What are the steps that can be taken to reduce peak fronting?

The most common cause - indeed, usually the only cause - of peak fronting is excessive sample load.

So to minimise peak fronting, all you usually need to do is to reduce the sample load on your column.

If you are using a 4.6 mm ID, 30 cm analytical column, this means a maximum injection volume of 20 microliters, sample concentration in the range of 10 microgm/ml.

S.K. Srinivas

I had the problem of peak fronting once in using a flow rate too fast to allow the analyte to equilibrate between the mobile phase and LC column. Try a slower flow if the above post doesn't apply!

I am using a 150X4.6mm,3.5µ column.My flow rate is 0.8ml/min.Injection volume is 10µl.All the peake have a very good shape except only one of the related substance.

The 4 most common causes of peak fronting....

1: Column Overload
2: Sample solvent is not compatible with mobile phase.
3: Co-eluting peak
4: Column is dead (or dying)

Based on the previous comment, it sounds like a co-eluting peak might be the problem here.

uh about the previous mesage, what kills the kollumn ?

Well, since we seem to be doing things in fours, how about:

1. plugging of the frit or bed
2. development of a head space as a result of mechanical or hydraulic shock
3. adsorption of contaminants from the mobile phase or sample
4. chemical attack on the silica or bonded phase

Pretty much everything I can think of can be shoehorned into one of those categories.

so, how to see the differnce between these 4 options?

Well, they really fall into two broad categories: physical problems (#1,2), and chemical problems (#3,4).

If the problem is physical, there is a high probability that the symptom (tailing, for example), will show up on all the peaks in the chromatogram. If this is the case, the usual practice is:
a) back flush the column to attempt to clean the frit.
b) change the inlet frit.
When you change the frit, you have an opportunity to look for a head space.

Adsorbed garbage is usually treated by flushing the column with either the strong component of the mobile phase or something that is a good solvent for whatever garbage you suspect is adsorbed (this assumes you know something about the history of the column, of course). If that fails, then diagnosis #4 is made by elimination.

There's a good treatment of all of this in the book "Troubleshooting LC Systems" by Dolan and Snyder. You can order from the publisher (Humana) or from LC Resources (http://www.lcresources.com/dlmain.htm)

By way of "truth in advertising", I work for LC Resources -- but it *is* a good book!

If you have frit plugging problems, your column head pressure will increase, but the chromatography stays the same. Keep track of pressure and peak width, and watch for trends.

A quick way to check for voids at the column head is to reverse the column. If the peak shape is dramatically improved, then you've probably found the problem. But THINK!! before doing this. If you are running a guard column, almost always a good idea, then you are dealing with two columns, either or both of which could have voids, and any garbage that may be on the frits. It is possible to plug your detector lines if you aren't careful.

I run a cartridge system with frits that are external to the column, so I can reverse the cartridge and not worry about junk washing off the frits. With modern short, small particle columns, changing an internal frit is no longer an option, although years ago some members of our group routinely salvaged columns this way.

If I'm not using the cartridge, I use a particle filter, so I can still reverse columns without worrying.

The exact problem is that I have three related compounds (a)carboxylic acid derivative (b)carboxyaldehyde derivative and (c)alcohol derivative of the compound.The carboxyaldehyde derivative is giving me a problem of broad peak shape.All other related compounds inncluding (a) and (c) have a very good peak shapes.

To Scott Fredrickson:
I have a problem with broad peaks and most presumably the cause lies in void volume at the column head.
Can I really just change the flow direction to check it ? (I'm not using any guard column)

If you are not using a guard column or particle filter, there is a chance of backwashing particles (pump seal bits, stuff from unfiltered samples, etc.) off the head of the column. To prevent damage, thoroughly flush the column after reversing it but before hooking it to your detector, so particles do not end up in the LC tubing. Those little SS tubes are not cheap!!

This will only diagnose the problem, not fix it--at least not fix it permanently. The problem will return, and you will need a new column unless you can remove the column frit and fill the void.

If you don't want to deal with guard columns (they can be annoying), at least put a particle filter in front of the column.

Three peaks, the first and last of which have good peak symmetry and the middle has peak asymmetry of the fronting variety. Sounds like a classic diluent vs. mobile phase problem. The diluent for the fronting chemical is a slightly better solvent for the compound than the mobile phase is. As a result, the compound is not deposited on the stationary phase as efficiently as needed and is carried along until the diluent plug is carried away by the mobile phase. To correct this, use a weaker diluent (slightly less solubility for the fronting chemical) or use a different mobile phase strong solvent (ACN instead of MeOH for example).

Another possibility for one broad peak in an otherwise good chromatogram could be due to the specific chemistry of that compound. The pH of the mobile phase should be checked and adjusted if necessary. It should be atleast 2 units away from the pKa value of the compound.

It is well documented that the rotation at the proline and similar bonds in diastereomers may cause changes in resolution. To find if this is a case one needs to change the temperature of separation. Instead of one peak with fronting you can get two peaks at low temperature or one nice peak at high tempearture.

I've already observed this kind of behaviour for aldehyde peaks. I suspect partial oxidation of the aldehyde into carboxylic acid during the separation, probably because the mobile phase and/or the sample diluent contains dissolved oxygen (uncomplete degasing). The aldehyde oxidises to the carboxylic acid, which elutes earlier ; this results in a fronting peak. I don't know whether there is literature on that subject, but I think it's a quite general problem with fragile aldehydes.

what are the main differences between gas-liquid and liquid-liquid chromatography?