Isotope Ratios and Chemical Analysis of the 1957 Brazilian Ubatuba Fragment (Powell, Swords, Rodeghier & Budinger, 2022)

Source (primary, peer-reviewed): Robert M. Powell (SCU), Michael D. Swords (CUFOS), Mark Rodeghier (CUFOS), Phyllis Budinger (SCU), "Isotope Ratios and Chemical Analysis of the 1957 Brazilian Ubatuba Fragment," Journal of Scientific Exploration, Vol. 36, No. 1 (Spring 2022), pp. 39ff. DOI: 10.31275/20222413. PLATINUM OPEN ACCESS (CC-BY-NC).
URL: https://journalofscientificexploration.org/index.php/jse/article/view/2415
Captured: 2026-06-08, full verbatim (PDF text extraction; figure/table layout not preserved). Provenance only; analysis lives on ../../sources/nolan-research-and-claims.
Why this matters / relation to the "silicon sample": this is the rigorous, published analysis of the Ubatuba material that Garry Nolan cites (the "Robert Powell paper"). It analyzes the magnesium piece (HR-ICPMS, 2017-18) and finds magnesium isotope ratios within terrestrial limits, trace-element ratios inconclusive — i.e. not demonstrably non-terrestrial — with the one oddity being extremely pure magnesium with a strontium impurity "not used in the manufacture of magnesium at the time." Nolan's separate "99.99% silicon" piece (anomalous Si isotope ratios, claimed to require extreme neutron bombardment) is a different chain-of-custody fragment, is interview-stage and unpublished (../transcripts/nolan-sol-foundation-materials-tests-2026-03-05), and should be weighed against — not conflated with — this peer-reviewed magnesium result.

Journal of
Scientific
Exploration
Anomalistics
and
Frontier
Science
39
journalofscientificexploration.org
JOURNAL OF SCIENTIFIC EXPLORATION • VOL. 36, NO 1 – SPRING 2022
RESEARCH
ARTICLE
Robert M. Powell
exploreSCU@gmail.com
https://orcid.org/0000-0001-7449-2499
Scientific Coalition for UAP
Studies, Fort Myers, Florida
Michael D. Swords
Center for UFO Studies
Chicago, Illinois
Mark Rodeghier
https://orcid.org/0000-0001-7258-0073
Center for UFO Studies
Chicago, Illinois
Phyllis Budinger
Scientific Coalition for UAP
Studies, Fort Myers, Florida
SUBMITTED January 22, 2022
ACCEPTED February 22, 2022
PUBLISHED May 22, 2022

https://doi.org/10.31275/20222413
PLATINUM OPEN ACCESS
Creative Commons License
4.0. CC-BY-NC. Attribution
required. No Commercial use.
Isotope Ratios and Chemical Analysis
of the 1957 Brazilian Ubatuba Fragment
HIGHLIGHTS
Analysis of the trace elements in a metallic sample from a 1957 ‘UFO’ were inconclusive
about its non-terrestrial origin. But tests indicated that the debris was mostly composed
of extremely pure magnesium with an odd strontium impurity. This formula was not used
in the manufacture of magnesium at the time.
ABSTRACT
A sample from the Ubatuba fragment collected in Brazil in 1957 was tested with the intent
of examining the isotope ratios of its primary element, magnesium, and the trace elements
strontium, barium, copper, and zinc. As background, the history of chemical testing of the
Ubatuba fragments during the 1960s-1980s at multiple labs with varying capabilities is
reviewed and then the remainder of the paper examines recent tests completed in 2017
and 2018 that for the first time used HR-ICPMS techniques to look at the isotopic ratios of
the minor constituents as well as the primary magnesium component of the sample. The
magnesium isotope ratios were found to fall within terrestrial limits while the results on
the isotope ratios of the trace elements were inconclusive. Recommendations are made
for improving the process of examining the trace elements.
KEYWORDS
Ubatuba, isotope, magnesium, HR-ICPMS, UFO
INTRODUCTION
The vast majority of UFO sightings involve only witness
testimony, without physical evidence that can be examined in
a laboratory. However, in a very small fraction of cases, there
has been material available for later analysis. These include
falls of angel hair (although these often evaporate rapidly),
slag-like material, powdery deposits, or physical samples
that appear manufactured. The Ubatuba, Brazil, samples,
the subject of this paper, fall into this latter category. Other
notable examples of physical traces include those in Campi
nas, Brazil (1954), Vaddo, Sweden (1954), Redding, CA (1969),
Delphos, Kansas (1971), and Carlisle, NY (1975).
The ideal UFO artifact useful for isotope testing should

be something physical in a solid state, 2) have been wit
nessed or otherwise determined to be clearly involved with
a UFO, and 3) have acceptable provenance and adequate
data collection. These criteria are not easy to meet, first
because the events themselves are rare (far less than 0.1%
of UFO reports), and also because of the lack of investiga
tive resources to meet points 2 and 3. Although the Uba
tuba samples are not conclusively tied to a UFO event, the
rarity of such a sample, the interesting circumstances of
how it became available, and the long history of prior test
ing motivated this new analysis.

