6

This URL lists the mass of Copper-63 as 62.9295975(6) and this other URL lists the mass as 62.939598. These values differ by almost exactly 0.01 which seems hard to explain by experimental error. Why is it that these values differ in a significant digit but have the same less-significant digits? Is one of them a typo of the other? What is the correct value? What is the origin of both of these values?

This discrepancy was noted by commentators in this article about a supposed cold-fusion reactor. According to those commentators, this value is relevant to the cold-fusion debate because it makes all the difference as to whether or not the supposed reaction is energetically feasible. The linked article concerns the same cold-fusion claim discussed in this previous physics.SE question.

EDIT: user9325, voix, and user3673 have indicated that the correct answer is 62.929... I have started a bounty on the origin of both the correct and incorrect values.

Community
  • 1
Dan Brumleve
  • 478
  • One is presumably a typo as both values agree in all figures of significance except the hundredths digit. – dmckee --- ex-moderator kitten May 21 '11 at 01:15
  • 1
    Funny that they are both on NIST's domain. – dmckee --- ex-moderator kitten May 21 '11 at 01:16
  • 1
    Wolfram alpha likes the 62.929 figure and credits http://amdc.in2p3.fr/web/nubase_en.html as the source – dmckee --- ex-moderator kitten May 21 '11 at 01:19
  • 1
    http://www.google.com/search?q=%2262.939598%22 has 169 results and http://www.google.com/search?q=%2262.929598%22 has 48 results. Disturbingly, Cu-63 is also referenced by 3 of the 7 results at http://www.google.com/search?q=%2262.919598%22! – Dan Brumleve May 21 '11 at 05:26
  • May I ask you which of the two numbers makes cold fusion feasible? – Phira May 21 '11 at 08:25
  • Apparently the "2" value. See the comments by JoulesBurn and RockyMtnGuy at the article I linked to. http://www.theoildrum.com/node/7942 – Dan Brumleve May 21 '11 at 08:32
  • 1
    Editing data collections is a painstaking business.

    For that reasons books like CRC Handbook, D'Ans-Lax or Landolt-Börnstein

    were rather expensive and time-consuming in editing and correction.

    And, that business is almost dead :=(

    The underttaking of NIST, therefore is very deservable, but as You see in this example, full of pitholes.

     – Georg May 21 '11 at 10:15

4 Answers4

10

Okay, so I did some poking around and the 66th-75th editions of the CRC Handbook of Chemistry and Physics all have the incorrect atomic mass of Cu-63 [62.939598], and from 76th edition on they seem to have figured it out.

Those isotope mass tables are put together from a number of sources, so it's hard (time consuming) to tell exactly where the error came from. I did notice however that starting in the 76th edition of the CRC, where they get it right, they start citing G. Audi and A.H. Wapstra, "The 1993 atomic mass evaluation", Nuc. Phys. A 565 (1993) 1-65. In editions 66-75, they were citing Audi and Wapstra's "The 1983 atomic mass table" which appeared in Nuc. Phys 432, 1 (1985).

Now, I looked at the 1993 version, and it has the correct 62.929.. mass, but I have not been able to find Wapstra and Audi's 1983 version of the same table, so I don't know if it was one error by the CRC which got carried over year after year, or if 62.939.. is in fact the value given in that paper.

I did find at least one piece of evidence which points to an error by the editors of the CRC in the 2nd edition of the Encyclopedia of Physics [Lerner and Trigg, 1991]. Their table of isotopes lists the correct 62.929... value and also cites Wapstra and Audi's 1983 paper.

I hope that satisfies everyone's curiosity, because I don't think I can do any better then that. ;-)

C Earnest
  • 677
8

Since both references give the same percentages and the same overall atomic weight, an easy calculation shows that this only works out for the number in the first link, therefore the second decimal should be 2.

(And I think it would be nice to contact the webmasters of the second link.)

Phira
  • 2,072
  • I am hoping the question can be answered with a reference to an original source. – Dan Brumleve May 21 '11 at 07:49
  • I am sure it can, but not by me, but I thought that it is still useful to be able to rule out one of the to sources, because now one could follow the reference trail of the other number. – Phira May 21 '11 at 08:02
  • I am also interested in the reference trail of the incorrect number. – Dan Brumleve May 21 '11 at 08:16
  • Dear @Dan, this is not a server about the history of administrative errors and typos in the least important numbers in the Universe. This is a server about physics. User9325's answer is much more solid answer than what you could get by any "original source". There is clearly one typo somewhere - someone did it and others copied it. It doesn't affect anyone except for those who find it important and they will surely check. – Luboš Motl May 26 '11 at 05:20
  • 1
    Luboš, if you read the comment threads that I have linked to, you will see that people are attempting to debunk this cold-fusion claim on the basis of this typo. Whether or not you believe in this nonsense it paints physicists and the discipline of physics in a poor light. My motivation in asking this question is to provide a better reference than NIST for this value so that this mistake will not be repeated in future discussions. I wholeheartedly believe that SE is capable of this. – Dan Brumleve May 26 '11 at 05:36
  • Furthermore I have a personal interest in learning more about how to do this sort of research in general. C Earnest's answer is my favorite so far. user9325's answer simply shows that one of the URLs contradicts its own extra data, but that is no reason to accept the 62.929 value; it is only a reason to reject the other. On the other hand, all the other answers support the 62.929 value with secondary sources. – Dan Brumleve May 26 '11 at 05:56
3

The 2003 Atomic Mass Evaluation: Cu(63) - 62.929597474

The 1995 Update to Atomic Mass Evaluation: Cu(63) - 62.929601079

The 1993 Update to Atomic Mass Evaluation: Cu(63) - 62.929600748

voix
  • 2,084
  • How to find the origin of any of these numbers? What were the experiments, who conducted them, where, when, etc.? I've tried to follow the references myself but I hit a dead end. – Dan Brumleve May 24 '11 at 06:36
  • Googling the 2003 value http://www.google.com/search?hl=en&q=%2262.929597474%22 I found near the top another discussion thread in which someone has used the 62.939 value to argue against the feasibility of this same cold-fusion: http://tickerforum.org/akcs-www?post=177539&ord=2364030#2364030 – Dan Brumleve May 24 '11 at 06:57
  • 1
    @Dan - A.H. Wapstra, G. Audi and C. Thibault "The AME2003 atomic mass evaluation" - part I, part II – voix May 26 '11 at 17:15
2

I've started a discussion on this at the Wikipedia article.

If I had to guess at the origin of this discrepancy it is a typo in one of the old CRC's where someone typed '3' instead of the correct digit of '2'. If that's the case there may have been merely an internal memo correcting the typo -- if that.

James Bowery
  • 1,327