DNA Precipitation: Ethanol vs. Isopropanol

Since our most popular article of all time (“The Basics: How Ethanol Precipitation of DNA and RNA Works”) was published, many of our readers have asked us to further explain the difference between precipitating DNA with ethanol vs. isopropanol and which is the better choice. So today, I'll meet the challenge and discuss the pros and cons of ethanol vs. isopropanol.

First, let’s review what we know about what is needed for precipitation of DNA or RNA with ethanol:

1. Salt to neutralize the charge on the nucleic acid backbone, causing the DNA to become less hydrophilic and fall out of solution.

2. Ice to chill the sample. Lower temperatures promote the flocculation of the nucleic acids so they form a larger complex that readily pellets under the centrifugal forces of a microcentrifuge.

3.  A nucleic acid concentration high enough to force the DNA out of solution (if the conc is not high enough, you can add a carrier nucleic acid or glycogen to enhance the recovery).

4. Centrifugation to pellet the sample

The difference between isopropanol and ethanol is the solubility of DNA in each solvent.

DNA is less soluble in isopropanol so it will fall out of solution faster and at a lower concentration, but the downside is that the salt will too. With ethanol, the DNA needs to be at a higher concentration to flocculate but the salt tends to stay soluble, even at cold temperatures.

DNA falls out of solution in 35% isopropanol and 0.5M salt. Using ethanol, the final concentration needs to be around 75% with 0.5M salt. So for the typical precipitation protocol, isopropanol is added from between 0.7-1 volumes of sample and ethanol is added at 2-2.5 volumes of sample.

The choice of which solvent to use depends largely on the volume of sample you need to precipitate.

If you are precipitating small volumes of DNA, and you can fit the required amount of solvent into the sample tube, then ice cold ethanol is the preferred choice. You can chill it (some people use liquid nitrogen or -80C to accelerate the precipitation) and precipitate more DNA without the salt contamination that would occur from chilling isopropanol. Afterwards you need to wash the pellet with 70% ethanol to remove salt.

Isopropanol use useful for precipitations where you have a large sample volume  (e.g. the eluate you get after using a Qiagen plasmid Maxi Kit) because less solvent is needed, so you can fit the whole lot in the (15ml) tube. But because salts are generally less soluble in isopropanol than in ethanol, they have more of a tendancy to co-precipitate with the DNA. So to lessen the chances of salt precipitation, isopropanol precipitations are carried our at room temperature with minimal incubation times.   Once the DNA or RNA pellet is recovered from the isopropanol, you’ll want to wash it with cold 70% ethanol to remove excess salt and to exchange the isopropanol with the more volatile ethanol. It is ok to chill the isopropanol precipitated sample, if you are sure that it is not excessively salty.

Because DNA is less soluble in isopropanol, isopropanol allows precipitation of larger species and lower concentrations of nucleic acids than ethanol, especially if you incubate it cold and long. If you do this, just remember to wash the pellet several times in  70% ethanol after pelleting, to reduce the amount of salt you carry over.


So how do you choose when to use isopropanol and when to use ethanol?

Use ethanol if:

  1. You have room to fit two volumes of ethanol to sample in your tube
  2. The sample needs to be stored for a long period of time and will be chilled
  3. You need to precipitate very small pieces of DNA or you have a very low concentration of sample so you want to chill it longer and colder.

Use Isopropanol if:

  1. You have limited in space in your tube and can fit only 1 volume of sample
  2. You need large molecular weight species because incubation at room temperature for short periods of time will not be conducive to precipitating small species of nucleic acid
  3. You are in a hurry and want to accelerate the precipitation of nucleic acids at room temperature

What do I prefer? I use ethanol over isopropanol for most cases, but will use isopropanol if I need to make everything fit in one tube. My preferred protocol is 2 volumes of ethanol and freeze at -20C for at least an hour or overnight for best results. I centrifuge the sample at full speed for 20 minutes to make sure I get everything down. I always wash with 70% ethanol and then centrifuge for 10-15 minutes and keep my eye on the pellet when I decant everything. You need to note or mark the side of the tube where the pellet is expected to be and don’t let it out of your sight when decanting the ethanol!

