I'm looking for the half-life of dNTPs, either as a whole or broken into individual bases, at 95 °C (or similar). A titration would be great if that exists. I can provide more specifics if need be, but broadly the idea would be 'able to polymerize' vs. 'unable'.

  • 1
    $\begingroup$ Interesting question. FWIW, I was able to get strong bands on a 6 hour long PCR reaction with Pfu polymerase, so it is unlikely that dNTP stability would be a very big concern under most circumstances. $\endgroup$
    – March Ho
    Aug 30, 2016 at 20:55
  • $\begingroup$ @MarchHo Thanks for the insight. To confirm, when you say 6 hour long PCR, that is the total length of the program, and not a total time of 6 hours at 95C (or 98C perhaps with Pfu)? $\endgroup$
    – metaditch
    Aug 30, 2016 at 21:11
  • 1
    $\begingroup$ The total time was 6 hours long, most of which was extension at 72C. I would say the time taken at 98C was roughly 25 minutes (30s per cycle) $\endgroup$
    – March Ho
    Aug 30, 2016 at 21:13
  • 7
    $\begingroup$ I think almost every molecular biologist must have wondered about this (as we always assume that (d)NTPs are unstable) but there seems to be no study that checks the stability of (d)NTPs at different temperatures. The closest I could find is this article which says that 40% of ATP survives after 1h at 95⁰C (10mM in distilled water). Mg²⁺ would increase the stability and stability of dATP may probably be higher. $\endgroup$
    Aug 31, 2016 at 6:02

1 Answer 1


Instead of searching for the stability of dNTPs, consider the stability of their constituent parts.

Caveat: ignoring the bonds between the different parts of dNTPs likely misrepresents the stability of the intact molecule, given that the covalent bonds between the nucleobase, sugar, and phosphate pieces each have their own relative stabilities. See my related Chemistry.SE post: Do the phosphoester and glycosidic bonds of dNTPs increase thermal stability?

Hydrolysis of Purines and Pyrimidines

Purines and Pyrimidines

In The stability of the RNA bases: implications for the origin of life,1 there is an Arrhenius plot showing the rate of decomposition of the most common nucleobases (A,T,C,G,U) at pH 7 in 50 mM phosphate buffer. According to ThermoFisher, the pH of PCR buffer is between 8 and 9.5, though the authors of the paper linked above observe that the pH rate profile for the decomposition of purines (A,G) is relatively flat between pH 5 - 10.

At 95° C, the least stable nucleobase has a half life of 10-1 years, or about 1 month.

Arrhenius plot for the decomposition of common nucleobases at pH 7

Fig. 2. Arrhenius plot for the decomposition of A, U, G, C, and T, pH 7. ∆, data from Garret and Tsau (10). Equations are as follows: log k(A) = -5902/T + 8.15; log k(U) = -7649/T + 11.76; log k(G) = -6330/T + 9.40; log k(C) = -5620/T + 8.69; log k(T) = -7709/T + 11.24. [red lines added for emphasis]

Decomposition of Pentose Sugars

pentose sugar

The same research group investigated the stability of the sugar moieties of nucleic acids in Rates of decomposition of ribose and other sugars: implications for chemical evolution.2 The analysis is slightly obscured by the use of deuterated water, though the authors observe that the rate of decomposition of ribose in H2O at 100° C and pH 7.0 is identical to that in D2O at 100° C and pD 7.4.

At 100° C and pD 7.4, ribose, the sugar component of RNA, has a half life of less than 2 hours. Interestingly, the authors note that 2-deoxyribose, the sugar component of DNA, decomposes 2.6 times slower than ribose with a half life just shy of 4 hours.

Stability of ribose and other sugars in water, along gradients of temperature and pH

Fig. 1. Rates of ribose decomposition as a function of pD. The curves were fitted by assuming a functional form k = k0/(1 + [H+]/Ka), where Ka and k0 are adjusted constants. [red lines added for emphasis]

Hydrolysis of Triphosphate


Getting to the heart of your question, i.e. at what point are dNTPs no longer "polymerizable", the stability of the triphosphate is vital, as the hydrolysis of the two distal phosphates and the creation of a phosphodiester bond with the remaining proximal phosphate is the rate-limiting step in DNA polymerization.

An older publication, The Hydrolysis of γ-Phenylpropyl Di- and Triphosphates,3 investigates the hydrolysis of nucleotide analogs at 95° C, over a pH range of 0.3 to 10. For the ATP analog, the rate of spontaneous hydrolysis to ADP is very low, 17.5 × 10−5 s−1 at pH ~5. Assuming first order kinetics, the half life of the triphosphate form is ~4000 seconds, or a little more than an hour at 95° C. In a buffered solution closer to a neutral pH, we can expect triphosphate to be even more stable.

Table 5. A Comparison of the Rates of Hydrolysis of γ-Phenylpropyl Di- and Triphosphates and ADP and ATP at 95° rate constants for the hydrolysis of ATP to ADP

A worked example

A PCR with one 2-minute initial melting and 20 cycles with 15-second melt steps will expose dNTPs to 420 seconds at 95° C. Considering only the hydrolysis of triphosphate, which is the least stable component of the 3-component dNTP molecule outlined above, the equation for exponential decay can be used to calculate the proportion of unincorporated dNTPs that are still "polymerizable" at the end of cycling, using the rate constant provided in Miller and Westheimer.3

$N_{\text{final}} = N_{\text{initial}} × e^{-kt}$

$N_{\text{final}} = e^{(-17.5 × 10^{-5} s^{-1})(420 s)} = 0.929$

For some perspective, a 50 μL PCR with 0.2 mM dNTPs contains enough raw materials to create 1.51 × 1013 copies of a 200 bp dsDNA amplicon, ignoring the contribution of primers to the amplicon body, and assuming perfect fidelity of the polymerase and primers in excess. A 20-cycle PCR with 106 molecules of template will yield 1.05 × 1012 amplicons, assuming perfect duplication at each cycle, meaning dNTPs are supplied in more than 10-fold excess and heat decomposition is unlikely to have an effect on PCR fidelity in this system.


  1. Levy M, Miller SL. The stability of the RNA bases: implications for the origin of life. Proc Natl Acad Sci USA. 1998 Jul 7;95(14):7933-8.
  2. Larralde R, Robertson MP, Miller SL. Rates of decomposition of ribose and other sugars: implications for chemical evolution. Proc Natl Acad Sci USA. 1995 Aug 29;92(18):8158-60.
  3. Miller DL, Westheimer FH. Hydrolysis of Gamma-Phenylpropyl Di- and Triphosphates. Science. 1965 Apr 30;148(3670):667.

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .