If you simply let a tube sit, the red cells will settle as they are denser. The time it takes to settle is, in fact, a clinical test called the erythrocyte sedimentation rate. Suffice it to say, for your purposes simply letting plasma (purple top tube) settle for an hour would generally be sufficient to have relatively red cell free plasma if you pipette from the top of the tube.

Blood is a colloid of RBCs and other cells. Colloids are often clarified by the use of a flocculant. The flocculant either neutralizes the surface charge on the suspended particles (cells in this case), or tethers particles together. Either way, the particles tend to clump together and settle much faster than they would otherwise.

I am not aware of this being done with blood before, but I am actually interested in achieving this for entirely different reasons. We need to harvest chicken RBCs for a non-medical use. If you run across any information relevant to flocculating blood, I'd love to see it.

Plasma or serum can be separated from whole blood without centrifugation by allowing the blood to just let stand. By gravity all the cells will settle down in due course of time (if time is not the question). If you allow the citrated blood to stand in a tube, the supernatant is the plasma. However, if you let the uncitrated whole blood to stand, the serum will separate in due course of time. If you are working with small volumes of blood this may not be as practical but may be tried. However, from your question I do not understand why you do not want to centrifuge. How does centrifugation affect your experiment?

Apart from let the red cells sediment as described by the colleagues before, which is the best way to achieve what you want, you could also coagulate the blood cells, e.g. By adding thrombin, if this does not affect your analysis

The negative charges (zeta potential) that hold red cells apart is from sialic acid. Elevated protein levels help neutralize this charge, hence the more rapid sedimentation rate when you are sick and have an elevation of acute phase reactants in the blood. you can also agglutinate (clump) the red cells in a type A or type O donor by adding the respective anti-A or Anti-B isohemagglutanin. This can result in complement mediated hemolysis, so depending upon exactly what you are measuring this may not be helpful.

Not sure it would be applicable in your case, but there may be a way to filter through a big-pore size filter keeping only the RBCs . Not sure though if the bacteria/viruses/whatever might also stick to the filter due to their adhesive properties.

Just thinking here (May be totally not applicable)...

In my experience with blood, although the cells would fall on the bottom in small volumes is really difficult to take the supernatant without disturbing them.

And here I found this site for HIV testing of plasma.

http://cen.acs.org/articles/91/web/2013/10/Low-Cost-Device-Quickly-Isolates.html

I hope it helps...

I found also a pdf file describing micro filter for virus testing on fields (from 2013) - but haven't read it completely yet.

http://faculty.uoit.ca/macdonald/Gong-MacDonald-FieldTested-2013.pdf

My Dear

If you simply let the test tube stand for minutes in cool room of coursr red cells will be precipitated and you have chance to collect serum as well as plasma.adding anticoagulate of course you will collect plasma.

Just let the blood rest at room temperature for few minutes and the cloth will separate from the serum. You will get less serum than if you had used the centrifuge, but, probably, it will be enough.

as the answer Ged gortin you can do it of course it is important know hematocrit of your sample and the volume of serum which you need .i don't suggest you for special tube due to it cost many and isn't cost effective and some flocculant aren;t good and may be damaiging your serum .it is better you take 5 cc blood and put it in a water bath

for at last one hour it may bring 50/50 serum

I agree with Omer Iqbal. Just let it stand. Erithrosedimentation rate is this. It would take the same time, aprox 1 hour or 2 for anticoagulated blood. Remember that normal sedimentation rate is up to 20 mm in 1 hour.

I agree with other colleagues to put your blood samples stand for while in room temperature.

Add thrombin

It's quiet simple.

for Plasma:

add EDTA (1 -1.8 mg/mL) to the tube before sampling. then keep the blood for 8-10 hours in fridge. you would get the plasma above the red blood cells phase.

Serum:

the same as above. leave the tube which is contained the blood (without EDTA) for 8-10 hours in fridge, after coagulation, the product in the upper phase would be serum.

you can accelerate the sedimentation by adding HydroxyEthylStarch (HES) to the sample.

Another method is to use a membrane as used in plasmapheresis to separate plasma/serum from blood (under gravity).

what @Seyed Mohamad Jazayeri is how it's done here. centrifugation helps but not necessary.

Thanks a lot for all your answers. We are developing a method to detect some bacteria in plasma. But the test must be really quickly. So, plasma (or serum) separation should be in less than 10 minutes and free-equipment. For this reason we wanted avoid centrifugation step. The normal method takes long time and doesn't fit well to our goal. I have read different solutions for lateral flow, membranes, etc. However we have not found commercial solutions.

Again, thanks again for your time!

Héctor, if you find something different please keep us up-dated. Thanks

Dear colleagues,

I have find some solutions basicaly by filtration. Marketed and prototypes. I don't have any idea about prices.I post webpages in case you are interested:

- http://cen.acs.org/articles/91/web/2013/10/Low-Cost-Device-Quickly-Isolates.html

- http://www.mdimembrane.com/Products/Immunodiagnostics/Products/Rapid%20Plasma%20Separation%20Device.html

Passive settling is time-consuming, filtration may exclude the bacteria under study if they are appreciably bound to cell membranes. Blood culture media stored at 37C has been successfully used in clinical laboratories for decades.

You can do it adding a colloid solution ( eg HES 450) at 30% v/v and incubating for 30min. See Munoz M et al BJA 2010

I also agree with all member answer. Blood will clot and plasma and serum will seperate easily.

When you sample used a collecting tube without an anti- coagulant,(mostly the red topped tube), then place it upside down.  After a day or so the clot would be attached to the lid of the tube, so you just carefully open the lid and the clot would come out with the lid and you remain with the serum.

just leave it in refrigerator for while

We have a blood card that separates it immediately and makes it easy to ship. 

http://www.arrayit.com/Microarray_Diagnostics/Blood_Cards/blood_cards.html

We have a plasma preparation card that prepares a few microliters of plasma from a drop of whole blood in 3 minutes.

http://novilytic.com/

Nice comments. I do agree with most of them. Simple and neat. Best of luck.

I want to produce larger volumes of fish serum for some research. If I merely allow it to settle, will there be proteases that affect the serum?

Ask Dr. Parakash of Stanford University if he can sell you one of his paperfuges. It is a centrifuge made out of paper and string. This scientist made this device to diagnose malaria parasites from blood.

https://www.youtube.com/watch?v=L5ppD07DMKQ

if the whole is allowed for a long period, it will be separated. However, the separation will not be appropriate when compared with samples that were separated using a centrifuge. Besides, the time is usually very short using a centrifuge compared with when allowed to stand.