If I use isopropanol, I avoid cold temperatures because of the excess salt that usually comes down with it. If I want to increase the yields precipitated, I prefer to leave it incubating at room temperature longer vs. chilling the sample. When the DNA is pelleted, the pellet is sometimes more difficult to see compared to the ethanol pellet. It can be clear and glassy. Make sure, again, to note the side of the tube where the pellet should be. Look for it before decanting the isopropanol and 70% ethanol wash. After washing with ethanol, the pellet becomes visible and white. I always make sure it doesn't slip off the side of the tube wall before decanting the supernatant. Allow the tube to drain upside down for a few minutes and then air dry or speed vac dry (5 minutes is enough) and then resuspend in buffer.

Finally, for dry DNA pellets, heating the sample in buffer at 50-60C will help the DNA dissolve faster and won’t damage the DNA. For RNA, heating can be used too (in water) at temps around 42C.  Overdried DNA and RNA will take longer to dissolve so make sure not to speed vac for too long.

So now you know the difference between ethanol and isopropanol and when to use which. If you have any questions, or anything to add, please drop a comment in below.

About the author: Suzanne Kennedy
Suzanne is Director of R&D at Mo Bio Laboratories in California, and the author of their blog, The Culture Dish. She has a PhD in Microbiology and Immunology from Virginia Commonwealth University.

See more from Suzanne Kennedy Visit their website Twitter

29 thoughts on “DNA Precipitation: Ethanol vs. Isopropanol”

  1. Jim H says:

    Agh! I still cringe when people cling to the false belief that cold temperature is necessary. True, it does affect solubility if you're letting gravity do the work, but it actually decreases yield if you're using a microfuge due to the increased viscosity of the solution at lower temperatures. This Inconvenient Truth was published in 1985 by BRL, but is still such an icon of ethanol precipitation that the "Cold Ethanol Denialists" refuse to budge (even the scientists at BRL-> Life Tech wouldn't change all of the technical manuals). See the Zeugin JA, Hartley JL reference in this wiki http://en.wikipedia.org/wiki/Ethanol_precipitation

    The main finding is that extending centrifugation time (instead of extended periods in -20C) maximizes the yield.

    1. Avatar of merajrizvi merajrizvi says:

      Hi Jim, Thanks for the valuable information. Does it really make a difference if the DNA settling down under the force of gravity or under the centrifugal force in a centrifuge? I usually keep the DNA at -20 degrees for a couple of hours after ethanol precipitation and it seems to me that the yield gets better although i have not compared the results quantitatively.


  2. Thanks Jim! This is great information and I wasn't familiar with the BRL papers.
    Are you the Jim Hartley from BRL who wrote the paper cited by Wikipedia?

    It is a very interesting paper but I have some comments.

    It would have been better if the incubations actually compared room temperature to 0C. These experiments only used cold temps. Even experiment 2, the centrifugations are all at 6C. It would have been good to see the difference between centrifugation at 6C vs. RT.

    This makes the Wiki entry inconsistent because it says "The best efficiency is achieved at room temperature." and there is no reference for that statement at the end.

    Figure 2 is the experiment to look at DNA recovery after incubation at cold temps, comparing 0 minutes vs. 10 minutes at 0C. Most people precipitate DNA longer than 10 minutes at 0-4C. The centrifugation experiment in figure 3 goes out to 30 minutes at 4C but not the incubation experiment. The best comparison would have been to incubate 1 hr at 4C vs. 10 minutes to see if there is a true difference. One could argue that the samples in experiment 3 were incubated at 4C longer in the 20-30 minute spins (because the centrifuge was chilled) and this may have enhanced the recovery.

    In experiment 1, where all temps seem to recover the same amount of DNA at all temperatures, the centrifugation was only 10 minutes, which experiment 3 shows is obviously not optimal. So it is possible that the colder temps at -20C and -70C would have recovered more DNA than the 4C temperature if the centrifugation was longer. Ok- unlikely, but possible. We don't really know what the outcome will be until we do the experiment.

    I do think this is a really good, interesting article. I am sure our readers will find it helpful.

    Thanks very much for reading and commenting!

  3. Marc says:

    Thank you guys so very much, I have spent the last 2 years wondering about this issues. This website is great! Thanks! Cheers from Geneva!

