From Lab Bench to Pilot Plant: A Milestone Roadmap

An artificial photosynthesis business plan reads very differently to a grant panel and a private investor if it doesn't show a credible path from bench chemistry to something a customer can actually buy. Most reviewers will look for evidence you understand Technology Readiness Level (TRL) staging, not just a market-size slide. The rough sequence below is what we build into bespoke plans for solar-fuels and catalysis ventures.

TRL is a nine-point scale originally developed by NASA and now standard across DOE, NSF, and Innovate UK grant applications: TRL 1-3 covers basic research and lab-validated concepts, TRL 4-6 covers component and system validation in an increasingly realistic environment, and TRL 7-9 covers a system proven in an actual operational setting through to full commercial deployment. If your plan doesn't state a current TRL and a target TRL for the funding round you're raising, most experienced reviewers will ask for it in the first meeting anyway — so it's worth putting the number on the page rather than leaving it as an implied assumption.

• Months 0–3 (TRL 2–3): Freedom-to-operate patent search against existing artificial photosynthesis IP holders; lock in a catalyst chemistry direction (e.g. a BiVO₄ photoanode with a cobalt-phosphate co-catalyst, a perovskite light absorber, or a metal-organic framework photocatalyst); file a first SBIR Phase I or Innovate UK smart grant application.
• Months 3–9 (TRL 3–4): Build and test a bench-scale reactor; benchmark solar-to-fuel efficiency against the published 5–10% typical range (with best-in-class lab results above 20%) so your own numbers can survive scrutiny.
• Months 9–15 (TRL 4): Secure SBIR Phase II or Innovate UK scale-up funding; hire two to three specialist scientists in photoelectrochemistry or heterogeneous catalysis — the single biggest cost line at this stage.
• Months 15–24 (TRL 5): Build a 10–100 square metre pilot photocatalytic panel reactor, following the flat-panel architecture Japan's ARPChem/NEDO consortium demonstrated; file the relevant environmental permit application in parallel, since approval timelines run in parallel with construction, not after it.
• Months 24–36 (TRL 5–6): Run continuous outdoor operation for 12+ months to generate real performance data (the standard NEDO's demonstration used to validate its numbers publicly); sign a first contract-research agreement or letter of intent with an industrial partner.
• Months 36–48+ (TRL 6–7): Raise a Series A specifically earmarked for a hectare-scale demonstration plant; begin negotiating a first product offtake agreement for hydrogen, methanol, or syngas.

The mistake we see most often in early plans: founders write this section like a normal small business launch checklist ("register the company, hire staff, open the doors") instead of a staged TRL roadmap. A grant assessor or deep-tech investor reads the absence of TRL language as a signal the founding team hasn't operated in this funding environment before.

Notice that every stage above is gated by a funding event, not a calendar date. A realistic plan states the milestone that releases the next tranche of money — "Series A closes once 12 months of continuous outdoor pilot data is available," not "Series A in Month 30." Grant panels and investors both read staged, milestone-gated roadmaps as a sign the team understands its own capital efficiency, whereas date-only roadmaps read as guesswork dressed up as planning.

It's also worth stating plainly which stage your venture is at today. A plan that opens with "we are currently at TRL 3, seeking funding to reach TRL 5" is easier for a reviewer to underwrite than one that implies the whole roadmap above is still theoretical. Reviewers who fund dozens of similar applications a year can tell within a paragraph whether a team has actually built anything yet.

What It Actually Costs to Fund an Artificial Photosynthesis Venture

This is not a $50,000 small-business launch. Getting from a lab proof-of-concept to a pilot-scale demonstration typically requires $250,000 to $20 million in the US (£200,000 to £15 million in the UK), spread across grant funding, seed equity, and eventually a Series A. The wide range reflects how far along the TRL curve you're trying to fund in a single plan — a bench-scale proof of concept is a very different ask to a hectare-scale demonstration plant.

Cost Breakdown

Unlike a retail or services business, where the biggest line item is usually premises or inventory, the cost structure here is dominated by two things: specialist people and one-off capital equipment. A single photoelectrochemistry PhD hire can cost more per year than an entire storefront fit-out in most other BPT categories on this site, and the pilot reactor itself is a bespoke piece of equipment, not something you can buy off a shelf and negotiate down.

