The Economist February 26th 2022 Science & technology 77
An American firm called EnergyX,
plans, however, to change that. Using a po
lymer membrane developed by Benny
Freeman of the University of Texas at Aus
tin the company intends, beginning later
this year, to filter lithium directly from
brine. Dr Freeman says the company’s pilot
plant, which will be able to fit into a stan
dard shipping container, should be able to
handle millions of litres of brine a day.
Once the process is perfected, he reckons it
will be able to extract at least 90% of the
lithium within a brine.
Speaking at the aaasmeeting, Dr Free
man explained that his inspiration was the
protein channels which control the flow of
metal ions in and out of biological cells. A
potassium channel, for example, is 10,000
times more permeable to potassium ions
than to those of sodium. Modern imaging
tools and supercomputers have revealed
the structure of these channels, permitting
them to be mimicked.
The upshot is a membrane pierced by
nanometresized pores made from rings of
carbon and oxygen atoms precisely ar
ranged to let lithium ions through while
slowing the passage of others, such as so
dium. To start with, these membranes will
enrich a brine’s lithium levels before it en
ters the ponds. Eventually, they could re
place the ponds by generating a pure and
concentrated solution of lithium hydrox
ide suitable for immediate industrial use.
Another way to improve the efficiency
of ponds was also proposed at the meeting,
by Seth Darling of Argonne National Lab
oratory, in Illinois. The actual evaporating
is done mainly by heat delivered as sun
light. But much of this is wasted. Either it
warms water below a pond’s surface—
which, not being in contact with the air, is
thus unavailable for evaporation—or it is
reradiated before it has had a chance to
liberate any water molecules.
Dr Darling worked out that covering a
pond with a material that converts light to
heat quickly would stop these unfortunate
losses by concentrating the warming effect
at the pond’s surface, thus promoting
evaporation. As long as that material was
also porous, it would then let the resulting
water vapour through and out into the air,
to be blown away. His first try was Chinese
calligraphy ink which, being viscous and a
lightabsorbing black, worked well—ex
cept that eventually (as ink does) it dis
solved in the water. Now, though, he pref
ers charcoal, which also has the green bo
nus of being made from farm waste.
Dr Darling says the result is a process
that can convert incident sunlight into
heat at the water’s surface with near 100%
efficiency. Besides improving lithium ex
traction, that could also help industries,
such as fracking and mining, that tend to
accumulate large ponds of waste water as a
byproduct of their activities. It could be
usedaswellasa lowcostwaytoturnsea
waterintofresh,bycapturingthewaterva
pourasit evaporated.
Moreover,ifDrFreeman’sandDrDar
ling’sideascangetridoftheneedtouse
vastevaporationponds,thatwouldopen
upnewsourcesoflithium.TheSaltonSea,
a lake in southern California, contains
hugequantitiesofthestuff,andinTexasa
lotofgroundwaterreleasedasa sideeffect
ofoilproductionisrichinit.Unfortunate
ly,neitheroftheseplaceshasthevastacre
agesofotherwisevaluelesslandrequired
for the sorts of inefficient evaporation
pondsusedinSouthAmerica.Shrinking
thesizesofthoseponds,orevengettingrid
ofthemaltogether,wouldchangethat.n
Tomorrow’sbatteriestoday
Thepsychologyofjustice
First impressions
T
he “satanic panic” that swept
through America in the 1980s and 1990s
held that thousands of ordinary people up
and down the country were secretly mem
bers of devilworshipping cults which
were abusing, raping and murdering chil
dren on an industrial scale. Alleged victims
made detailed allegations, often after ther
apy designed to “recover” memories that
had supposedly been buried in the after
math of trauma. Many people went to pri
son. None of it was true.
One aftereffect of the panic was to ce
ment in the minds of both the public and
the justice system the idea that eyewitness
testimony is unreliable. That fitted with
experiments by psychologists such as Eliz
abeth Loftus, which demonstrated just
how malleable memories can be. The Inno
cence Project, an American charity, exam
ined 375 cases of wrongful conviction for
all sorts of crimes, and found misidentifi
cation of suspects by witnesses was a fac
tor in around 70% of them.
But at the annual meeting of the Amer
ican Association for the Advancement of
Science, John Wixted, a psychologist at the
University of California, San Diego, argued
that this institutional distrust has gone too
far. Eyewitness memories, he said, can in
fact be very reliable—if they are tested in
the right circumstances.
The key to reliability, said Dr Wixted, is
the confidence of witnesses in their as
sessments. Experiments suggest that
when witnesses to a simulated crime are
confident of having identified the suspect
in a later photo lineup, they are almost al
ways correct. Similarly, if they are sure the
suspect is not present, that is likely to be
right too. Only when a witness is unsure
does a risk of misidentification arise. A
field study conducted in 2016 by Houston’s
police came to similar conclusions.
The problem is that this confidence is
trustworthy only the first time the ques
tion is asked. One of the unavoidable frus
trations of quantum mechanics is that
measuring a particle’s position or energy
irretrievably alters it. Something similar,
said Dr Wixted, happens with memories.
The very act of testing them contaminates
every other test that comes after. Assessing
people’s faces for a possible match, for ex
ample, lodges them in a witness’s memory.
Once that has happened, anything from
police encouragement to the highpres
sure environment of a courtroom can twist
subsequent attempts at recollection.
Dr Wixted drew a comparison with evi
dence such as dnasamples. Improper han
dling can contaminate these. That does not
mean dnatests are inherently unreliable,
but it does mean the technology must be
used carefully. The same, he says, is true of
witnesses. The answer, as he and Dr Loftus
argue in a recently published paper, is to
test a witness’s memory as fairly as possi
ble, and—crucially—to do so only once.
Decades after the Satanic panic, the
matter remains important. Dr Wixted cited
the case of Charles Don Flores, a prisoner
awaiting execution for a murder commit
ted in 1998. Initially, when shown a lineup
that included Mr Flores, a crucial witness
said none of the people matched her recol
lection. (She had recalled a white man with
long hair. Mr Flores is of Latin American
extraction, and had short hair then.) By the
time the case came to trial a year later, she
had changed her mind, and Mr Flores was
convicted. His appeal on the basis of the
witness’s change of mind has been denied.
Dr Wixted, however,suggests she was like
ly to have beenright the first time and
wrong the second.n
Eyewitness evidence can be more
reliable than thought