  4. Jim H says:


    No, I'm Jim Hardy. Dr Hartley is now a PI at NCI-Frederick, with a number of other brilliant colleagues who "left" when Invitrogen shut us down. He's doing absolutely remarkable work on cell free protein expression systems and I continue to see him often. [see http://fredcobio.wordpress.com/2008/11/25/nature-methods-a-la-frederick/

    I worked for Jil, who sadly hung up her lab coat for a career in Marketing. I would love to see someone reproduce or improve on this experiment, after 20 years!

  5. John Mackay says:

    There was also a great paper in methods of enzymology (late 80's I think?) that discussed the different salts used. From memory it also described that cold temp wasn't required and centrifuge time (esp for NH4Ac – 30 minutes) was the issue.

    I'm interested to know what centrifuge times most people use for their precipitations? Many protocols are just 10 minutes but habit (ha!) usually sees mine at 15 or 20 mins (except for 30 mins for NH4Ac)

  6. Marc says:

    I for one have no other choice but to centrifuge precipitated DNA (maxiprep, isopropanol, total volume ~25ml) at less than 5000g, so I let the centrifuge run for at least 1 hour, 20°C, and usually get good amounts of DNA…except for some plasmids.
    Actually I wish there were some tutorial on maxiprep optimization in this website. I have tried the PDM medium as suggested here: http://bitesizebio.com/2008/04/28/pimp-your-plasmid-growth-medium/
    It allows for more bacteria to grow (per ml), and thus yield to more DNA…but also more protein, RNAs…
    I am working in the lentiviral vector field and have noticed that low maxiprep yields (<400ug) usually lead to poor transfection (because of the level of “contamination” I assume), so I am looking for a way to get high amount of highly pure DNA…
    It is probably not the right place to ask for this, but if you have any hint, I’d be glad to hear it.

  7. Hi Marc,
    If the plasmids are low copy, you can try a protocol used for large constructs such as P1 clones and BACs where you start with a much larger volume, precipitate the DNA after the solution 3 step to condense it down, and then resuspend in the buffer for anion-exchange in a volume that fits in one load. I can direct you to a protocol if you need.

    Make sure you are not over growing the plasmid and you can avoid excessive carbohydrate in the final DNA. The E.coli strain has an impact on this too. You might want to change cells.

    Sometimes plasmids grow poorly in bacteria because of the insert. There are specialty competent cells you can try that help with toxic clones or unstable plasmids.

    Last- have you tried centrifuging at 4C vs. -20C? That article Jim posted said that centrifuging too cold can reduce yields because it becomes more difficult for the DNA to move through the viscous liquid. RT or 4C might be better.


  8. Dan says:

    Great article–but one other thing I've been looking for is a discussion of the fragment sizes brought down by different precipitation methods. For instance, if you have a 30 bp oligonucleotide you wish to purify, is one method better than another? What are the lower fragment size limits of each method? And while we are at it, any advice concerning perhaps concentrating ds versus ss DNA? Any thoughts? Anybody? Bueller?

  9. Hi Dan,
    I am not familiar with any studies looking at size but the current data from the papers Jim mentioned say that longer centrifugation is important for getting high recovery. Although the data is not strong, he suggests room temperature centrifugation so the DNA has an easier time moving through the liquid.
    I still will use cold incubation because it makes sense that precipitation or flocculation will be enhanced in cold.

    To purify a 30 bp nucleotide, I would use something like a sephadex resin instead of precipitation, if you can.

  10. D says:

    I've been having trouble with precipitating DNA. When I try to the isopropanol precipitation, I end up with a separation of layers. One organic (oily) phase and one aqueous phase. No amount of mixing will give me a proper DNA pellet.

    Have you ever seen this before?

  11. Hi D,
    I have heard this from others before but never experienced it. Are you doing a phenol extraction and maybe carried over some of the organic phase into the aqueous? Might you be taking the wrong phase of the extraction (it should be the upper and not the lower)?
    Tell us more about what you are doing?


  12. DK says:

    Cooling during alcohol precipitation of DNA does one thing very reliably though:

    It makes your DNA dirtier. With cooling, one ends up with more protein in the pellet.

  13. Ka says:

    Hi there,

    I have a basic doubt. After transcribing cDNA from total RNA with Omniscript kit Qiagen)to perform real time PCR, do I need to precipitate it? If I do, will the pellet also contain RNA fragments and DNTPs, that were left from the RT reaction, together with the cDNA? If that is true, the quantification of cDNA by Nanodrop is superestimated, right?