40
JOURNAL OF SCIENTIFIC EXPLORATION • VOL. 36, NO 1 – SPRING 2022
journalofscientificexploration.org
ANALYSIS OF THE 1957 BRAZILIAN UBATUBA FRAGMENT
Robert Powell, Michael Swords, Mark Rodeghier, Phyllis Budinger
Sample History
The Ubatuba sample originated in Brazil in mid-1957.
It was first mentioned in a Rio de Janeiro newspaper, El
Globo, on September 14, 1957, with the title “A Fragment
from a Flying Disk!” Pieces of a supposed flying disk were
provided to the newspaper by a subscriber whose name
was illegible but wrote in a very educated manner. The
subscriber claimed that he had been fishing near the
coastal town of Ubatuba when he saw a flying disk that
was climbing rapidly just before it exploded into a shower
of thousands of fiery fragments. Most of the fragments
fell into the sea except for a few that fell on the beach and
were collected. Three samples of a very light, dull, grayish
metal that were received by the newspaper were turned
over to Dr. Olavo Fontes. Fontes had the samples tested in
Brazil before he turned the samples over to Coral Loren
zen of the Aerial Phenomenon Research Organization in
late 1957. In 1987 Lorenzen turned the samples over to
Dr. Peter Sturrock of Stanford University. The latter is the
source of the sample tested by the authors of this paper
(Sturrock, 2001).
It is not possible to know if these samples originated
from an unknown aerial explosion or from the town of
Ubatuba, Brazil. We can only be reasonably confident that
the sample originated in Brazil sometime before Septem
ber 1957 as that is when the samples first appeared at the
Brazilian newspaper office of El Globo (Sturrock, 2001).
This paper will examine the possibility of whether the
material has a possible extraterrestrial origin based on an
examination of the isotope ratios of the elements within
the sample. Independent of the history of the samples,
the isotope ratios of the elements within the sample
should fall within the range of elements that originate on
Earth. The isotope distribution of elements vary across
the planet and those ranges have been established (CRC,
1998–1999).
Bulk Chemical Makeup
Close to a dozen different labs examined the ele
mental makeup of the Ubatuba sample from the 1950s
through the 1990s. The surface chemical constituents
will be ignored due to the lack of proper handling and
contamination from contact with the ground. The bulk
chemical makeup of the material from various labs will be
reviewed as these tests involve removal of surface coat
ings prior to analysis, which eliminates the surface con
tamination issues.
Dr. Fontes arranged for the first testing of these
samples, completed in Brazil in November 1957 by a Bra
zilian government lab, the Mineral Production Lab. Tests
on the material using a Hilger Spectrograph detected no
trace elements present, indicating 100% magnesium (Mg)
as far as the equipment could determine (APRO Bulletin,
1960). This seems to be the origin of statements that
the Ubatuba sample was extraterrestrial because it was
100% pure magnesium. This was an erroneous conclusion
as the material was not truly pure Mg; the trace elements
were not detectable with the equipment used in Brazil at
the time (Sturrock, 2001).
The Ubatuba sample was next tested in September
1958 at Oak Ridge National Laboratories by chemists Dr.
Ellison Taylor and Dr. Cyrus Feldman, physicist Dr. T. A.
Welton, and metallurgist Dr. Robert Gray. The chemi
cal analysis was made by burning a small amount of the
sample with an arc and using an Applied Materials two-
meter grating spectroscope to measure the elemental
makeup. The results showed an overall purity of the Mg
measured as 99.8%. Trace elements detected included
iron (Fe), silicon (Si), and aluminum (Al) in the 100–1000
parts per million (ppm) range and calcium (Ca) and copper
(Cu) in the 1–10 ppm range. (It is worthwhile to note that
two elements that will be found in future tests, barium
(Ba) and strontium (Sr), could not be measured if below
1200 ppm.) In their report, the scientists noted that the
sample had fissures within the Mg crystalline lattice that
indicated oxidation at high temperature, supportive of
the story as to how it was found (ORNL report, 1958).
Indications of high temperature oxidation based on lat
tice inspections were also noted in later examinations
of the material by two metallurgists (Walker & Johnson,
1970; Walker, 1992). It was also noted that the form of
the Mg ruled out pyrotechnics, and the properties of Mg
ruled out this being any type of aircraft or missile that had
burned up in the atmosphere. Oak Ridge National Labs
believed it could still be man-made and that the sample
was sufficiently interesting that it warranted further
study (ORNL report, 1958).
Dow Chemical was the next laboratory to test the
samples, in December 1961. They used an electron beam
that sputtered into the sample to a distance of 1-4 microns.
The system’s sensitivity was generally 1000 ppm but with
lower levels of detection for certain elements. Results
indicated the purity of the Mg at 99.98% with traces of
Sr and Ba at 30 ppm and Ca at 100 ppm (Sturrock, 2001).
February 1968 was the year that the University of
Colorado UFO Study (Condon Committee) under the
supervision of Dr. Roy Craig evaluated the Ubatuba sam
ples. Craig used the government’s Alcohol & Tobacco
Laboratory to complete the testing. The lab removed the
surface areas of the sample with hydrochloric acid and
then rinsed it in distilled water before they utilized neu
tron activation followed by measurement of gamma ray
activity using gamma-ray spectroscopy to measure the

