Before you can answer the question(s) you will first have to develop a method and then run a loading study. That will tell you how much sample can be safely loaded on the column while still being able to purify the portion you want. When you have that information, you will be able to calculate the amount of mobile phase used based on the flow rate and run times.

When you need to learn about something, please spend the time researching the answer on your own. That is part of the learning process. Will you be using low pressure or high pressure LC to perform the purification ? Be clear about what you wish to do. Start with a web keyword search to learn about LC and HPLC for peptide purification. Read up on your exact sample. Determine its purity using an analytical method first so you understand how much purification will be needed. Characterize the sample so you can be sure it is the right peak (UV spectra or other chemical or physical properties). Look for example methods (on the web). Learn what you can and also contact someone local to you (not on the web) for personal assistance.

Before you can answer the question(s) you will first have to develop a method and then run a loading study. That will tell you how much sample can be safely loaded on the column while still being able to purify the portion you want. When you have that information, you will be able to calculate the amount of mobile phase used based on the flow rate and run times.

When you need to learn about something, please spend the time researching the answer on your own. That is part of the learning process. Will you be using low pressure or high pressure LC to perform the purification ? Be clear about what you wish to do. Start with a web keyword search to learn about LC and HPLC for peptide purification. Read up on your exact sample. Determine its purity using an analytical method first so you understand how much purification will be needed. Characterize the sample so you can be sure it is the right peak (UV spectra or other chemical or physical properties). Look for example methods (on the web). Learn what you can and also contact someone local to you (not on the web) for personal assistance.

Thank you for your prompt reply Dr. Bill. I have some basic knowledge about HPLC. I have already purified few samples with HPLC. I have developed a gradient method for purifying my peptide compounds. I dissolved my compound in 0.100 g in  5 ml ACN solvent. I got 3  peaks in HPLC  all of equal intensity. For purification, each time I am injecting 0.200 ml/injection. If I go beyond that concentration, flat topped signals were noted. No separate peaks.  I can go with 0.2 ml/injection condition. So I need at least 25 injections to purify my compound at 4ml/min eluent rate.  Is this normal ? or Am I wasting the solvents? Do I need to optimize flow rate and mobile phase again?   

One thing you might do is try dissolving the sample in a "weaker" solvent. Please note the poster link below. One example shows a compound dissolved in DMSO, which limited load capacity (for that polar compound). Dissolving the sample in acetonitrile is worse, since acetonitrile is "stronger" than DMSO. If your peptide is very non-polar, DMSO may be a better choice as it is weaker than acetonitrile. If you have a polar peptide (as mine was), dissolution in water is a better choice. You may need to "pre-purify" to clean material that prevents the peptide from dissolving in a weaker solvent. The pre-purification can also reduce the other impurities. A mixture of water and acetonitrile may work for you too.

http://www.teledyneisco.com/en-us/liquidChromatography/Chromatography%20Documents/Posters/Pre-Purify%20Peptides%20with%20Flash%20Chromatography.pdf#search=peptide

You have not provided any detailed information about your actual HPLC method so no constructive comments can be made regarding the quality of the method. Here are some very general comments based on the info provided so far.

 "Do I need to optimize flow rate and mobile phase again? "

You always need to optimize the method used to resolve the sample apart. In this case, you need to optimize the method for purification. This is first done on the analytical scale, then scaled up, as needed. Try to find a method which does so quickly (short time), with lots of baseline space between the peaks (excellent resolution) and uses a simple mobile phase. Verify sample loadability and that the sample dissolves 100% (not 99%) in the solution and mobile phase used (Use a weaker solution, not pure ACN!). *This is especially important for gradient analysis. Do not precipitate the sample out during the run.

You implied that your detector is saturated. Is resolution lost? Do not assume that resolution is lost because the signal is off scale. For purification, we typically use a detector with variable or narrow path-length flow cell (reduce the path-length to reduce sensitivity while maintaining selectivity at the same wavelength) to better understand where resolution is lost. Do you have this capability? Is resolution lost?

If you are currently limited to 200 ul injections, can you transfer your method to a larger column (no idea what sized or type column you have now, but scale-up solves this)? That is what scale-up is all about, trying to maximize sample load to purify in as few runs as possible, but maintain purity levels. You may be able to run the scaled up method with just one injection (or just a few, depending on what type of injection system you have). Note: Prep runs often use lots of mobile phase so 25 analytical runs may still use as much mobile phase as one prep run. There is no way to know how much mobile phase will be needed until after you develop and optimize the method of separation.

As you are new to HPLC, please get some local help with your project. These are very fundamental questions and someone local to you who is experienced in liquid chromatography is in a better position to help you with this project than someone via the web.

For sample preparation please do not use ACN as also mentioned above. Even though the ACN may get dilute in the mobile phase after injection, it will lead to less peptide binding and elution of some peptide before the gradient.

For sample preparation, either dissolve the peptide in 5%ACN in H20 with 0.5% TFA or If the peptide samples are in a buffsr than mix 3 parts of the sample with 1 part of sample buffer (20% ACN in H20 with 2% TFA). Finally Filter through 0.22 micron filter. 

To avoid flat top peaks, increase the attenuation value of detector in HPLC control software. Also, if the peptide contain the phe tyr trp residue, record spectra at 280 nm instead of lower wavelengths which give high mAu values.

You didn't mention the column dimensions, a 10x250 mm semi prep column can take up to 50 mg protein load. So you need to optimize for maximum resolution and yield.

I hope that your gradient is preceded by a run to achieve baseline post injection. 

Like my usual semi-prep run includes 

Injection followed by 5 column volume (CV) mobile phase A, then elution with 0-60% mobile phase B in 10 CV.  

Thank you Dr. Bill. I will do some optimization.

I will try with DMSO solvent also as suggested by Dr. Jack.

Thank you Dr. Mangal Singh. we have XBridge BEH C18 OBD Prep Column, 130Å, 5 µm, 19 mm X 250 mm, 1/pkg column. Solvent flow rate is 4ml/min. I have done all 25 injections got pure peptide. Three peaks of equal intensity at Rt= 12.5 14.1 and and 20. My compound peak is at 12.5. Like Dr. Bill's suggestion I will do the optimization. If you have any other suggestion also please post that. 

Thank you Dr. Kou Hayakawa. I will also test my sample with 3-phase system. Our  HPLC model can support 3-phase system. I'll give a try.