• Lab-scale R&D + proof of concept (catalyst synthesis, bench reactor): $250K–$1.2M (£200K–£950K)
• IP filing, freedom-to-operate search, patent prosecution: $40K–$250K (£30K–£200K)
• Pilot-scale photocatalytic panel reactor (10–100 sq m demonstration unit): $500K–$4M (£400K–£3.2M)
• Environmental permitting (US state + EPA / UK Environment Agency): $15K–$150K (£12K–£120K)
• Specialist scientific hires (photoelectrochemistry, catalysis PhDs): $180K–$900K/yr (£150K–£720K/yr)
• Demonstration-to-commercial pilot plant (hectare-scale panel array): $4M–$15M+ (£3.2M–£12M+)

Funding Routes

The capital stack for this niche is grant-first, not loan-first. In the US, the DOE's SBIR programme funds over $300 million annually across 60+ clean-energy topics, and the Department of Energy separately committed $100 million over five years to solar-fuels research, split between the Caltech-led Liquid Sunlight Alliance (~$60M) and the UNC Chapel Hill-led CHASE programme (~$40M). The NSF SBIR/STTR Chemical Technologies track offers up to $305,000 in non-dilutive funding for six-to-18-month R&D projects. Research on ARPA-E-funded cleantech startups found that winning an early federal innovation award roughly doubles the probability of subsequently closing a venture capital round — which is why we build the grant-to-equity bridge explicitly into every bespoke plan, rather than treating grants and VC as separate fundraising tracks.

In the UK, Innovate UK smart grants are the closest equivalent, typically awarding £100,000–£2 million for early-stage clean energy R&D, usually requiring private co-investment. Smart grants generally cover 60-70% of eligible project costs for a small business, meaning a £340,000 award still requires the founding team to source the remaining third from angels, a university proof-of-concept fund, or their own capital — a detail worth modelling explicitly rather than assuming the grant covers everything. UK-based ventures collaborating with an EU research partner may also be eligible for Horizon Europe cluster funding under the clean energy pillar, though association status has shifted in recent years and should be checked at application time rather than assumed.

If your venture is closer to electrolysis-based hydrogen production than photocatalysis, it's worth comparing your capital stack against our green hydrogen business plan template, which covers a slightly different (and slightly more mature) funding environment. Founders working on carbon capture as the CO₂ feedstock side of an artificial photosynthesis process may also find our carbon capture and sequestration business plan template useful for the upstream half of the value chain.

Core Technology & Lab Stack

Investors reading a deep-tech plan will look for specifics, not a vague reference to "clean technology." A credible artificial photosynthesis plan names the actual materials, hardware, and software the team is working with.

Catalyst & Materials

• BiVO₄ (bismuth vanadate) photoanodes — one of the most widely studied light-absorbing materials for water splitting
• Cobalt-phosphate (Co-Pi) co-catalysts — used to accelerate the oxygen-evolution half-reaction
• Perovskite-based light absorbers — higher theoretical efficiency, but with stability challenges under continuous outdoor operation
• Metal-organic framework (MOF) photocatalysts — an emerging materials class for CO₂-reduction pathways

Reactor Hardware

• Photocatalytic panel reactors — flat-panel arrays, the architecture behind the ARPChem/NEDO 100 sq m demonstration built with the University of Tokyo, Fujifilm, TOTO, Mitsubishi Chemical, Shinshu University and Meiji University
• Photoelectrochemical (PEC) tandem cells — stacked light absorbers designed to capture a broader slice of the solar spectrum
• Gas separation membranes — required wherever hydrogen and oxygen are generated in the same reactor volume, for safety as much as product purity

Modelling, Simulation & Analytics

• COMSOL Multiphysics — reactor-level mass transport and light-absorption modelling
• DFT packages (e.g. VASP, Gaussian) — computational catalyst screening before committing lab time to synthesis
• Aspen Plus — process-level techno-economic modelling once you move past bench scale
• Gas chromatography (GC) and a potentiostat — the baseline analytical kit for quantifying hydrogen/oxygen/CO₂ output and electrochemical performance
• Google Patents / Justia Patents — free tools for an initial freedom-to-operate screen before commissioning a paid FTO search

None of this needs to be exhaustive in the plan itself, but naming the specific materials and tools your team is actually using signals technical credibility in a way that generic language about "proprietary technology" does not.

