Water,
Journal Year:
2024,
Volume and Issue:
16(18), P. 2626 - 2626
Published: Sept. 16, 2024
Turning
to
green
technologies
in
wastewater
treatment
is
a
well-known
global
trend.
The
use
of
natural
sorbents
plant
origin
or
phytosorbents
order
purify
water
from
various
types
pollutants
becoming
more
and
popular.
This
solves
several
important
problems
at
once:
the
harmless
materials,
reducing
cost
processing,
waste
disposal.
Moreover,
there
increase
agricultural,
food,
woodworking,
other
industries.
review
presents
data
on
modern
mainly
vegetable
waste,
as
technologies.
Natural
materials
remove
ion
metals,
dyes,
crude
oil
petroleum
products,
organic
non-organic
contaminants.
techniques
obtaining
raw
are
considered.
methods
for
activation
modification
phytosorbents,
which
provide
greater
sorption
efficiency,
presented.
adsorption
mechanisms
contaminants
examined,
model
descriptions
shown.
It
has
been
revealed
that
effectiveness
interaction
depends
presence
functional
groups.
Studies
over
past
twenty
years
have
shown
good
prospects
such
practice.
Water,
Journal Year:
2024,
Volume and Issue:
16(23), P. 3380 - 3380
Published: Nov. 24, 2024
This
study
presents
an
innovative
approach
utilizing
artificial
intelligence
(AI)
for
the
prediction
and
classification
of
water
quality
parameters
based
on
physico-chemical
measurements.
The
primary
objective
was
to
enhance
accuracy,
speed,
accessibility
monitoring.
Data
collected
from
various
samples
in
Algeria
were
analyzed
determine
key
such
as
conductivity,
turbidity,
pH,
total
dissolved
solids
(TDS).
These
measurements
integrated
into
deep
neural
networks
(DNNs)
predict
indices
sodium
adsorption
ratio
(SAR),
magnesium
hazard
(MH),
percentage
(SP),
Kelley’s
(KR),
potential
salinity
(PS),
exchangeable
(ESP),
well
Water
Quality
Index
(WQI)
Irrigation
(IWQI).
DNNs
model,
optimized
through
selection
activation
functions
hidden
layers,
demonstrated
high
precision,
with
a
correlation
coefficient
(R)
0.9994
low
root
mean
square
error
(RMSE)
0.0020.
AI-driven
methodology
significantly
reduces
reliance
traditional
laboratory
analyses,
offering
real-time
assessments
that
are
adaptable
local
conditions
environmentally
sustainable.
provides
practical
solution
resource
managers,
particularly
resource-limited
regions,
efficiently
monitor
make
informed
decisions
public
health
agricultural
applications.
International Journal of Chemical Reactor Engineering,
Journal Year:
2025,
Volume and Issue:
unknown
Published: April 26, 2025
Abstract
The
consumption
of
vegetables
generates
a
lot
waste,
allowing
their
use
as
biomass.
biosorbent
based
on
turnip
leaves
were
prepared;
they
processed
in
the
raw
state
(TL)
and
activated
with
H
3
PO
4
(TLA),
then
tested
to
remove
Crystal
Violet
dye
CV
from
aqueous
solutions.
Adsorbents
characterized
using
(FTIR),
(SEM),
(XRD),
(TGA)
pH
PZC
,
revealing
that
functional
groups
OH,
C-H,
C=O
C-O-C
mainly
responsible
for
adsorption
CV.
Scanning
electron
microscope
(SEM)
imaging
revealed
cellulose
fibers
multicellular
structure
initially
linked
lignin
hemicellulose,
which
dissociated
after
chemical
treatment,
XRD
analysis
confirmed
amorphous
nature
structure,
attributed
presence
hemicellulose.
kinetic
study
showed
best
suited
models
describe
experimental
data
pseudo-second-order
(PSO)
model
TLA
TL.
isotherms
different
followed
Sips
isotherm
maximum
capacity
(qmax)
635.54
mg/g
TL
621.76
TLA.
optimal
conditions
found
be
an
adsorbent
mass
20
mg,
temperature
25
°C.
concentration
10
mg/L
respective
contact
times
150
min
120
TLA,
leading
elimination
yields
92.60
%
97.56
%.
mechanism
could
explained
by
electrostatic
interactions
between
negatively
charged
surface
positively
group
dye.
A
thermodynamic
was
carried
out
process
solutions
this
exothermic
spontaneous
due
approximate
values
ΔH
(−25.26
−20.69
TLA)
ΔG.
predictive
multi-component
system
studied
Support
Vector
Machine
(SVM)
model.
Two
SVM
approaches
developed
compared.
first
involved
method
integrated
optimization
algorithm,
while
second
used
more
recent
efficient
method,
Dragonfly
(DA)
conjunction
method.
evaluation
accuracy
three
commonly
statistical
measures:
mean
squared
error
(RMSE),
coefficient
determination
(R2)
correlation
(R).
included
five
important
variables,
136
observations:
weight
(TL
or
TLA),
initial
(c
0
mg/L),
dose
foliar
waste
(in
time
minutes)
final
(Cf).
programming
MATLAB
software.
results
demonstrated
DA-SVM
optimized
RBF-Gaussian
kernel
function
had
excellent
prediction
ability,
R2
0.997,
R
0.998
RMSE
1.0809.
