You can essentially do it without it but Ficoll paque can facilitate separation of white and red blood cells visually. What i mean is that by centrifuging blood, you will get red blood cells and granulocytes at the bottom. On top of them, white blood cells will appear as a white buffy coat and then another layer of plasma. The problem of not using Picoll is that WBC represent only 1% of whole blood, so this layer can be hard to distinct and collect, so that you can stain for proper antibody your white blood cell cocktail and isolate neutrophils. Therefore, by using Picoll, you will basically introduce a new transparent layer, between red blood cells and white blood cells. This also reduces RBC contamination. From bottom to top, blood constituents will appear as red (red blood cells and granulocytes), transparent ficoll, white buffy coat (white blood cells and platelets), yellowish (plasma and platelets). If you process a blood sample of 5ml per se, you can use at least 2 ml of ficoll, so that you make the distinction clear. Remember Red blood cells represent about 45% of blood, making it almost 2 ml in a five ml sample. Hope this helped
Article Isolation and Functional Analysis of Human Neutrophils
Common themes are either PBMC prep and removal followed by dextran sedimentation of RBCs, higher density separation media(ie 1.114g/ml vs 1.077 in PBMC isolation media) or discontinuous gradients. Generally a RBC lysis step is also necessary.
This question has been asked lots of times previously so you may find more information by search example below.
One suggestion is to use Ficoll with two density gradients (1,078 and 1,119 g / ml) and collect a "cloud" deposited over the largest gradient. You will get between 80-90% neutrophils. Good experiment.