41
journalofscientificexploration.org
JOURNAL OF SCIENTIFIC EXPLORATION • VOL. 36, NO 1 – SPRING 2022
Robert Powell, Michael Swords, Mark Rodeghier, Phyllis Budinger ANALYSIS OF THE 1957 BRAZILIAN UBATUBA FRAGMENT
sample’s elemental makeup. The purity of the Mg was
99.9% with Sr at 500 ppm, Ba at 160 ppm, zinc (Zn) at
500 ppm, manganese (Mn) at 35 ppm, chromium (Cr) at
32 ppm, and copper (Cu) at 3 ppm. The lab’s published
measurement errors varied by element but they were all
roughly +/–20% of the measured value. (NOLAT Report,
1968). Craig noted in the Condon report that the traces of
Sr were not known to be added to commercial Mg (Condon
Committee, 1968). This lab also measured the Mg isotope
ratios, which will be discussed later in this paper.
The capability to test <100 ppm and parts per bil
lion (ppb) levels of trace elements advanced considerably
in the 1990s with the advent of the inductively coupled
plasma mass spectrometer (ICP-MS). For background,
ICP-MS is a type of mass spectrometry that uses an
inductively coupled plasma to ionize the sample. It
atomizes the sample and creates atomic and small
polyatomic ions, which are then detected . . . It
can detect different isotopes of the same element,
which makes it a versatile tool in isotopic labeling.
(Wikipedia, 2022)
Sturrock had two samples tested in March 1997 by
Elemental Research in Vancouver using a laser ablation
ICP-MS. With this new technique 37 trace elements were
detected. This is not unexpected as many of these ele
ments were in the ppb range and almost any metal tested
with such a system will have many trace contaminants.
Only the significant trace elements will be discussed.
Calcium was detected at 4600 ppm and 3230 ppm and
titanium (Ti) at 283 ppm in one sample. The Ca is much
higher than ever reported previously. It is likely that this
is due to surface contamination as laser ablation vapor
izes the sample from the surface downward. Surface
analysis tests had shown large amounts of sodium (Na),
Ca, Mn, and Ti all present in seawater, which was near
where the samples were claimed to have been found. The
more interesting trace contaminants in order of amount
were as follows: Sr (916 ppm and 568 ppm), Ba (301 ppm
and 248 ppm), Zn (27.8 ppm and 17.5 ppm), Cu (3.0 ppm
and 16.6 ppm), tin (Sn) (7.7 ppm and 11.3 ppm), and lead
(Pb) (7.1 ppm and 10.5 ppm) (Sturrock, 2001).
In summary, the Ubatuba sample’s bulk constituents
were approximately 99.8% Mg with the main two trace
metals being Sr (500–000 ppm) and Ba (150-300 ppm).
Present in smaller amounts were Zn (<100 ppm), Cu
(3–20 ppm), and both Pb & Sn (7–11 ppm). The use of Sr
as a trace element in Mg had been done by Dow Chemical
but was not considered a normal practice during the time
this sample was first obtained (Sturrock, 2001). Although
finding such a pure form of Mg with trace levels of Sr in
Brazil in 1957 was very unusual and difficult to explain,
this was not sufficient evidence to establish a non-ter
restrial origin for the sample. The next tests discussed
attempted to do that by examining the isotope ratios of
the Mg in the Ubatuba sample.
Isotopic Analysis of Mg
The value of isotopic analysis rests with the unique
isotope ratios of elements that originate in different loca
tions on earth, different time periods, other areas of our
solar system, or in other star systems. We know the range
of isotopes that can occur naturally on Earth, so unless a
material is treated in such a way as to shift the isotope
ratios of a given element, we expect it to fall within the
normal terrestrial range. Elements that originate in mete
orites, on the Moon, Mars, etc., will have slightly differ
ent isotope ratios for some of the elements such as Mg,
Cr, Sr, Ba, Ti, and Ni due to the inhomogeneous develop
ment of the solar system’s protoplanetary disk as well as
the decay of 26Al (Paton et al., 2013; Young & Galy, 2004;
Lugaro et al., 2018). A material from outside of our solar
system would be expected to have even more variation
in isotope ratios (Lugaro et al., 2018; Vangioni & Olive,
2019). The isotope ratio in and of itself cannot prove that
an object has an extraterrestrial origin but it is a clue as to
a material’s origin.
The first analysis of Mg isotopes in the Ubatuba sample
was done in February of 1968 by Craig as part of the Condon
Committee and was performed using neutron activation by
the National Office Laboratory of the Alcohol and Tobacco
Division (NOLAT, 1968). The analysis looked at only one of
the three isotopes of Mg, 26Mg. As reference, the nominal
terrestrial abundance values for these isotopes are: 24Mg
= 78.99%, 25Mg = 10.00%, 26Mg = 11.01% (the abundances
of the common isotopes of an element in a sample should
add to 100.0%, so in the Ubatuba sample, 24Mg + 25Mg +
26Mg =100.0% within measurement error). The report indi
cated that the isotope abundance of 26Mg was 14.3% with
an error of 0.7% and noted that the value “is in reasonable
agreement with 26Mg in the literature.” (NOLAT, 1968) This
statement is incorrect, and to further compound the error,
Craig, in the Condon Committee report, left out the value
obtained by the lab and simply stated (Condon Report,
1968), “. . . the Brazil sample did not differ significantly in
26Mg isotope content from other magnesium samples (p.
142).” It is difficult to understand how Craig, with a PhD in
chemistry, could have made this error, but he did. A chem
ist’s bible, The Handbook of Chemistry and Physics, lists the
nominal abundance of 26Mg as 11.01% (CRC, 1998–1999).
The terrestrial range of this magnesium isotope is 10.99%
to 11.03% (USGS, 2006). This gross error was not known

