Isotopic shifts in mercury in fluorescent light bulbs

It’s been known for some time that the ratios of isotopes of mercury in fluorescent light bulbs depart from the natural ratios. The explanation given for this effect is that the different isotopes of mercury are fractionated into different areas of the light bulb. Some of the isotopes preferentially enter into the glass tube and others remain in the bulk reservoir. A question that has come up from time to time is how solid this explanation is, and whether other possibilities such as LENR might have been prematurely ruled out. A version of this question recently came up in a thread on the Rossi v. Darden lawsuit. In reply, Paradigmnoia provided a reference to the doctoral dissertation of Chris Mead, in which he reported that (1) he saw significant changes in ratios in the isotopes of Hg above their natural abundances, as seen in earlier studies; and that (2) he observed that whole-lamp ratios for Hg were not significantly different from the natural ratios. Mead’s dissertation appears to describe work presented several years earlier at a 2010 meeting of the American Geophysical Union.

How was this research carried out, is it solid, and what conclusions can be drawn from it? Is this research adequate to contraindicate transmutations via LENR as a possible explanation for the isotopic changes in Hg in fluorescent light bulbs, either in general or in this specific instance? Are there flaws or unexamined assumptions that weaken the conclusions? How does this study compare to other studies?

One interesting detail to come to light in the Mead work is that there are two kinds of fractionation — mass-dependent fractionation and mass-independent fractionation. Mead saw no straightforward relationship between isotope mass and extent of fractionation (p. 27), and the pattern of fractionation was entirely different from that reported in a number of earlier studies. Mead proposes a mechanism for this mass-independent fractionation in which the abundance of an Hg isotope is inversely proportional to its fractionation and relates the amount of "trapped Hg" (Hg trapped in the glass) to the amount of Hg in the bulk reservoir.

Twelve 14-watt bulbs were used in all, 3 as blanks. Whole-lamp assays were carried out with 2 bulbs, one with 0 hours of operation and the other with 3600 hours; trapped-Hg assays were done with 5 bulbs; and bulk-Hg only assays were done with 2 bulbs. Following is a table from the dissertation that summarizes the results (p. 29):

My first complaint: there was appears to have been only one whole-lamp assay that was done on a bulb that had any operation time.

To emphasize another detail: the pattern of fractionation seen in Mead's study was very different from that observed in other studies:

Quote from Mead, p. 27,

The trapped Hg of used CFL show unusually large isotopic fractionation (Figure 3.2, Table 3.1), the pattern of which is entirely different from that which has been observed in previous Hg isotope research aside from intentional isotope enrichment (Bergquist & Blum, 2007; Biswas, Blum, Bergquist, Keeler, & Xie, 2008; Estrade, Carignan, & Donard, 2011; Estrade, Carignan, Sonke, & Donard, 2009; Gehrke, Blum, & Marvin-Dipasquale, 2011; Ghosh, Xu, Humayun, & Odom, 2008; Jackson, Whittle, Evans, & Muir, 2008; Laffont et al., 2009; Sherman, Blum, Keeler, Demers, & Dvonch, 2012; Sherman et al., 2009; Sonke et al., 2011; Zheng & Hintelmann, 2009, 2010a, 2010b).

Just to be clear, I not really arguing for LENR in this case or in most cases of Hg fractionation; I'm wondering whether it might not be worth discussing and examining the moving parts on this forum, and playing a little devil's advocate, as Hg fractionation bears some resemblance to results in many LENR transmutation studies.

Quote from Paradigmnoia

How much energy can be expected to be made from bumping Hg isotopes around in the manner that they are demonstrated to occur in the glass (or the reservoir)?

(if it were LENR)

Just my personal guess: any LENR would not entail fusion, i.e., from bumping Hg isotopes around. Any changes would occur from induced fission or induced alpha decay, decreasing the overall amount of Hg differentially for different isotopes. Others who take LENR seriously will have their own take on things, of course.

Also, there's a missing ingredient which might or might not be important: hydrogen and/or deuterium.

