• Understand the reasons for the current low use of on-line LC
• Identify industrial applications with good payback

• Identify areas of weakness and "close the gap"

Achievements

A review of the literature on on-line LC was compiled, that also included comments from previous researchers on the reasons for past failures in process applications of LC. The main conclusion was that sampling and sample delivery were the major limitations of on-line applications of LC. Hence, the project focussed on improved methods of sampling and used designs of sampling probes from Zeneca as the starting point of the revised project.

The probes from Zeneca (Mk 1 and Mk 2) were further evaluated and found to have some deficiencies. This led to the design of two new probes (Mk 3 and Mk 4).

The Mk 4 probe has been constructed, is under evaluation and continues to be developed.

The Mk 3 probe has been fabricated and will be constructed soon.

As part of the implementation of the Mk 4 probe, reprogramming of the control system was necessary to provide the required flexibility.

Deliverables

• Two new sampling probes designed and fabricated
• IMB report on On-line LC ( 99/P7/1)
• Project report on designs of Mk 3 and Mk 4 sampling probes
• Project report on fast LC analysis strategies (< 2 min) with application to batch reactions

Objectives versus Achievements

• Reasons for low use of on-line LC identified
• Potential applications identified, but no industrial processes investigated practically
• Project direction altered to concentrate on sampling probes suitable for LC and other on-line techniques

Deliverables

Progress July 1998 to November 1998
Document Ref: 99/P7/1
Issued: 26 March 1999
Executive Summary

"Two companies launched on-line HPLC in the 1980’s, Applied Automation and Waters. Both companies had some successful installations, but the volume was not sufficient to support a product line. The major reasons for this include sampling and sample preparation difficulties, cost of the purchase and disposal of solvents, hazards of flammable solvents, high maintenance costs, the need for specialised staff and a lack of willingness to train staff, long sample times and the general complexity of the instrument. These all increased costs and outweighed the benefits of on-line sample preparation, although the other factors have played a significant part in the lack of popularity of the technique for on-line process analysis.

Recently, there have been successful applications, particularly in analysis of river water and fermentation monitoring. The success in these areas has been achieved through the use of reliable sampling systems based on ultrafiltration technology. Developments in HPLC and data processing should enable the complexity, running and maintenance costs of the instrument to be significantly reduced, whilst increasing the reliability of the data produced. An initial design for a sample valve has been produced for CPACT by Labman Automation, which if successful will simplify the sample process from many fine chemical and pharmaceutical applications, and with modifications may handle more difficult samples.

On-line HPLC offers significant advantages over other methods such as spectroscopic and flow-injection analysis, as complex mixtures can be simultaneously analysed for a number of components over a wide concentration range, with relatively simple calibration. Combining reliable sampling with lower cost analysers should make the technique viable. It will have most benefits for those applications where the advantages justify the possibly greater costs of the instrument. Areas identified so far are development, pilot plant and troubleshooting applications. Other potential applications have been identified by most of the manufacturing company members of CPACT in the project team."