42
JOURNAL OF SCIENTIFIC EXPLORATION • VOL. 36, NO 1 – SPRING 2022
journalofscientificexploration.org
ANALYSIS OF THE 1957 BRAZILIAN UBATUBA FRAGMENT
Robert Powell, Michael Swords, Mark Rodeghier, Phyllis Budinger
until Dr. Michael Swords obtained and then reviewed the
actual papers of Craig from the Texas A&M Cushing Library
in 2008. A further error was found by Brad Sparks in 2018
when he noted that the lab calculations used the weight of
the wrong Ubatuba sample (there were two samples pre
pared) and if the correct weight had been used then the
26Mg value would be 23.1% (Sparks, 2019). It is clear that
whether the 26Mg was 14.3% or 23.1%, this first attempt at
measuring the isotope ratio was a failure.
The second analysis of Mg isotopes was done by Stur
rock in the spring of 1997 and the testing was completed by
Charles Evans and Associates using SIMS (Secondary Ion
Mass Spectrometry) instrumentation. This analysis evalu
ated all three common magnesium isotopes. The exact
values for each isotope obtained were not provided but
instead the ratios of the isotopes were displayed in a chart
that compared the ratio of 26Mg/24Mg to 25Mg/24Mg. Figure
1 (from Sturrock, 2001) shows these ratios for an Ubatuba
sample as well as four different samples provided by DOW
Chemical. Also plotted is the point corresponding to the
ratios for the nominal values of the Mg isotopes: 26Mg/24Mg
= 0.1394 and 25Mg/24Mg = 0.1266.
The last attempt to measure the magnesium iso
topes was made later in 1997 by Sturrock using Elemental
Research in Vancouver, Canada. They used a laser ablation
ICP-MS. Sturrock indicated that there were results for two
Ubatuba samples and two magnesium standards, one of
Figure 1. Plot of the determination of the isotope ratios 25Mg/24Mg and 26Mg/24Mg. From figure 1 from Sturrock (2001).
(a) DOW CP is the DOW sample of triply sublimed magnesium used in the Colorado Project; (b) DOW A is a sample of triply
sublimed magnesium supplied by DOW Chemical to Peter Sturrock; (c) SU-A is one of the Ubatuba samples; (d) Baker A is
a magnesium standard used by the Johnson Space Flight Center; (e) DOW E is another sample of triply sublimed magne
sium supplied by DOW Chemical to Peter Sturrock; and (f) the added green circle is the calculated nominal value based on
the established nominal values for the 3 magnesium isotopes.
which came from the National Institute of Standards and
Technology (NIST). However, his paper displays results
from only one of the Ubatuba samples, the two standard
samples, and states the other Ubatuba results “looks” just
like the other one. The three displayed magnesium isotope
distributions look similar, indicating a terrestrial origin for
the Ubatuba sample as shown in Figure 2 (from Sturrock,
2001). The data counts for the graphs are supplied in the
paper but not the actual magnesium isotope percentages.
The authors have converted the data counts in Figure 2
into percentage values, which are shown in Table 1. Uncer
tainties for these ratios are not available as errors were
not published with figure 2 in Sturrock (2001). Even so,
the deviations of the values of the Ubatuba sample from
the terrestrial norms are very small in magnitude and not
inconsistent with a terrestrial source (differences could be
due to the capability of the equipment used).
In summary, the three past attempts to measure
magnesium isotopes in the Ubatuba samples have been
inconclusive. We next report on more recent results using
modern equipment that allows measurement of not only
the magnesium isotopic distribution but the isotopic com
ponents of the trace metals in the sample that are in the
10–100 ppm range. This will allow for examination of more
metals than just the magnesium which makes up 99.88%
of the sample with trace metals accounting for the remain
ing 1200 ppm.