    Thank you for the tips and help.
    Best wishes from Brazil,

    1. You don't need to precipitate it after RT. You should go direct into the PCR step and use not more than 10% of the reaction volume with the RT reaction. The only reason to clean up the RT reaction would be to quantitate it first. If you simply precipitate the RT reaction without any clean up to remove protein, you will co-precipitate some of the reactants you don't want. The easiest way to clean up the RT is to use a commercial PCR clean up kit or to use phenol-chloroform first and then precipitate.
      I don't know if the dNTPs and RNA fragments would co-isolate with the cDNA during the precipitation but I would think that if you centrifuge it long and hard enough, you will get some of it coming down with your cDNA and if there is enough nucleic acid present, the small stuff will co-precipitate with the large stuff and they'll stick together and all pellet together.
      So better to remove it before precipitation to be sure.

  14. Hi Suzanne,
    I've been working with RNA for a couple of months now and couldn't seem to make a single protocol fit for the isolation of RNA of different plant samples. The method which uses LiCl precipitation would give me a good curve (and a good 260/280 ratio) with max. abosorbance at 260 nm for one plant sample. Other samples give out a peak at 265 – 269 nm, and still others would have no peak at all. Some of the pellets I get would be brown colored.. I usually use 100-300 mg of homogenized leaves as my starting material…I tried using RNA isolation kit based on the Chomczynski and Sacchi protocol but this gives me a constant peak at 269nm with poor 260/280 ratio for my samples.
    Do you have any suggestions for this results? Thanks. :-)

  15. Avatar of Jorge M Jorge M says:

    Hi Suzanne,

    Nice blog.  I am having some trouble with purifying viral dsDNA.  Is there any significant different in solubility between this kind of DNA an mammalian?
    For example, last attempt I saw a lot of "DNA" after adding ethanol (+1/10 NaAcetate).  I centrifuged it for one hour, and the pellet was tiny.  I centrifuged further for 4 extra hours, and I got much more DNA.
    Comments will be appreciated.
  16. Avatar of melody melody says:

    Hi! How long will it take to heat the sample in buffer at 50-60C? (for dry dna pellets)

    1. Avatar of Adam S Adam S says:

      If DNA falls out of solution at around 75%, why is it that we only add 66% final (2 volumes)?

  17. Pingback: Homepage
  18. Avatar of Heather Heather says:

    This is really useful information, thanks! My lab does not have temperature-controlled centrifuges though, and they do heat up the samples quite a lot after just 10 min at full speed. I wonder if the benefits of a 20 or 30 min centrifugation might be counteracted by the samples being heated? Is some of the DNA likely to resuspend?

  19. Pingback: URL
  20. Pingback: Cheap Oakleys
  21. Pingback: Ray Ban Wayfarer
  22. Avatar of vinod vinod says:

    Hi Suzanne,
    I am also facing some problems with BAC DNA isolation as the quality is not very high. I am using isopropanol at RT for DNA precipitation. What should i do to get pure DNA for fingerprinting.


  23. Avatar of Tiago Bruno says:

    Hello everybody. Thanks for the nice post about the differences about preciptating DNA and RNa in isopropanol and ethanol. I have a question about the cloud that appers after isopropanol addition in rna extractio. That is rna? And the strange fillaments changing the light refraction? Another useful help i would like to receive from you guys is why sometimes my rna do not preciptate in the bottom after centrifugation. It is so frustating. I make extractions from heart of mice and i should get at least 100-300 ug, but iam only getting 20-60. Thnx very much.

    1. Hi Tiago,
      Thanks for your question. I would say, yes, exactly, if you see filaments falling out of solution, that should be the nucleic acids. When that happens the yields are high. Often times you do not see it like that but I will sometimes see opaqueness forming and then it goes away.
      From heart muscle you should not be getting other contaminants such as polysaccharides forming and causing precipitation. What method are you using for extraction of RNA?
      Heart is a difficult sample type because it is a fibrous tissue. we have a post coming up very soon- this week on fibrous tissues and RNA. How much tissue do you start with? 100-300 ug of RNA from heart would have to be a very large starting amount. Tell us more about the method you use for RNA isolation and we can figure out where it is going.
      When you centrifuge to pellet the RNA, spin for 20 -30 minutes at full speed (16,000 x g or higher). What time and speed are you using?

Speak Your Mind

Nucleic Acid Purification and Analysis