This paper explores emerging infrastructure that will enable the discovery and re-use of scientific data.  
One of the outcomes of e-science is the increasing amount of data being produced in digital form with 
potential for re-use. In order to facilitate re-use such data needs to be managed both in terms of 
curation and preservation. An essential aspect of curation is enabling the discovery of that data. 
As publication, discovery, and re-use become recognised as aspects of the life-cycle of research data, 
then the overlap between management of raw and derived data 
(datasets produced by experimentation and data collection) and the management of bibliographic data 
(in particular metadata describing digital resources) become obvious.  There are overlaps as regards 
supporting technologies e.g. OAI-PMH, and overlaps between services that might be offered e.g. linking 
datasets and journal articles. There are increasing commonalities in the concerns of those developing 
digital libraries and those creating and managing research data. In the UK the eBank project is 
funded by the Joint Information System's Committee (JISC) as part of its 
Semantic Grid programme in order to demonstrate how dissemination of research data can be made easier.
Currently once a research experiment is finished the initial dissemination may be via a letter or 
communication, followed later by a more detailed explanation in a full paper. eBank will 
demonstrate how data might be 'published' directly in open archives and subsequently linked to 
final peer-reviewed journal articles. As the availability of journal article e-prints increases 
users would benefit a service linking back to the raw or processed data. The objective is to make 
'publishing' and discovery of research data easier, in order to facilitate re-use of data in 
further experiments.  The eBank project represents an opportunity to bring together the digital 
library community with the Grid / computer science community. An eBank demonstrator is being 
developed within the chemistry domain (crystallography) with a view to assessing the feasibility of a 
generic approach across other disciplines. The Combechem Project at the University of Southampton 
offers an ideal research testbed since large quantities of heterogeneous data are being generated 
including electronic lab books, crystallography data and physical chemistry data. 
The demonstrator shows how an e-prints repository of peer-reviewed journal articles can include 
links to associated metadata describing research data. The demonstrator is using the technical 
architecture currently being deployed within the ePrints UK Project. The ePrints UK architecture 
supports the harvesting of metadata from e-print archives in UK academic institutions and elsewhere 
using the OAI Protocol for Metadata Harvesting (OAI-PMH). The eBank project will augment this 
work by harvesting metadata desribing research data from the University of Southampton institutional 
'e-data repository'. Research data is deposited in an OAI compliant institutional repository at the 
University of Southampton, and metadata describing the data holdings made available for harvesting. 
The workflow of capturing data from experimentation through the various stages of derived data is being 
undertaken as part of the complementary Combechem project. An enhanced version of the e-prints.org 
software accommodates self-archiving by chemists. The ePrints UK project harvests the metadata and makes 
appropriate links to published journal articles. 
The paper focuses in particular on the eBank architecture linking the research data repository at the 
University of Southampton to a central ePrints UK search service hosted at the University of Bath. 
The paper also gives an account of the proposed data flow with associated metadata schemas, suggesting a 
simple generic schema for a high level description of research data.