Many problems in an LC system show up as changes in the chromatogram. Some of these can be solved by changes in the equipment; however, others require modification of the assay procedure. Selecting the proper column type and mobile phase are keys to "good chromatography

Thanks Igor, its a capillary column, 300um ID, so don't think its overloading. Rgd the connections I've checked them but in any case if there were mixing chambers should have affected all the peaks where as in my case only some peaks are broad/tail. Will try and new column as you suggested

Is there a pattern (are only early eluting or only late eluting peptides affected)? Did you check you buffer system?

No pattern (see attached screenshot). 

Solvents are pretty standard, 2%ACN/0.2%FA for loading, for gradient A: 0.1%FA, B: 80% ACN /0.1%FA. All prepared about a week ago using LCMS grade reagents. 

looks like column and analyte interaction because some peaks are good so the one with tailing has more activity. Pick the column with more inert (not just use the column you find in the drawer). Don't want to suggest TFA because you use MS detector unless you try it on the UV detector first.

http://www.waters.com/waters/en_US/Peptide-Separation-Technology-Columns/nav.htm?cid=513078&semCBU-PeptideSeparationTechnologyColumnsNonBrandUS&gclid=CPrwsuGD7cMCFWQV7AoddQEA1Q&locale=en_US

read this about the column

See that there is no much void volume in the trap column holder and the switch valveset up. Also try to load directly bypassing these two to the analytical column directly with minimum sample possible (to save analytical column) and if you find betterment then these two are the culprit.  

To circumvent the TFA problem with MS you might try to add 2,2,2,-Trifluoroethanol (TFE) to your buffer.

Thanks all for the great suggestions. I installed a new column today and all seems to be ok. Still not sure why only some peaks were affected previously?

because your new column is less active than the old one and that why it works.

Please elaborate, what do you mean by less active? 

It may exposed the silanol site that may absorb your analyte more, then you got peak tailing. For other peak that get sharp peak the analyte may not react so strongly to the silanol site.  You need well encapped C18 column. 

Thanks Narong, makes sense. 

Daniel,  yes trap gets loaded and eluted in same direction. I like this configuration since trap can protect the column this way

It may be the column overloaded. Reduce the amount of the sample injected to see if it improves.

I think you should change your column.