I want to reproduce an HPLC system described in a paper published 10 years ago. I'm only interested in running determinations, not doing research in HPLC methods or non-trivial optimizations of the published technique.

I already know the general specifications for the system, mobile phase, column type, etc. A problem is that I can't reproduce the system exactly; none of the components are still being made except the column. Therefore, I have to design a functionally equivalent system from parts I can buy today.

Never having designed (or operated) an HPLC before, I have a few questions. It appears to be a fairly common practice to design a system out of components from several different manufacturers. What are the risks involved in doing this? I imagine that if any problems crop up, each manufacturer would point at the other's equipment as being the cause of any fault, so I'd be assuming a greater burden of qualifying the equipment if I mix vendors. Also, the "full line" vendors equipment is designed to fit together, so it seems like mixing Lego blocks and Lincoln Logs. And yet, there are some vendors who _only_ make one type of component like a fluorescence detector or a post-column reactor; how do those guys make a living if everybody buys complete systems from one vendor? Only Waters appears to have all the components I need; Agilent and Shimadzu don't seem to offer a post-column reactor.

How much support can I expect from the manufacturer for installation and set-up? If I say I need this pump, that injector, etc. will the manufacturer send someone out to bolt it all together for me? Will they check out the flow rate, gradient accuracy, and run some standards to verify it works, or do I need to do all that myself (or hire a consultant to do the OQ for me)?

What would be ideal is if a vendor offered to build me a functionally equivalent system to the one described in the paper, with running the same standard described in the paper and getting similar resolution of the peaks as the acceptance criterion. Does anybody do that?

If I need a consultant to help get things up and running, how do I find that sort of consultant?
What should I look for in evaluating the competency of a chromatograpy consultant? How much should I expect to pay, in terms of hourly rate?

IF YOUR GOAL IS ANALYSIS AND YOU DON'T WANT PROBLEMS JUST BUY YOUR SYSTEM FROM WATERS. THE EQUIPMENT WILL HAVE PREDICTABLE PERFORMANCE AND THE SERVICE IS GOOD. MILLENIUM CDS IS BETTER THAN MOST AQUISITION SOFTWARE. IF YOU ARE ON TOO TIGHT A BUDGET I WOULD BUY A USED WATERS OR HP SYSTEM FROM A RESELLER INSTEAD OF BUYING JUNK FROM SOMEBODY ELSE. DON'T GO TO A CONSULTANT UNTIL YOU AT LEAST LOOK AT A MANUAL. HPLC IS ABOUT AS SIMPLE AS ANALYTICAL CHEMISTRY GETS. IF YOU DESIRE SOME HELP GO TO THE UNIVERSITY CHEMISTRY DEPARTMENT AND PAY THE GRADUATE STUDENT WHO IS RUNNING THE INSTRUMENTAL ANALYSIS LABORATORY A COUPLE OF BUCKS TO HELP. KEEP IN MIND THOUGH THAT THE REQUIREMENTS TO PUBLISH DON'T INCLUDE VALIDATION SUCH AS RUGGEDNESS OR ROBUSTNESS. OFTEN RESULTS OBTAINED FOR THE PAPER AREN'T EXACTLY WHAT YOU WILL SEE IN YOUR HANDS. DEPENDING ON THE NUMBER OF SAMPLES YOU MAY JUST DECIDE TO SEND THE SAMPLES TO A CONTRACT LABORATORY WHO HAS SKILLED CHEMISTS TO GIVE YOU YOUR RESULTS.

GOOD LUCK

I am amazed as to why anyone would want to reproduce a system exactly from 10 years ago. The column is the most important part of the system and if you can still buy the exactly the same type, you shouldn't have too much difficulty in reproducing the original separation. Although we all have our old favourites- in cars the old Beetle and Mini spring to mind- few would argue that the more modern equivalents are better. The only thing that I might worry about is the difference in gradient delay volumes of the 2 instruments-is that what's worrying you?

No, it's an isocratic method. However, there's another protocol for doing the same determination using a gradient method, so I figure it would be prudent to buy a binary pump. Also, that would make solvent mixing and operation of the column more convenient.