43
journalofscientificexploration.org
JOURNAL OF SCIENTIFIC EXPLORATION • VOL. 36, NO 1 – SPRING 2022
Robert Powell, Michael Swords, Mark Rodeghier, Phyllis Budinger ANALYSIS OF THE 1957 BRAZILIAN UBATUBA FRAGMENT
Figure 2. Isotope composition as determined by Elemental Research using ICP-MS (from figure 2 in Sturrock, 2001). Top
display is of the Ubatuba Mg sample SU-H; center display is Mg sample ALFA-a; and bottom display is NIST sample ISO-A.

44
JOURNAL OF SCIENTIFIC EXPLORATION • VOL. 36, NO 1 – SPRING 2022
journalofscientificexploration.org
ANALYSIS OF THE 1957 BRAZILIAN UBATUBA FRAGMENT
Robert Powell, Michael Swords, Mark Rodeghier, Phyllis Budinger
2017–2018 ISOTOPIC ANALYSIS
OF MG, SR, BA, ZN, AND CU
During the twenty years since the Ubatuba sample
was last studied, the ability to analyze isotope ratios has
improved significantly. Not only can the isotopes of Mg be
analyzed, but so can the minor constituents that are in the
100–1000 ppm range. Based on the previous analysis of
the chemical constituents of the Ubatuba sample, we iden
tified the trace elements of Sr, Ba, Zn, and Cu as targets for
isotope analysis.
Michael Swords obtained from Peter Sturrock a small
piece from an original Ubatuba sample. Cerium Laborato
ries in Austin, Texas, was identified as the lab of choice to
do the isotope analysis. Cerium has an ISO 17025 accred
ited laboratory for testing and calibration and is well-
known in the nanotechnology field. The instrument used
for isotope analysis was their High Resolution Inductively
Coupled Plasma Mass Spectrometer (HR-ICPMS) manufac
tured by Thermo Finnigan and known as the Finnigan Ele
ment2. The equipment is capable of measuring below parts
per trillion levels of trace elements. This same system has
been used to measure Pb isotopes in peat bogs (Krachler
et al., 2004), Sr isotopes in sediments (Hulme et al., 2008),
and is capable of separating most of the elemental iso
topes (Element Series, 2020).
Two of us (MS and RP) met with Dr. Tim Hossain, the
chief scientist at Cerium, along with the director of their
lab and the chemist who would be running the HR-ICPMS.
The sample and testing procedures were discussed and
agreed upon. The sample was prepared in a Class-1 clean
area, which means there is no more than one particle larger
than 0.5 microns in size per cubic meter of air. It was agreed
that 50 angstroms (5 x 10–7 cm) of the sample’s outside
surface would be removed in order to eliminate contami
nates introduced from handling. From the cleaned sample
0.00570 grams was dissolved in 1% ultra-pure nitric acid at
22 °C for 30 minutes. The Sr, Cu, Zn, and Ba were diluted
100:1 while a solution to test the Mg was diluted 10,000:1.
The latter was required due to the large amount of Mg
making up the sample. Calibration of the HR-ICPMS was
done using NIST traceable standards.
The more highly diluted Mg sample was tested first on
July 28, 2017. The results are shown in Table 2. The 26Mg
abundance of 10.58% in the Ubatuba sample was substan
tively different from the nominal abundance of 11.01%. No
standard error was provided by Cerium so we cannot con
clude that the 26Mg abundance was outside of the terres
trial norm.
The results of the isotope analysis on the lower con
centration elements were received on September 8, 2017,
and are shown in Table 3. Unfortunately, again no standard
deviation error was provided by the lab.
Strontium had the most unusual variations with the
Ubatuba 84Sr abundance at 0.74% compared to the nomi
nal value of 0.56% and the 86Sr at 9.10% compared to the
nominal value of 9.86%.
Once these results were reviewed, we planned a
second lab analysis of the same dissolved samples that
had been originally prepared by Cerium Labs. This would
provide independent verification of the Cerium results. We
had hoped to use a university laboratory for the second
analysis. Among the labs contacted were the University of
Texas, Rice University, University of Maryland, and the Uni
versity of Houston. We were unsuccessful in getting any
university lab to commit to analyzing our sample once they
asked about the source of the sample and were told it was
of an unknown source of almost pure magnesium that had
supposedly burned up in the atmosphere. Unsuccessful in
the university arena, another commercial laboratory was
chosen: ICP and ICP-MS Services in Cleveland, Ohio. The lab
was run by Dr. Arthur Varnes and used a Thermo Scientific
iCAP-Q inductively coupled plasma mass spectrometer.
TABLE 1. Mg isotope abundance for distributions shown
in Figure 2.
Isotope
NIST Standard
ISO-A (%)
Ubatuba
SU-H (%)
Standard
Sample
ALFA-a (%)
24Mg
78.99
79.25
78.76
25Mg
10.00
9.81
10.07
26Mg
11.01
10.99
11.16
TABLE 2. Ubatuba sample Magnesium isotopes abun
dances from a 10,000:1 sample dilution from Cerium
Labs with nominal abundance for comparison.
Isotope
Nominal
Abundance (%)
Ubatuba
Unknown (%)
24Mg
78.99
79.31
25Mg
10.00
10.10
26Mg
11.01
10.58

