IBM Instruments Inc

---

Focus would like to have a large mainframe, perhaps with mag tape capability, that could do that easily." Powerful minicomputers and mainframes will probably continue to play a significant part in the overall market for chromatography data systems. But there can be little doubt that there is an overall trend toward downsizing and networking. Although Sohio's mainframe-based data system is doing a good job, Kosman remains open to the idea of increasing the laboratory's reliance on micros at some point in the future. "Perhaps the ideal would be a network, including micros and a large mainframe," he says. "The micros could handle the data acquisition, and the mainframe could be used for fancy workup and archiving of data." The most visible micros on the current scene are the IBM personal computers (including the IBM PC, the PC-XT, and the PC-AT), and more and more chromatography software is being written for these seemingly ubiquitous machines. At this point, at least five companies are offering IBM PC-based chromatography data systems: Nelson Analytical (Model 3000), Dynamic Solutions (Maxima), Spectra-Physics (Labnet-XT/AT), Tekni-

vent (PC-Chrom), and Data Translation (Chromlab). IBM PC-based software is also available for more specialized chromatographic applications, such as diode array LC detection (LKB). "People want to use the chromatography software, but they also want to use standard word-processing, data base, and spreadsheet programs," claims Spectra-Physics's Terrance Rooney. "The availability of these additional programs on the IBM PCs offers a tremendous advantage relative to companies that don't offer IBMcompatible software." Rumors continue to be bandied about that this or that instrument company is rewriting scientific software for the IBM PC. Communications standards are lacking No matter what the architecture of a particular laboratory's data system, the incompatibility of computers from different vendors has created a barrier that makes it difficult for some laboratories to consider new approaches. "Once you're locked in with a particular system, it's hard to lock yourself out," explains Sohio's Kosman. "We have a lot of money invested in one or

two manufacturers' equipment, and this makes it difficult to say, 'Another vendor has a much better product, so let's switch over.' " In many laboratories the selection of new chromatography data equipment is constrained by a tendency to stick with one or two vendors whose products are familiar to staffers and fit in well with equipment presently in place. Terrance Rooney agrees that hardware incompatibilities are a significant problem: "We're trying to set up protocols so Spectra-Physics products can talk to all kinds of units—computers from Perkin-Elmer, HewlettPackard, Varian, DEC [Digital Equipment Corp.], even IBM mainframes. But it's still not clear how people are going to be able to do that. There is no easy solution. It's more than a matter of just plugging the wires together." Indeed, many of the chromatographers we interviewed are concerned about communications standards and place standardized protocols near the top on their wish lists of matters they would like to see vendors address. "It's sometimes hard to connect an older GC or LC around the lab with new equipment from another manufacturer," says T. K. Rehfeldt, senior

GCT

IBM Instruments just raised GC/IR to a higher power. Now, one company gives you both high-resolution separations and trace component characterizations simultaneously in a single integrated system.

1258 A · ANALYTICAL CHEMISTRY, VOL. 57, NO. 12, OCTOBER 1985

Focus research scientist at Sherwin-Williams Research Center in Chicago. "And it's not easy to translate data bases from one system to another. We would like to be able to transmit data from small computers to larger systems, but that's often not possible. There is a lack of suitable standards for data transmission." Vendors of chromatography data systems have begun to respond to such pleas. For example, Nelson Analytical's Series 6000 system, which is based on the HP 1000 A Series minicomputer, can pass raw data, methods, and data management files to remote computers such as DEC PDPs and IBM mainframes. The Waters 840 data station can also communicate with DEC minicomputers and IBM mainframes. Optimization and data bases Commercial chromatography software packages are designed to handle data from both GCs and LCs because these forms of chromatography are similar in many respects. However, there are significant differences, including the fact that although carrier gas selection in GC is fairly straightforward, deciding among the seeming-

ly infinite variety of solvent combinations used in modern LC can be very difficult. LC solvent optimization software has been designed to help ease this decision-making process. The primary benefit of solvent optimization software is that it saves time. "The first thing people do when they buy our optimization system is go back and run separations they've already optimized themselves, to see how well the software does," says Spectra-Physics's Terrance Rooney. "And they frequently find the results are remarkably similar. But where it might have taken them anywhere from a week to a month to optimize the separation, our system will take 10 to 12 hours." In operation, LC optimization algorithms attempt to increase the selectivity (alpha values) of overlapping chromatographic peaks (Berridge, John C. Trends in Analytical Chemistry 1984, 3[1], 5-10). The alpha parameter, which equals the ratio of the capacity factors (k') of two adjacent peaks, is increased through adjustments in the composition of the mobile phase. Although some vendors do not provide detailed information about their

optimization algorithms, most seem to use a simplex procedure or some modification thereof. The simplex algorithm is an extrapolative procedure in which movement toward the chromatographic optimum is determined by the results of a sequential series of experiments. An algorithm based on a solvent selectivity triangle used to be sold with Du Pont's Sentinel LC system—the software was coded into the instrument controller—but this was not perceived as a successful product, and it was eventually withdrawn from the market. "Optimization programs are very important to us," says Joseph Kosman of Sohio. "We use them extensively. We particularly like the systems that are totally automated—you put the sample in and the machine does all the work itself. This makes it possible to do methods development at night or on weekends. A system that requires a person to close the loop is not as valuable." At this point, three analytical instrument companies—Perkin-Elmer, Spectra-Physics, and LDC/Milton Roy—provide optimization programs, but a number of other vendors, such as Waters, have products in prépara-

/FTIR The IBM Instruments family of integrated GC /FTIR systems lets you select a single instrument that, for the first time, includes all these features: • Capillary GC with our unique dual-oven design (GC ) to provide enhanced, high-resolution chromatographic separations. • Sample enrichment capabilities to facilitate trace component detection and analysis.

• An innovative interface engineered to minimize dead volume and maintain optimal resolution throughout our GC /FTIR system. • A powerful data processor that permits the display of real-time spectra, Gram-Schmidt chromatograms and up to four other selective chromatograms. • An advanced GC/IR combination backed by the applications sup-

port and technical service of IBM Instruments. To learn more about our GCVFTIR, call us today at 1-800-243-7054 (in Connecticut, 1-800-952-1073), or write to: IBM Instruments, Inc., Orchard Park, EO. Box 332, Danbury, CT 06810.

Integrated solutions for the laboratory

IBM Instruments Inc. ANALYTICAL CHEMISTRY, VOL. 57, NO. 12, OCTOBER 1985 · 1259 A