The choice between these materials isn't cosmetic — it determines your cost base and your regulatory exposure at the same time. A BiVO₄/Co-Pi system is comparatively cheap and well-characterised in the literature, which makes it easier to defend efficiency claims to a grant panel, but it currently tops out below the highest lab efficiencies reported for perovskite-based tandem cells. Perovskite absorbers can push solar-to-fuel efficiency higher, but introduce lead-content and long-term stability questions that a reviewer familiar with the photovoltaics industry will ask about directly. MOF photocatalysts are the least mature of the four but are the most active area of current CO₂-reduction research, which can be a genuine differentiator in a grant application if your team has real expertise there rather than a passing interest.

On the hardware side, a flat-panel reactor (the ARPChem approach) is simpler to permit and cheaper to build per square metre than a tandem PEC cell, but a tandem cell can capture a wider slice of the solar spectrum and therefore reach higher efficiency per unit area. Most early-stage plans we review pick the flat-panel route first, specifically because it's the architecture with the most publicly available performance data to benchmark against — which matters when your entire pitch rests on convincing someone your projected numbers are realistic.

Permits, Patents & Regulatory Requirements

United States

• State environmental permit + EPA oversight for chemical process emissions ($15K–$150K in permitting and consultant fees, 6–18 months)
• NFPA 2 Hydrogen Technologies Code compliance for on-site storage and handling
• Bayh-Dole Act compliance if the venture's core IP originated in federally-funded university research — this affects who owns what, not whether you can operate
• OSHA Process Safety Management programme for reactive chemical processes

United Kingdom

• Environmental permit from the Environment Agency — hydrogen production is classified as a Part A(1) 4.2(a)(i) inorganic chemicals activity; a simplified "standard rules" permit is available for low-impact installations, taking roughly 3–6 months
• COMAH (Control of Major Accident Hazards) assessment from the HSE if storing hydrogen above threshold quantities
• Planning permission from the local authority for an industrial demonstration site (£500–£5,000 in fees, 8–13 weeks for standard applications)

Japan (Illustrative Other Jurisdiction)

Japan's High Pressure Gas Safety Act governs hydrogen storage and handling. Large public-private demonstration projects — such as the ARPChem/NEDO 100 square metre photocatalytic panel system — typically operate inside a METI/NEDO-sponsored consortium structure rather than through a stand-alone commercial permitting process, which is worth understanding if you're evaluating a joint-venture or licensing route into the Japanese market.

One freedom-to-operate note worth building into your own IP section: organisations including FUJIFILM Corporation and the Japan Technological Research Association of Artificial Photosynthetic Chemical Process (ARPChem) hold a dense patent portfolio in this exact chemistry space. A patent search early — even a free one using Google Patents or Justia before commissioning a paid freedom-to-operate opinion — is standard due diligence, not a nice-to-have.

Beyond the FTO search, most bespoke plans we write for this niche include a short IP strategy paragraph answering three questions an investor will ask regardless of whether it's stated explicitly in the plan: does the venture own its core IP outright, license it from a university under an exclusive or non-exclusive arrangement, or build on top of an expired patent (the earliest artificial photosynthesis patents, filed in the early 2010s, are now approaching the end of their 20-year term in some jurisdictions, which occasionally opens a specific chemistry to new entrants). Getting this wrong — claiming ownership of IP that's actually licensed with restrictive terms — is one of the fastest ways to fail investor diligence once term sheets are on the table.

Revenue Model & Unit Economics

Revenue in this niche arrives in stages, and a plan that shows only one revenue line (product sales) will read as naive to anyone who has funded a deep-tech venture before. The realistic sequence is: non-dilutive grant income first, then contract-research fees from industrial partners, then IP licensing royalties once the catalyst or process IP matures, and only later, product sales under an offtake agreement (hydrogen, methanol, or syngas).

Margin doesn't really exist as a concept until you reach that last stage — most ventures are pre-revenue for 3–7+ years while operating at TRL 3–6. Once a venture reaches offtake scale, gross margin depends entirely on whether production cost undercuts the current $6–$12 per kilogram hydrogen benchmark reported across photoelectrochemical and solar-electrolysis systems.

Worked Example (Illustrative Model, Not Verified Operating Data)

Take a 100 square metre photocatalytic panel system — the scale Japan's ARPChem/NEDO consortium actually demonstrated outdoors. At roughly 5 kWh per square metre per day of solar irradiance and a 10% solar-to-hydrogen efficiency (mid-range for deployed photoelectrochemical systems), that panel yields approximately 1.5 kg of hydrogen per day. Scaled to a 1-hectare (10,000 sq m) commercial pilot at the same efficiency, that becomes roughly 150 kg per day.