45
journalofscientificexploration.org
JOURNAL OF SCIENTIFIC EXPLORATION • VOL. 36, NO 1 – SPRING 2022
Robert Powell, Michael Swords, Mark Rodeghier, Phyllis Budinger ANALYSIS OF THE 1957 BRAZILIAN UBATUBA FRAGMENT
The ICP-MS used was calibrated based on NIST traceable
reference solutions with concentrations of Ba, Cu, Mg, Sr,
and Zn in the 1000 ppm range. The lab report from ICP-MS
Services was received on July 3, 2018.
The results of the Mg isotopic abundances will be
examined first. Table 4 shows for comparison the same
information as Table 2 plus the addition of the results from
ICP-MS Services in Cleveland, which included information
on error of measurement allowing us to calculate a 95%
confidence interval. The Cleveland results are consistent
with terrestrial abundances for 26Mg and 25Mg, although not
quite for 24Mg. As noted earlier, non-terrestrial sources can
lead to different abundances. For example, from results of
models reported in Vangioni and Olive (2019), we calculate
the abundances produced by a star with 15 solar masses
(only stars with 2 or more solar masses produce magne
sium) as 24Mg = 82.4%, 25Mg = 8.7%, 26Mg = 8.91%, quite
different from terrestrial values. Conversely, abundances
for stars in the Hyades cluster are very close to terrestrial
values because these stars are in the Local Group with the
Sun (Yong et al., 2004).
It is most useful to look at the ratio of 26Mg and 25Mg to
the dominant form 24Mg to determine possible origin. This
was first examined by Sturrock as shown in Figure 1. Figure
3 includes this same information, with additional data for
Mg isotopes from a meteorite, Dead Sea Mg, Amazon River
Mg, and the North Atlantic as well as the values obtained
on the Ubatuba sample from the Austin and Cleveland labs.
The results from the Cleveland lab fall on the same diago
nal line as other terrestrial samples while the Austin lab
falls well off the line. Since both of these labs are measur
ing the same sample solute, it is most likely that some error
occurred in the measurement of the Mg isotopes by the
Austin lab. (As comparison to a potential non-terrestrial
source, the point corresponding to the two ratios derived
from the abundances reported above for a 15 solar mass
star falls well outside the plot area.)
TABLE 3. Ubatuba sample Strontium, Copper, Zinc and Barium Isotopes abundances from a
100:1 sample dilution from Cerium Labs with nominal abundances for comparison.
Isotope
Nominal
Abundance (%)
Ubatuba
Unknown (%)
Absolute Percent
Difference*
Strontium
84Sr
0.56
0.74
32%
86Sr
9.86
9.10
8%
87Sr
7.00
7.03
0%
88Sr
82.58
83.12
1%
Copper
63Cu
69.17
70.26
2%
65Cu
30.83
29.74
4%
Zinc
64Zn
48.6
49.41
2%
66Zn
27.9
27.73
1%
67Zn
4.1
4.09
0%
68Zn
18.8
18.77
0%
Barium
134Ba
2.42
2.28
6%
135Ba
6.59
6.27
5%
136Ba
7.85
7.55
4%
137Ba
11.23
10.85
3%
138Ba
71.7
73.05
2%

Absolute percent difference = (Abs(Unknown – Nominal)/Nominal)*100

46
JOURNAL OF SCIENTIFIC EXPLORATION • VOL. 36, NO 1 – SPRING 2022
journalofscientificexploration.org
ANALYSIS OF THE 1957 BRAZILIAN UBATUBA FRAGMENT
Robert Powell, Michael Swords, Mark Rodeghier, Phyllis Budinger
The other trace isotopes measured by ICP-MS Services
are listed along with the results from Cerium Labs in Table
5. Given the results for Mg we place more weight on the
ICP-MS Services analysis. There is no consistent varia
tion between the two labs as compared to the nominal
terrestrial isotope abundance measured for the various
elements. 84Sr and 87Sr were below nominally measured
terrestrial values as measured by ICP-MS Services yet the
Cerium Lab value for those isotopes were above the nomi
nal terrestrial values. The same can be seen with some of
TABLE 4. Ubatuba sample Magnesium isotopes abundances in 10,000:1 sample dilution from Austin
and Cleveland labs with nominal abundances for comparison.
Isotope
Nominal Abun
dance (%)
Ubatuba
Unknown
Austin (%)
Ubatuba
Unknown Cleve
land (%)
Cleveland 95%
Confidence
Interval
24Mg
78.99
79.31
79.28
79.16 – 79.40
25Mg
10.00
10.10
9.94
9.64 – 10.24
26Mg
11.01
10.58
10.85
10.70 – 11.00
Figure 3. Magnesium isotope fractionation, plot of ratio of 26Mg/24Mg with ratio of 25Mg/24Mg from a variety of samples.
the Zn isotopes. ICP-MS Services measured 64Zn as below
the nominal value for the Ubatuba sample and measured
67Zn and 68Zn above the nominal values while Cerium Labs
measured 64Zn above the nominal value and the rest of the
Zn isotopes as meeting the nominal terrestrial values.
DISCUSSION AND CONCLUSIONS
One of the key lessons from this paper are the diffi
culties inherent in identifying whether the elements in a

