Funding for strategically important "national services" in chemistry will no longer be ring-fenced by the Engineering and Physical Sciences Research Council.
National services will now have to compete with all other research grant applicants for "responsive mode" funding rather than competing with each other for "directed mode" funding.
A high-profile casualty of the new funding system is the National Service for Computational Chemistry Software (NSCCS) at Imperial College London, which is to close next March.
The service provides central computer facilities and access to software and training for UK chemistry researchers and has been in existence, in various forms, for nearly 40 years.
In a letter posted on the NSCCS website, Michael Robb, professor of chemistry at Imperial, says: "Our grant went to a responsive mode panel on 1 July and was ranked 14th. The EPSRC funded only seven out of 64 grants from that meeting. Our proposal was judged alongside all the other proposals on the basis of scientific output, but with no direction from the EPSRC with regards to its strategic importance."
An online petition condemning the decision has attracted 179 signatures. Signatory David Clary, a professor of chemistry at the University of Oxford and president of the Faraday Division of the Royal Society of Chemistry, said: "The NSCCS has been highly cost-effective and is significantly enhancing the research of hundreds of chemists and molecular scientists. He urged the EPSRC to reconsider the decision.
Other academics have pointed out that responsive-mode funding has already been depleted and if national services are also "rolled in", investigator-driven science will suffer.
Other EPSRC national chemistry services include the National Crystallography Service at the University of Southampton, the National Mass Spectrometry Service Centre at the University of Swansea, and the Electron Paramagnetic Resonance Service at the University of Manchester. The Manchester grant is up for renewal this year.
The EPSRC decided to withdraw funding from the Chemical Database Service at Daresbury in March 2007, but the providers of the original service secured a contract for a reduced service.
The EPSRC said that funding of national services in chemistry through responsive-mode bids was consistent with practice in other areas within the EPSRC.
A spokeswoman said: "Within the current physical sciences portfolio we believe it important to tension the funding of such services against the funding of the high-quality research projects we receive.
"The decision not to fund the NSCCS was taken on the advice of independent scientists through the peer review system.
"The proposal was judged alongside others submitted and it did not rank highly enough to be funded. Decisions are based on the criteria of supporting research of the highest quality.