Against a modelled production cost of $9–$11/kg (the mid-point of the $6.1–$12.1/kg range reported for PEC and PV-electrolysis systems), a 1-hectare pilot sold under a green-premium offtake agreement could generate roughly $360,000–$510,000 in annual revenue against a comparable cost base. That thin, sometimes negative, margin at pilot scale is exactly why most ventures in this space stay grant-funded until either panel efficiency improves or offtake pricing does — and it's a number worth stating plainly in your own plan rather than glossing over.

Additional near-term revenue for many teams comes from contract research for industrial partners (chemical manufacturers wanting a cleaner feedstock route) and licensing fees once a catalyst formulation is patent-protected — both of which can fund operations well before a first offtake agreement is signed.

As a rough guide (composite estimates, not published deal terms for any named company), early contract-research agreements in adjacent catalysis and cleantech spaces typically run $50,000–$400,000 per project, scoped around a specific technical deliverable (a characterised catalyst sample, a validated efficiency benchmark, a feasibility report) rather than an open-ended retainer. IP licensing royalties for early-stage, unproven process chemistry typically fall in a 2–8% of net sales range once a licensee is generating product revenue, often paired with a smaller upfront or milestone-based licensing fee to compensate for the licensee taking on scale-up risk. None of these figures should be presented in your own plan as guaranteed outcomes — they're a starting point for a negotiation, and your financial model should show what happens to runway if licensing income arrives a year later than planned, which is the single most common variance we see against first-draft founder forecasts in this space.

The two biggest levers in any version of this model are solar-to-fuel efficiency and offtake price, and they move in opposite directions to the investor's benefit and your own. A founder who assumes 20% efficiency (a lab best-case, not a deployed-system average) and a $14/kg offtake price (the high end of the green-premium range) will show a business that looks profitable years earlier than one built on the deployed-system average of 5-14% efficiency and a $10/kg offtake price. Both numbers are defensible in isolation — the discipline is in choosing one consistent scenario, stating the assumption plainly, and showing a downside case alongside it rather than presenting only the optimistic version.

What the Market Data Actually Shows

Market-size estimates for artificial photosynthesis vary more than almost any other niche we cover — which is itself a useful thing to explain in your own plan, because an investor will notice the discrepancy if you cite just one source. P&S Market Research sizes the global market at $80.8 million in 2024, growing at a 15.2% CAGR to reach $248.0 million by 2032. Global Growth Insights puts 2025 at $143.09 million, rising to $171.91 million in 2026. Precedence Research projects $317.96 million by 2034, while Research Nester forecasts $896.37 million by 2035 at a 20.14% CAGR.

The spread exists because there's no dedicated NAICS or SIC code for this activity, and almost no company yet books revenue purely from artificial photosynthesis — most figures are built from adjacent hydrogen, catalysis, and carbon-utilisation markets, not from audited company financials. When you write your own market section, cite a range and explain why, rather than picking whichever single number looks most impressive.

Rather than picking a side, the more useful move in your own plan is to state what the four estimates agree on: double-digit CAGR (roughly 15–20% depending on the firm), and a market still measured in the hundreds of millions rather than billions — several orders of magnitude smaller than the wider hydrogen or renewable-energy markets it sits inside. That framing is more defensible under investor questioning than quoting the single largest number you can find.

Who's Actually Building This

Large-company activity in this space includes Siemens Energy, Panasonic, ENGIE, Toshiba, Fujitsu, Mitsubishi Chemical, Toyota Central R&D Labs, and Evonik. Among independent ventures, SunHydrogen Inc (formerly HyperSolar) develops a nanoparticle system designed to mimic photosynthesis and separate hydrogen from water, with a stated intent to license the technology rather than build production plants itself. Twelve, a Bay Area carbon-transformation company, closed a $645 million funding round led by TPG Rise Climate to build its AirPlant One facility in Moses Lake, Washington, converting captured CO₂ into E-Jet fuel with a claimed lifecycle emissions reduction of up to 90% versus fossil jet fuel.