47
journalofscientificexploration.org
JOURNAL OF SCIENTIFIC EXPLORATION • VOL. 36, NO 1 – SPRING 2022
Robert Powell, Michael Swords, Mark Rodeghier, Phyllis Budinger ANALYSIS OF THE 1957 BRAZILIAN UBATUBA FRAGMENT
sensitivity at the expense of the trace metal isotopic analy
sis. Should the Ubatuba samples be tested again, it would
be advisable to chemically separate the heavier trace iso
topes from the Mg so that all elements in the sample tested
are in the same concentration range. Low ppm levels of
trace elements can be tested effectively once separated
and there now exist even more accurate testing capability
with multiple collectors on a HR-ICPMS. These systems are
known as Multicollector Inductively Coupled Plasma Mass
Spectrometers (MC-ICPMS) and most exist at universities
that do isotope analysis.
No conclusion can be drawn from the data obtained
from Cerium Labs and ICPMS-Services as to whether the
Ubatuba sample consists of trace elements whose iso
topes do not match the normal range of terrestrial isotopic
ratios. Although both labs found isotope distributions out
side of normal terrestrial values, the lab values were not
consistent with each other.
One definite strangeness with the Ubatuba sample
does remain. All testing consistently indicates that the
Ubatuba sample is 99.88% pure magnesium with traces of
Sr, Ba, Zn, and Cu. The strontium impurity is not a normal
material are composed of isotopes that match the nominal
terrestrial isotope values. One of the primary challenges is
identifying two labs with sufficient experience in measur
ing specific isotopes so that duplicate results can be veri
fied. Although both labs were very experienced in the use
of HR-ICPMS and ICPMS systems, neither were experts in
the specific isotopes being measured. This is a real chal
lenge as most labs with experience in isotopic analysis of
specific elements are at universities. Most of these labs
have their own projects and it is difficult to get a univer
sity lab to do an analysis on an outside project especially
if the sample source is tied to a subject that is not of inter
est to the university, and perhaps as controversial as the
UFO phenomenon. As the demand for isotopic analysis
increases, hopefully there will be an increase in the exper
tise at commercial labs as these are less concerned with
sample origin.
The difficulty in obtaining consistent isotope values
in the trace metals between the two labs may be related
to the high concentration of Mg in the samples which was
three orders of magnitude greater than the trace metals.
This can lead to swamping of the HR-ICPMS detector
TABLE 5. Ubatuba sample Sr, Cu, Zn, Ba Isotopes abundance in 100:1 sample dilution from Austin & Cleve
land labs with nominal abundances for comparison. Standard deviation included for Cleveland data.
Element
Isotope
Nominal Abundance (%)
Ubatuba
Unknown
Austin (%)
Ubatuba
Unknown
Cleveland (%)
Strontium
84Sr
0.56
0.74
0.53 +/–0.01
86Sr
9.86
9.10
9.85 +/– 0.31
87Sr
7.00
7.03
6.83 +/– 0.05
88Sr
82.58
83.12
82.77 +/– 0.32
Copper
63Cu
69.17
70.26
69.14 +/–0.47
65Cu
30.83
29.74
30.86 +/–0.42
Zinc
64Zn
48.6
49.41
46.99 +/– .024
66Zn
27.9
27.73
28.00 +/– 0.13
67Zn
4.1
4.09
4.37 +/– 0.04
68Zn
18.8
18.77
19.70 +/– 0.20
Barium
134Ba
2.42
2.28
2.41 +/– 0.01
135Ba
6.59
6.27
6.65 +/– 0.03
136Ba
7.85
7.55
7.85 +/– 0.04
137Ba
11.23
10.85
11.21 +/– 0.08
138Ba
71.7
73.05
71.67 +/– 0.10

