Adherents of the titanium argument say the Moon's surface looks manufactured. Apollo samples, they point out, held far more titanium than comparable rocks on Earth, a metal humans prize for jets and spacecraft because it is light, strong and hard to refine. Add reports of brass, mica and uranium-236, an isotope they describe as a reactor by-product that does not occur naturally, and the lunar dust reads to them like the debris of an engineered hull. The measurements behind each claim are published. The question is whether titanium in lunar rock and a uranium isotope measured in a laboratory need an industrial explanation, or have ordinary ones on record.
The titanium itself has been mapped from orbit. Europlanet reported in October 2011 on a map built by Mark Robinson's team from around 4,000 Lunar Reconnaissance Orbiter Wide Angle Camera images (https://www.europlanet.org/epsc-dps-201114-subtly-shaded-map-of-moon-reveals-titanium-treasure-troves/). The method was calibrated against Apollo and Luna samples. "In the mare titanium abundances range from about one percent to a little more than ten percent. In the highlands, everywhere TiO2 is less than one percent." Robinson said scientists "still don't really understand why we find much higher abundances of titanium on the Moon compared to similar types of rocks on Earth," and that the richness implies the Moon's interior had less oxygen when it formed. The titanium sits mostly in the mineral ilmenite, "a compound containing iron, titanium and oxygen."
The camera team's data release states the limits. The Lunar Reconnaissance Orbiter Camera page for the WAC TiO2 Abundance Map (https://data.lroc.im-ldi.com/lroc/view_rdr/WAC_TIO2) describes TiO2 in weight percent derived from the ratio of two ultraviolet bands, using about 124,300 images taken from January 2010 to May 2013. The ratio map was converted to abundance "from a linear correlation between the TiO2 contents of returned lunar soil samples" and the ratio at each sampling site. Areas below the detection limit of 2 weight percent are set to 1, and the page warns that values above 15 percent "should not be trusted." The product cites Sato and colleagues in Icarus, 2017. These are figures for titanium dioxide in soil, not metal, and the same calibration ties every orbital value back to a returned sample.
A single rock shows what "ten percent titanium" means. NASA's Lunar Sample Compendium entry for Apollo 11 sample 10017 (https://curator.jsc.nasa.gov/lunar/lsc/10017.pdf), compiled by Charles Meyer in 2011, describes "a fine-grained, vesicular, ilmenite basalt of the high-K variety," 973 grams, crystallised about 3.6 billion years ago. Its chemical analyses give TiO2 at about 11 to 12 percent. The mineral counts put ilmenite at roughly 15 to 24 percent of the rock by volume, alongside pyroxene and plagioclase. The titanium is real and abundant, and it is bound in an oxide mineral inside a volcanic rock that cooled from lava. No analysis in the entry reports titanium metal.
Ilmenite is also the ordinary starting point on Earth. The U.S. Geological Survey's Mineral Commodity Summaries 2024 entry for titanium mineral concentrates (https://pubs.usgs.gov/periodicals/mcs2024/mcs2024-titanium-minerals.pdf) reports that one U.S. company recovered "ilmenite and rutile concentrates from its surface-mining operations" in Georgia and Florida in 2023. More than 95 percent of these concentrates went to titanium dioxide pigment, with the rest used for welding-rod coatings, carbides, chemicals and titanium metal. Most titanium mined in the United States becomes pigment for paint, paper and plastics, not aircraft. A titanium-rich lunar basalt is closer to an ore body than to a refined hull.
The uranium-236 report is real, and its history is documented by the scientists who made it. P. R. Fields, H. Diamond, D. N. Metta, D. J. Rokop and C. M. Stevens of Argonne National Laboratory published "237Np, 236U, and other actinides on the moon" in the Proceedings of the Third Lunar Science Conference in 1972 (https://articles.adsabs.harvard.edu/pdf/1972LPSC....3.1637F). They reported "widely varying amounts" of 236U in samples including 12070, and argued "that lunar 236U is produced mostly by solar flare protons irradiating the indigenous 238U." They also cited a separate paper reporting 236U in terrestrial uranium. The same mechanism predicted neptunium-237, which "was sought and found in sample 12070." Their explanation was natural irradiation, not a reactor.
Four years later the same group revisited the measurement. Fields, Diamond, Metta and Rokop's "Airborne 236U and its effect on lunar sample analysis," published in the 1976 Lunar Science abstracts volume (https://articles.adsabs.harvard.edu/pdf/1976LPI.....7..251F), says blanks were clean "until late 1972 when a blank run concurrently with Apollo 16 samples showed the presence of 236U contamination." A search "did identify 236U in the laboratory air." Cheesecloth dust-stops on an air intake were high in it, and so was outside air nearby. They concluded "that the 236U content of laboratory air varied with time," and they listed anomalies in their earlier data: neptunium was found in one portion of sample 12070 but none in another portion from the same box. The isotope ratios, they wrote, were "perhaps consistent" with "residues from atmospheric testing" rather than a reactor-irradiated source.
Laid out in order, the titanium record is a story about ilmenite: mare basalts carry roughly one to more than ten percent titanium dioxide, bound in an iron-titanium oxide, which is also what miners on Earth dig up. The uranium-236 was first explained as natural irradiation and later found in laboratory air. The brass and mica claims come without a cited sample report.
What Stays Contested
Why the Moon's maria are so rich in titanium is still an open question in lunar geochemistry, as Robinson's remark shows. The industrial reading answers it with a builder. The sample analyses answer it with an oxygen-poor interior and a common mineral. Until an analysis reports refined metal or alloy in a lunar sample, the hull reading has no measured sample behind it, while the ilmenite reading rests on catalogued, analysed basalts such as 10017.
