[STM]

I am contemplating on either using either an in-line filter (0.5 um frit) alone for my analytical column or also use a guard column.

For cost reasons, my choice is biased towards an in-line filter and not a guard column.

Kindly advice on the benefits and disadvantage of using either.

What would you recommend?

Thank you

[clarkstill]

I'm afraid I'd go for the more expensive guard column option.

The filter will remove insoluble suspended particulates, but won't prevent contaminating the column with slow-moving contaminants. I do a lot of normal-phase chiral HPLC, and rely on the guard column to prevent deposition of any highly polar contaminants on the analytical column which would not elute or elute very slowly.

I understand the cost/benefit analysis - obviously a filter is better than nothing, but the cost of replacing your analytical column if it becomes contaminated is a great deal more than the cost of replacing a guard column.

Ultimately it comes down to what you can afford. If you are scrupulous about the purity of samples that you use you might get away without a guard column, but there's certainly a risk.

[Arkcon]

Another thing to consider is your analytical needs.  As was said, if your samples are contaminated, a guard column will save your analytical column.  But if you're looking for contaminants, you'll have to re-verify that your method still detects them by checking with spiked samples. 

If your samples are really dirty with highly retentive substances, your guard may foul rapidly.  Note: a guard column can be 1/4 to 1/5 the price of an analytical column.  They can also be as much as 1/3 the price.  They don't generally go higher than that.  But if you foul 3, you might as well just get another analytical.  Then again, how much work do you have to do -- when is this project done, is the question.

Query: you have to remove particulates from the system anyway.  Why is a filter a possible answer to your problem?  Are your samples that dirty?  Is a component not as soluble in mobile phase as you'd like?  Can you filter before analysis?  Note: all answers may require a re-verification of your method.

[STM]

Thank you for your response. My experiment at times involves clean samples (done in MilliQ water or buffered solution) while at other times it involves natural water samples which come with a matrix of compounds as expected. My reason for not wanting to use a guard column, apart from the cost-benefit, is that I want to minimize backpressure also.

But like you said, I have to consider my analytical needs.

Thank you

[Arkcon]

Adding another column may increase back-pressure, at least slightly.  Pretty obvious when you think about it -- more media, even a tiny amount, is more particles in the path for pressure.  Guards are designed to minimize this, but it can be noticeable.

Adding a column, even a tiny one, adds dead volume, even a little bit.  Some are better than others, i.e. the guard that screws into the column.  However, you may get slight retention time shifts, or, if you must plumb the system yourself -- with guard column housing, tubing, etc, you may get significant RT shifts.