordDOI | |
| 1998 | Hiroki Fujiwara | Ultrasonic Effects on Electroorganic Processes. Part 10. Product‐Selectivity in Mixed Kolbe Electrolysis of Carboxylic Acids | Registered | Corpus recordDOI |
| 1996 | Michael Harenbrock | Electroorganic Synthesis, 62!! Anodic Heterocoupling (Mixed Kolbe Electrolysis) of Carbohydrate Carboxylic Acids with Alkanoic Acids to C ‐Glycosides | Registered | Corpus recordDOI |
| 1993 | Elisabeth Klocke | Electroorganie Synthesis, 55 [1] . Influences on the Selectivity of the Kolbe versus the Non‐Kolbe Electrolysis in the Anodic Decarboxylation of Carboxylic Acids | Registered | Corpus recordDOI |
| 1984 | P. F. Levy | Kolbe Electrolysis of Mixtures of Aliphatic Organic Acids | Registered | Corpus recordDOI |
| 1979 | H.J. Schäfer | Recent synthetic applications of the Kolbe electrolysis | Registered | Corpus recordDOI |
| 1959 | William B. Smith | The Kolbe Electrolysis as a Source of Free Radicals in Solution | Registered | Corpus recordDOI |
A corpus record opens only on a deployment that holds the paper; staging and production differ. “Usable by a model” counts values that pass the model’s row rules, and none is harmonised yet.
Our to-dos (MESSAI)
In priority order. Estimates are rough and assume one person part-time.
- Kolbe parameter set~2–3 daysNot started
Canonical names and units for the ten parameters and seven outcomes in the map (proposed:
carboxylate_total_concentration,degree_of_neutralisation,anode_current_density,kolbe_ce_fuel,specific_energy_fuel_volume…), sent to Max to check. Fields that get filtered become database columns. Everything else depends on this. - Kolbe literature corpus~1–2 weeksNeeds Max’s list
The 20–50 papers he picks plus a search for the rest, fetched, extracted with a Kolbe-specific spec, harmonized. On 7 Oct 2026 the corpus held one Kolbe paper. By 11 Oct 2026, 54 more had been added to the staging corpus, 47 of them with an abstract, and 434 values had been extracted from the 9 whose open-access full text could be fetched; none is yet harmonised or reviewed. The other 45 have no extracted values; on 31 Oct 2026 the open-access PDFs of 11 papers were stored in MESSAI, 3 of them among those 45. Which of the rest are paywalled is not recorded, so no count is given. This is the main source of cross-study evidence. What the deployment you are on holds now is counted under “How it works”.
- Encode his paper as within-study effects~2 days after dataWaiting on data
Concentration, composition and cell-design effects from his runs, checked against the trends he reported.
- Parameter map v1 with numbers~1 week after the two aboveLater
Replace the evidence states on this page with effect sizes, intervals and the number of studies behind each cell. Add the links to MESSAI’s knowledge graph.
- Predictor v1~2 weeksLater
Hierarchical model with study as a random effect, conformal intervals, ranked drivers, and a leave-one-study-out validation report including where it fails.
- Interactive prediction page~3–4 daysLater
/collab/pohlin English and German: parameter inputs, prediction with interval, drivers, and the live parameter map. - Next-measurement ranking~1 weekLater
Rank the empty cells by how much measuring them would narrow the predictions Max cares about, and turn the top ones into suggested runs.
- Supporting fixes~2 days + roadmapNot started
Product-rate unit harmonization and reference-electrode conversion to SHE; the measurement upload route (roadmap C3-M1, planned 21 Nov) so lab runs land with “measured” provenance.
Deliverables to UFZ
L0 is the day his literature list arrives, T0 the day his run table arrives. Nothing is promised before the data exists.
- This weekParameter set and run-table template
The proposed Kolbe parameters and units for him to correct, the CSV template, and the literature request.
- L0 + 2 weeksKolbe evidence table
What the literature has measured, under which conditions, with source snippets. Shows at a glance which parameters the field has studied and which it hasn’t.
- T0 + 1 weekHis paper encoded
Within-study effects from his runs, reproducing the trends he published, or a note on what doesn’t reconcile.
- T0 + 4 weeksParameter map with numbers
Effect size, interval and number of studies for each filled cell; empty cells ranked by importance.
- T0 + 6 weeksPredictor v1 and validation
Predictions with intervals and drivers, tested by leaving one study out at a time.
- T0 + 8 weeksInteractive page, suggested runs, methods note
The prediction tool in MESSAI, the top suggested measurements, a co-authored methods note draft, and the dataset on Zenodo with a DOI if UFZ agrees.
What we ask of Max and UFZ
- The run table behind the 2026 paper, plus unpublished runs, especially any that vary current density, pH, temperature or electrode, and the failed ones.
- A list of 20–50 Kolbe and electrobiorefinery papers whose data he considers reliable enough to pool with his own, with any he would exclude and why.
- A check of the parameter set and of the “known direction” cells in the map.
- 30 minutes every two weeks to review results, and agreement on data terms and authorship.
Open questions for Max
- Which parameters besides concentration and composition did he vary, in the paper or in unpublished runs? Current density, pH, temperature, electrode?
- Which cells in the map would change his next experiment if we could predict them?
- Does he suspect interactions? For example, does the ≈0.1 mol/L threshold move with current density or chain length?
- Are the “known direction” cells right for his system? Platinum versus carbon, high current density favouring Kolbe.
- When we combine his runs with published Kolbe results from other labs, which papers or groups are reliable enough to include, and which should we leave out (different electrode, missing conditions, doubtful yields)?
- Real broth: how do impurities and real titres compare with synthetic mixtures?
- Would he be willing to run a small number of experiments that the model proposes, feed the results back so the model updates, and repeat? (an iterative design-of-experiments cycle; each round narrows the uncertain cells in the map)