Reference for paper discussed in THH's post:

Chris Mead, James R. Lyons, Thomas M. Johnson, and Ariel D. Anbar, "Unique Hg Stable Isotope Signatures of Compact Fluorescent Lamp-Sourced Hg," Environ. Sci. Technol., 2013, 47 (6), pp 2542–2547, DOI: 10.1021/es303940p.

http://pubs.acs.org/doi/abs/10.1021/es303940p

The 2013 Environ. Sci. Technol. paper appears to be a restatement of the research discussed at the 2010 conference and in the 2014 dissertation.

While the anti-correlated trends in isotopes for lamps B, E and F (in-glass assays), and G and H (whole-lamp assays), are striking, note that the graphs show only a subset of the lamps and do not include lamps A (whole-lamp) and C and D (in-glass). There is, then, a danger of cherry-picking here, on top of the challenges presented by the small sample size (12 light bulbs, 9 in operation). And indeed, lamp D is seen to be something of an inexplicable outlier (p. 2544).

Since only ~ 2 percent of the Hg leaves the amalgam pellet (1 percent in vapor phase and 1 percent trapped in the glass), a mechanism that operated effectively in the vapor phase and ineffectively if at all in the amalgam pellet would be expected to result in large changes in the in-glass assays and trivial changes in the whole-lamp assays. Correlation is not a panacea and can be seductive but dangerous. If the anti-correlated trend in relative amounts of isotopes between in-glass and whole-lamp assays was spurious, we could still surely draw conclusions about something happening in connection with the in-glass assays but perhaps nothing happening with regard to the whole-lamp assays. Table 1 from the paper gives the 2-sigma values for the delta calculations, to which the whole-lamp values are without exception smaller than or equal in magnitude (i.e., of questionable significance), while the in-glass assays are generally greater (i.e., are significant).

So the shifts in the trapped-Hg assays seem significant, while those in the whole-lamp assays are somewhat in the noise (<= 2 sigma).

In addition, the destructive assays used (both in-glass and whole-lamp) will not distinguish between in situ fractionation of Hg isotopes, on one hand, and a LENR-mediated differential depletion of Hg across all isotopes within the reservoir, on the other.

The cap-delta calculation was used for the graphs showing the anti-correlation. This calculation imports assumptions about the underlying mechanism from the mass-dependent fractionation (MDF) theory in the form of isotope-specific scaling factors, including a change in sign in one case (see p. 2543). By contrast, the delta notation makes fewer assumptions and merely compares the amount of an isotope to the NIST 198Hg amount. I am curious what these two graphs would look like using the delta notation values, although I assume the anti-correlation would remain.

The very different pattern seen by Mead et al. in comparison to previous studies indicates at minimum that the present study is not representative. They further write: "The observed pattern of fractionation is characterized by enrichment in all isotopes relative to 202Hg, the most abundant isotope. This pattern is consistent among the different CFL measured, indicating that there is no systematic trend in the magnitude or pattern of fractionation with time." Lack of a systematic trend over time is a surprising conclusion to draw if one assumes that there is a systematic process of fractionation at work.

A nice way to distinguish between the standard MIF/MDF hypotheses and the particular LENR depletion hypothesis I suggested above would be:

• There is nothing in MIF/MDF to suggest that additional impurity elements will collect in the bulb. If there are lots of impurity elements, this would be something to explain away; e.g., fluorescent light bulbs are amazingly dirty things or something. So all else equal, the expected result from MIF/MDF would be that after many hours of operation there is still basically just mercury together with the other elements known to exist at the start.
• By contrast, if Hg atoms are being induced to alpha decay and/or fission, the expectation is that there will be other impurity elements lighter than Hg as well as possibly helium, in proportion to hours of operation and in proportion to the shifts that are seen in isotopes of mercury.

This prediction does not deal with other LENR possibilities, but one can come up with a similar set of predictions on a case by case basis.

It seems that one of the primary failure modes of fluorescent lamps is a loss of mercury, presumed to occur through adsorption to the glass. But if we go with Mead et al., above, only ~ 1 percent of the mercury will be trapped in the glass, and they were doing 10,000+ hour runs (= 13+ months). I am not sure how to reconcile these two propositions.