My main concern is uncertainty from never having touched an HPLC before. I'm worried that there's some unpredictable gotcha that won't become apparent until I try to get it running.

Depending on the instrument vendor, you might be able to get them to have their applications people specify a system, run it on that system, and have "acceptable" performance for that method established as part of their installation qualification. It might be worth a try. The worst they could do is say no.

We operate 22 HPLC's from three different vendors in our lab. They all produce equivalent results with the 200 methods we run. Some sytems are easier to operate, others require less maintence. I without a doubt would stick entirly with one vendor to simplify installation. All of the HPLC manufactures will install and calibrate the system for you. I also recomend following their PM schedues. They will even do that for you...for a price.
There are lots of gotcha's with HPLC methodology, but they have little to do with the manufacuturer and more to do with the technique.

I also agree with getting a complete system from a single vendor if possible, eliminates "I can't work on or be responsible for this" issues. We've used Alltech and other detectors on Agilent with good results, but realize that any subsequent troubleshooting falls mainlyon our shoulders. It's bad enough these days with computer vendors blaming software, and vice-versa...

I agree 100% with the statement above. As for the other comment that "the column is the most important part of the system", I'll have to disagree. As far as duplicating some elses analysis, yes, the column is very important, but in general, the pumping and injection systems are the most important parts of the system. I have Agilent 1090's, and 1100's, Water's old stuff and an Alliance. The 1090's are as rugged as they come but might not have quite as sensitive a UV detector. The 1100's are very good, but there are some sample carryover issues (the 1090 has the best injection rinsing system I've seen). The Allaince has the highest sensitivity PDA I've seen, but the software is not designed for everyday R&D work. There were some injection port carryover issues originally, but the seal pack has been re-designed and I believe it is better.
Waters stands by their instruments longer than Agilent. The 1090's are out going out of support soon while an ancient Waters 710B autoinjector of ours was just completely re-built and will probably run for another 10 years.

The way I understand it for Agilent is that going "out of support" is not a death knell for the instrument, it just means that parts availability would not be absolutely guaranteed. Most likely the most "popular" parts would be in stock for quite a while, and Agilent would still repair on a "best effort" guarantee.

However, "best effort" is not always very good. We had a situation where we tried to order a new keyboard for an integrator (3396A). We were told the integrator was not supported and the part was not available. No attempt was made to find out if a keyboard for a later model would work. Fortunately, we were considering buying a new GC at the time so I called the sales rep and said that lack of parts support could affect our purchase decision. Amazingly, the sales rep was able to find an equivalent part number. This has not been the only time I have not been impressed by Agilent support. For whatever my opinion is worth, if we did not already have HP/Agilent equipment, I don't think I would be likely to purchase new equipment from them.

Support may considerably vary from one country to another (or even from one region to another in the USA). We always had a good support from Agilent (in France).

I agree with what has been said. For gradients, I have found high pressure mixing pumps work more reliably for me than low pressure mixing pumps.

Noname,
The advice has been very good. I would ask about your application, for example, how difficult is the separation. If the analyte peak is well separated, forget about equipment functioning differently. Only for difficult separations, ruggedness becomes a problem.
A guarantee for a method is mostly to give you a way to back out of equipment. Unless the local support is good, method support tends to be like new car problems from the dealer, just keep you quiet until the warrantee runs out.
Where are you located? The local dealer makes a big difference. For example, in Southern California, Cole Scientific reps Shimatzu. They do a hell of a good job of support and repair making an instrument bought from them more valuable.
Another question is very personal. Are you a chemist yet and if not when do you expect to be? Sounds awful to ask and I am sorry if I hurt your feelings, but it is to a point. Everybody works after school and gradually gets the feeling that given enough time, they can solve any chemical problem. After that time, they run instruments like artists play instruments, with ease and skill that makes it seem effortless and easy. If you are getting close, HPLC will fill you with joy, with its relationships with chemical and physical principles and the ease of analysis compared to wet methods. A chemist could probably do your method on any modern equipment and you can also. Take the attitude that you can and you will.
Jim

I agree with technical support in a particular location be one of your selection criteria. One vendor may have good support in a particular country/region, and others have poor support, and vice versa.