48
JOURNAL OF SCIENTIFIC EXPLORATION • VOL. 36, NO 1 – SPRING 2022
journalofscientificexploration.org
ANALYSIS OF THE 1957 BRAZILIAN UBATUBA FRAGMENT
Robert Powell, Michael Swords, Mark Rodeghier, Phyllis Budinger
by-product in the manufacture of magnesium and would
have been intentionally added. Dr. Beaman and Dr. Solaski
of DOW Chemical were surprised by the presence of stron
tium as was Dr. Couling of Battelle Labs (Sturrock, 2001). It
is still a mystery as to how high purity magnesium with the
addition of strontium impurities showed up at a Brazilian
newspaper office in 1957.
IMPLICATIONS AND APPLICATIONS
The current study illustrates both the promise, and
difficulties, of doing elemental analysis on samples from
anomalous events, whether UFO-related or otherwise. The
study aims to distinguish a possible extraterrestrial arti
fact from terrestrial ones, but this is not a straightforward
exercise. Limitations of instrument precision and the cost
of testing make definitive results difficult to achieve, even
with the improvements in analytical technology (unless a
sample has isotope ratios far outside normal abundances).
So too do developments in materials science, which con
struct ever-more esoteric materials (although this is less
of a complication with a sample from an earlier era, as with
Ubatuba).
The use of isotope analysis within the field of ufol
ogy should be expanded to analyze landing traces, and
close encounter cases generally, where odd substances
or extreme effects are seen. Such cases could conceivably
have remnant materials of abnormal isotope ratios. A con
certed focus on this type of evidence by investigators and
UFO organizations is an essential strategy.
Thinking beyond the UFO box, if any researcher
obtains unusual seeming material from incidents like
cryptozoological encounters, strange falls from the sky,
or paranormal objects such as apports, this approach can
be employed. We need more reliable, physical data on all
types of paranormal events, and purported physical sam
ples allow the most in-depth studies. However, this is lim
ited by the serious expense of these tests. They cannot be
casually applied.
REFERENCES
APRO. (1960, March). APRO Bulletin, p. 3.
Condon Committee. (1968, October 31). Condon Report.
University of Colorado, pp. 392–396.
CRC. (1998–1999). CRC handbook of chemistry and physics
(79th ed.) (pp. 1–10). CRC Press.
Element Series HR-ICP-MS. (2020). Thermofisher.

https:www.thermofisher.com/order/catalog/product/
IQLAAEGAAMFABWMAFC#/IQLAAEGAAMFABWMAFC
Hulme, S., Wheat, C., Coggon, R., McManus, J., & Expedi
tion 301 Scientists. (2008). Data report: Trace ele
ment, Sr isotope, and Ge/SI composition of fluid and
sediments in ridge-flank low-temperature hydrother
mal environments. Proceedings International Ocean
Discovery Program, 301, 1–17.
Krachler, M., Le Roux G., Kober B., & Shotyk W. (2004).
Optimising accuracy and precision of lead isotope
measurement (206Pb, 207Pb, 208Pb) in acid digests
of peat with ICP-SMS using individual mass discrimi
nation correction. Journal of Analytical Atomic Spec
tometry, 19, 354–361.
Lugaro, M., Karakas, A., Peto, M., & Plachy, E. (2018). Do
meteoritic silicon carbide grains originate from
asymptotic giant branch stars of super-solar metal
licity? Geochimica et Cosmochimisca Acta.
NOLAT. (1968, February 27). National Office Laboratory
of the Alcohol and Tobacco Division report: Neutron
Activation Analysis of Artifacts Associated with UFO
Phenomena [Letter from Assistant Chief Maynard
J. Pro to Dr. Roy Craig]. Dr. Roy Craig files, Cushing
Archival Library, Texas A&M University.
ORNL. (1958, September). Oak Ridge National Laboratories
report. Dr. Roy Craig files, Cushing Archival Library,
Texas A&M University.
Paton, C., Schiller, M. & Bizzarro, M. (2013). Identification
of an 84Sr-cepleted carrier in primitive meteorites
and implications for thermal processing in the solar
protoplanetary disk. The Astrophysical Journal Letters,
763, L40 (6 pp.).
Sparks, B. (2019). UFO encyclopedia, 3rd edition. Jerome
Clark, Volume 2: N–Z. pp. 1146–1153.
Sturrock, P. (2001). Composition analysis of the Brazil mag
nesium. Journal of Scientific Exploration, 15(1), 69–89.
USGS. (2006, March) Compilation of Minimum and Maximum
Isotope Ratios of Selected Elements in Naturally Occur
ring Terrestrial Materials and Reagents, (p. 1). US Geo
logical Survey Report 01-4222.
Vangioni, E., & Olive, K. A. (2019). The cosmic evolution
of magnesium isotopes. Monthly Notices of the Royal
Astronomical Society, 484(3), 3561–3572. https://doi.
org/10.1093/mnras/stz210
Walker, W. W. (1992). Scientific Studies of the Ubatuba
Magnesium Fragments.
Walker, W. W., & Johnson, R. W. (1970). Further studies on
the Ubatuba magnesium samples.
Wikipedia. (2022). Inductively coupled plasma mass
spectometry.
https://en.wikipedia.org/wiki/
Inductively_coupled_plasma_mass_spectrometry
Yong, D., Lambert, D., Paulson, D., & Prieto, C. (2004). Mg
isotope ratios in Hyades stars. The Astrophysical Jour
nal, 603(2), 697–707.
Young, E. D., & Galy, A. (2004). The isotope geochemistry
and cosmochemistry of magnesium. Reviews in Miner
alogy and Geochemistry, 55, 197–230.