Carbon Removal Investment Pipeline

What this research found

A venture-style screen of the carbon dioxide removal (CDR) field pulled together $176.3M of federal grant awards, 15 USPTO patent filings and company funding records to find early-stage investment targets. Every grant and patent was sorted into a technology category and given an estimated technology readiness level (TRL), and the ranking produced a five-company shortlist spanning ocean alkalinity enhancement, enhanced weathering and direct air capture, alongside 19 named researchers as a founder and hiring pipeline.

• Federal money is concentrated in technology-agnostic programs. Of the $176.3M in grants analysed, $142.4M — 81% — sits in broad multi-technology awards such as a $100M Department of Energy pilot program and ARPA-E's $36M SEA-CO2, rather than in any single CDR approach.
• Direct air capture is the most commercially active category, holding 7 of the 15 patents and the highest average TRL at 4.64. Ocean alkalinity enhancement is the earliest-stage at TRL 2.25, despite drawing $26.4M in grants across just 2 awards.
• Enhanced weathering shows patent activity out of proportion to its public support: 3 patents against $2.0M of grant funding. Biomass and mineralization has 2 patents and no grant funding at all, which the analysis flags as a funding gap.
• Within the same category, grant-funded work scored more mature than patented work — TRL 6.0 against 3.86 for direct air capture, and 5.5 against 2.0 for enhanced weathering.
• The shortlist pairs two stealth or pre-seed ocean CDR companies at TRL 2, Running Tide Technologies and Ocean-Based Climate Solutions, with three Series A companies at TRL 5: Lithos Carbon ($6.3M, 2022), UNDO Carbon ($12M, 2023) and Noya Labs ($11M, 2023, led by a16z). Average TRL across the five is 3.8.
• Of 19 ranked talent targets, 3 already hold startup roles — including Ebb Carbon co-founder and CTO Matthew Eisaman and ZeoDAC chief executive Christopher Jones — leaving 16 academics with no company affiliation, averaging 3.8 papers each in the dataset.

How it was done

Fourteen grant records from NSF, the Department of Energy and ARPA-E covering 2015 to 2024 were cleaned and combined with 15 USPTO patent filings, and each of the 29 records was assigned to one of four technology categories by keyword matching on its title and abstract. A second keyword pass placed each record in a low, medium or high readiness band, and funding, patent counts and average TRL were then aggregated by category to separate patent-heavy from grant-heavy sectors. Company stage and funding came from Crunchbase, PitchBook and company websites; candidates were scored to favor stealth and pre-seed over later rounds, capped at two companies per technology, and filtered to TRL 2–6. Researcher momentum was scored from publication and citation counts to build the talent list, and the results were written up as an investment report with five deal sheets plus a 12-slide committee deck.

Data sources

• NSF, Department of Energy and ARPA-E grant awards, 2015–2024 — 14 records totaling $176.3M
• USPTO patent filings, 2015–2024 — 15 records
• Crossref, OpenAlex and Semantic Scholar publication records
• Crunchbase, PitchBook and company websites for funding stage and team data

Outputs produced

• raw_grants.json

  Raw grant data from NSF API and web search (14 grants totaling $176.2M)

• grants_cleaned.csv

  Structured CSV with Awardee, Title, Abstract, Amount, Date, PI, Agency columns

• 01_search_grants.py

  NSF Awards API search implementation script

• 02_enhance_grant_data.py

  CDR relevance filtering and data enhancement script

• patents_y02c.json

  15 CDR patents (CPC Y02C) with inventors, assignees, publication dates

• publications_cdr.json

  149 publications (42 CDR-relevant) with authors, citations, affiliations, DOIs

• cross_reference_matches.csv

  4 direct grant-patent matches with full context and commercial signals

• orphan_patents.csv

  1 university patent identified as potential commercialization target