On the research side, the Joint Center for Artificial Photosynthesis (JCAP) at Caltech, led by principal investigator Nathan Lewis, has operated since 2011. The Caltech-led Liquid Sunlight Alliance (LiSA) and the UNC Chapel Hill-led CHASE programme split the DOE's $100 million solar-fuels commitment roughly $60 million and $40 million respectively. In Japan, the ARPChem consortium (funded through NEDO, working with the University of Tokyo, Fujifilm, TOTO, Mitsubishi Chemical, Shinshu University and Meiji University) built and ran the 100 square metre outdoor demonstration panel referenced throughout this guide — still the most-cited real-world deployment at this scale.

Three Layers of Competition, Not One

A credible competitor-mapping section for this niche should separate three distinct types of player, because they compete on different things and pose different risks to a new entrant.

Most first-draft plans we see only address the first layer — naming Siemens or Mitsubishi as "the competition" — and miss that the more immediate competitive risk is often a better-funded venture-backed independent working the exact same catalyst chemistry, or a consortium publishing a result that undercuts your own efficiency claims before you've raised your next round.

More Questions Founders Ask

What's the difference between artificial photosynthesis and green hydrogen electrolysis?
Green hydrogen electrolysis uses electricity (often from solar or wind panels) to split water in a separate electrolyser. Artificial photosynthesis systems typically combine light capture and water splitting in a single integrated device (a photocatalyst or photoelectrochemical cell), skipping the intermediate electricity step. In practice, the cost comparison is close: roughly $11.4/kg for a 9.8%-efficient PEC system versus $12.1/kg for panel-plus-electrolysis, or as low as $6.1/kg for grid-supplemented electrolysis.

Which universities and labs are leading this research?
Caltech (via JCAP and the Liquid Sunlight Alliance), UNC Chapel Hill (via CHASE), and the University of Tokyo and Shinshu University in Japan (via the ARPChem/NEDO consortium) are among the most consistently cited research programmes. If your plan references academic collaborators, these are the institutions an investor will expect to recognise.

Who is actually funding this space right now?
Government agencies (the US DOE, NSF, and Japan's NEDO) fund the earliest research stages. Private capital arrives later and in large amounts once a venture reaches demonstration scale — Twelve's $645 million round, led by TPG Rise Climate, is the clearest example of what late-stage capital looks like in this niche, though it's a scale most founders will spend years working toward.

Is a university spinout the right structure for this kind of venture?
Often, yes. Many founders in this space license core IP from a university tech transfer office rather than owning it outright from day one. That arrangement needs to satisfy investors that the licence is broad, long, and secure enough to build a company on — a weak or narrow licence is one of the more common reasons deep-tech spinout plans get rejected at diligence.

How long before an artificial photosynthesis venture can actually sell a product?
Based on the funding and demonstration timelines referenced throughout this guide — a DOE programme running five years, a NEDO demonstration validated over a full year of continuous operation, and a typical seed-to-Series-A gap of 18–36 months — most credible plans put first commercial product revenue somewhere between year 4 and year 8 after founding, not year 1 or 2. A plan that promises earlier revenue without naming the specific offtake customer and price already agreed will read as optimistic rather than ambitious to an experienced reviewer.

Sample Business Plan Preview

Here's an extract from the kind of plan our team writes for solar-fuels and deep-tech ventures — so you can see exactly what you'll get:

What's in the Template

Every Avvale business plan template includes these sections, pre-structured for your industry. For a deep-tech venture like this one, the structure is deliberately different to our standard small-business template — it's built to satisfy a grant assessment panel and a private investor reading the same document, which is a harder writing problem than most business plans solve for.

• Executive Summary — Your venture at a glance, written to hold a grant assessor's or investor's attention in the first 60 seconds
• Company Overview — Legal structure, IP/licensing position, and founding story
• Technology & TRL Roadmap — Where the science stands today and the staged path to demonstration scale
• Industry Analysis — Market size (with a defensible range, not a single cherry-picked number), key players, and regulatory requirements
• Competitor & IP Analysis — Freedom-to-operate positioning and how you differentiate from adjacent players
• Funding Strategy — The grant-to-equity bridge, with named programmes and realistic amounts
• Operations Plan — Facility, permitting timeline, and key scientific hires
• Management Team — Founder bios, scientific advisory board, and key hires planned

The optional Financial Forecast add-on (included in our $300/£250 and $1,000/£800 packages) provides a 5-year Excel model with cash flow, break-even analysis, and capital requirements staged by TRL milestone rather than a flat monthly burn rate — a structure most generic financial-model templates don't offer.

Frequently Asked Questions