New Scientist - USA (2021-11-20)

(Antfer) #1
20 November 2021 | New Scientist | 13

AN ANCIENT diamond found
in Botswana contains a never-
before-seen mineral that came
from deep inside Earth. The
mineral – named davemaoite –
provides a unique window
into deep-Earth chemistry.
Originally dug up in a mine in
Orapa, Botswana, the diamond
is about 4 millimetres wide and
weighs 81 milligrams. A dealer
sold it in 1987 to a scientist
at the California Institute of
Technology who was studying
diamonds, but neither the
dealer nor the scientist had
any idea how special it was.
The diamond, which is now at
the Natural History Museum of
Los Angeles County in California,
was more recently analysed
by Oliver Tschauner at the
University of Nevada, Las Vegas.
Most diamonds form 120 to
250 kilometres underground,
but some are born in Earth’s
lower mantle, which begins
660 kilometres below the
surface. Tschauner and his
colleagues realised that the

Botswanan diamond was
from the lower mantle. After
examining it using X-rays,
they discovered tiny crystals of
another mineral trapped inside.
They used a laser to cut into
the diamond to extract these
crystals and see what they
were made of. The crystals
turned out to be a form of
calcium silicate that was
theorised to exist in the lower
mantle but had never actually
been observed before. The
molecules within it take on a
particular cubic arrangement
known as a perovskite structure.
The atomic composition
of this particular perovskite –
which primarily contains
calcium, silicon and oxygen –
shows it could only have formed
under the extreme conditions
experienced in the lower
mantle, where the pressure
is more than 200,000 times
that found at Earth’s surface.
Tschauner and his colleagues
named the new calcium
silicate mineral davemaoite in
honour of deep-Earth scientist
Ho-Kwang “Dave” Mao at
the Carnegie Institution for
Science in Washington DC

(Science, doi. org/g5x4).
Normally, davemaoite’s
crystal structure would break
apart if it was brought up to
Earth’s surface because of the
massive drop in pressure. But
because it was trapped inside a
rigid diamond, it was preserved
on its long journey up to the
Orapa mine, which probably
took between 100 million to
1.5 billion years.
“When we broke open
the diamond, the davemaoite
stayed intact for about a second,
then we saw it expand and
bulge under the microscope
and basically turn into glass,”
says Tschauner.

Davemaoite is thought
to make up about 5 per cent
of Earth’s lower mantle, and
is important because it is
theorised that the mineral
can also host radioactive
elements like uranium,
thorium and potassium-
that heat Earth as they decay.
“Without these radioactive
elements, the Earth would
have cooled by now,” says
David Phillips at the University
of Melbourne in Australia.
Tschauner and other
researchers are continuing
their hunt for super-deep
diamonds in the hope of
finding more hidden minerals.
This is an arduous process,
since there are no simple ways
of distinguishing between
diamonds with shallow or deep
origins, and no way of knowing
where the deep variety will be
found. “The issue is, we don’t
know quite where to look,”
says Tschauner.  ❚

“When we broke open the
diamond, the davemaoite
stayed intact for only
about a second”

Geology

AA

RO

N^ C

ELE

ST
IAN

/NA

TU

RA

L^ H

IST

OR

Y^ M

US

EU
M^ O

F^ L
OS

AN

GE
LE
S^ C

OU

NT

Y

New mineral from the deep


A diamond from Earth’s lower mantle contains crystals of a mineral
we have never seen before, now named davemaoite, finds Alice Klein

The tiny diamond that
contained the new
mineral davemaoite

Machine learning

Carissa Wong

AN ARTIFICIAL intelligence tool can
quickly suggest possible candidates
for the chemical structures of
psychoactive “designer drugs”
from a simple analysis. The tool
could fast-track the development
of lab tests that screen for drugs

with similar effects to cocaine
and heroin, but which aren’t
detectable by current tests.
“Our method could cut down
the time required to identify a
new designer drug from weeks
or months to just hours,” says
Michael Skinnider at the University
of British Columbia in Canada.
Skinnider and his colleagues
created a machine learning tool
called DarkNPS by training it with
the chemical structures of about
1700 known designer drugs. The
training set included tandem mass
spectrometry results for each drug,
which provided information on
the mass of the molecule and the
elements it contained. This allowed
the AI to identify patterns between
tandem mass spectrometry data
and chemical structures.
Given tandem mass spectrometry
data for a previously unseen drug,
DarkNPS could then guess the
molecular structure with an
accuracy of 51 per cent. This
increased to 86 per cent if the AI
gave its top 10 guesses, meaning it
could be most useful for narrowing
the search (Nature Machine
Intelligence, doi.org/g52q).
“This could... make it possible
to identify new designer drugs
much sooner after they’ve hit
the market,” says Skinnider. ❚

AI tool could be used
to make tests to
spot new drugs fast

2mm

SC


IEN


CE
PIC


TU


RE
CO


/SP


L


The structure
of a chemical
dictates what
effects it might
have in the body
Free download pdf