Gradient thin layer chromatography of coumarins and furocoumarins


Glowniak, K.; Matysik, G.; Bieganowska, M.; Soczewinski, E.

Chromatographia 22(7-12): 307-310

1986


Extracts of four umbelliferous plants containing coumarins and furocoumarins were chromatographed in sandwich chambers with a glass distributor. Best separations were obtained for stepwise gradient elution in the concentration range of 2.5 to 15% ethyl acetate in chloroform. Seven eluent fractions with increasing ethyl acetate concentrations were introduced directly under the distributor of the sandwich chamber.

Gradient
Thin
-Layer
Chromatography
of
Coumarins
and
Furocoumarins
K.
Glowniak*
/
G.
Matysik
/
M.
Bieganowska
/
E.
Soczewitiski
Department
of
Inorganic
and
Analytical
Chemistry,
Department
of
Pharmacognosy
*
,
Medical
Academy,
Staszica
6,
20-081
Lublin,
Poland
Key
Words
Thin
-layer
chromatography
Stepwise
gradient
elution
Coumarins
Furocoumarins
Summary
Extracts
of
four
umbelliferous
plants
containing
cou-
marins
and
furocoumarins
were
chromatographed
in
sandwich
chambers
with
a
glass
distributor.
Best
separa-
tions
were
obtained
for
stepwise
gradient
elution
in
the
concentration
range
of
2.5
to
15%
ethyl
acetate
in
chloroform.
Seven
eluent
fractions
with
increasing
ethyl
acetate
concentrations
were
introduced
directly
under
the
distributor
of
the
sandwich
chamber.
Introduction
The
separation
of
simple
coumarins
and
especially
furo-
coumarins
occurring
in
plant
materials
is
difficult.
Systema-
tic
investigations
on
a
large
number
of
plant
species
were
carried
out
using
conventional
thin
-layer
chromatography
[1,
2].
The
method
is
useful
for
chemotaxonomic
studies
[3,
4]
for
analysis
of
coumarins
and
their
metabolites
[5-8]
for
optimization
of
HPLC
systems
[9,
10]
and
for
prepara-
tive
separation
both
on
a
thin
-layer
and
in
a
column
[7,
11,
12].
The
application
of
the
sandwich
chamber
with
a
glass
distributor
[13]
permits
the
use
of
gradient
elution
TLC
which
greatly
increases
the
separation
efficiency
of
com-
plex
samples
containing
components
with
wide
range
of
polarities
[13-16].
The
increased
efficiency
is
due
to
dis-
placement
effects
in
the
initial
band.
In
turn,
this
is
due
to
the
flow
of
a
mobile
phase
with
increasing
polarity
causing
the
compression
of
the
individual
solute
bands
so
that
even
wide
starting
bands
(large
samples)
are
resolved
into
narrow
zones;
even
components
present
in
small
amounts
can
be
detected.
Free
coumarins
and
furocoumarins
are
usually
extracted
from
the
plant
material
with
petroleum
ether;
therefore,
ether
extracts
of
fruits
from
the
family
of
Umbelliferae
described
in
the
literature
as
the
source
of
coumarins
were
analysed,
such
as
hogweed
(Heracleum
sphondylium
L)
[17]
and
common
parsnip
(Pastinaca
sativa
L)
[18].
Besides
of
these
known
sources
of
coumarins,
the
extracts
of
fruits
of
skirret
(Slum
sisarum
L)
and
Libanotis
intermedia
Rupr.,
whose
compositions
are
much
less
known,
were
investigated
[19].
In
order
to
select
the
optimal
separation
conditions
of
the
plant
extracts,
the
effect
of
the
eluent
composition
on
the
retention
of
several
coumarins
(osthol,
umbelliferon)
and
furocoumarins
(isopimpinelline,
imperatorine,
bergapten,
xanthotoxine)
was
investigated
[9,
10].
These
compounds
occurring
both
in
the
fruits
of
Heracleum
sphondylium
L
as
well
as
in
parsnip
are
difficult
to
separate
(especially
furocoumarins)
due
to
their
similar
chemical
structures.
The
conditions
for
gradient
elution
chromatography
of
the
plant
extracts
were
selected
on
the
basis
of
data
ob-
tained
for
isocratic
elution.
Experimental
Plant
extracts
were
obtained
by
continuous
extraction
of
the
powdered
material
with
petroleum
ether
in
a
cir-
culation
extractor.
The
volume
of
the
extract
(in
cm
3
)
corresponded
to
the
mass
of
the
material
(in
grams).
The
test
solutes
were
obtained
from
the
Danish
Royal
School
of
Pharmacy,
Copenhagen.
The
elution
was
carried
out
in
sandwich
chambers
with
a
glass
distributor
[13]
using
10
X
10cm
precoated
silica
plates
for
high-performance
TLC
with
or
without
a
pre
-
concentration
zone.
1.0-1.5p1
of
0.1
%
w/v
solutions
of
the
test
solutes
in
methanol
were
spotted.
Two
-com-
ponent
eluents
contained
a
polar
modifier
(diisopropyl
ether,
ethyl
acetate,
methyl
ethyl
ketone)
and
a
nonpolar
diluent.
Stepwise
gradient
elution
was
carried
out
as
de-
scribed
earlier
[13-16,
20-21]
by
introducing
consecutive-
ly
0.2
ml
fractions
of
the
eluent
under
the
distributor,
according
to
a
predetermined
program.
In
order
to
eliminate
errors
in
the
comparison
of
the
R
F
values
of
test
solutes
with
those
of
the
separated
zones
of
the
extract
(caused
by
displacement
effects),
the
test
solutes
and
the
extracts
were
alternatively
spotted
in
separate
experiments
on
the
preconcentration
zone
of
the
plate
in
short
distances
so
that
the
starting
spots
partly
overlapped.
During
the
elution
process
the
zones
were
flat-
tened
and
their
ends
overlapped
forming
a
slightly
thicker
zone.
Chromatographia
Vol.
22,
No.
7-12,
December
1986
Originals
0009-5893/86/7-12
0307-04
$
03.00/0
CD
1986
Friedr.
Vieweg
&
Sohn
Verlagsgesel
lschaft
mbH
307
Results
and
Discussion
The
results
of
the
first
series
of
experiments
(isocratic
development)
are
shown
in
Figs.
1-5
as
R
F
values
plotted
against
the
per
cent
concentration
of
the
polar
modifier
(dichloromethane,
diisopropyl
ether,
methyl
ethyl
ketone
or
ethyl
acetate)
in
the
eluent.
For
most
cases
the
R
F
vs.
%
plots
were
approximately
linear
in
the
concentration
ranges
investigated.
The
weakly
polar
eluent
toluene
+
dichloromethane
sep-
arated
well
the
coumarins
osthol
and
umbelliferone,
with
R
F
0.6
0.4
3
5
2
0.2
30
SO
70
90
%
CH
2
CL.
z
Fig.
1
R
F
values
of
coumarins
plotted
against
the
per
cent
concentration
of
dichloromethane
(CH
2
C12);
diluent:
toluene.
1
osthol;
2
isopimpinellin;
3
imperatorin;
4
bergapten;
5
xantho-
toxin;
6
umbelliferone.
R
F
0.8
0.6
3
0.4
6
A
5
0.2
40
60
80
100
Pr
2
0
Fig.
2
R
F
values
of
coumarins
plotted
against
the
per
cent
concentration
of
diisopropyl
ether
(iPr
2
0);
diluent:
n-heptane.
For
the
coumarins
see
Fig.
1.
differentiated
molecular
structure:
the
prenyl
radical
in
the
osthol
structure
decreased
the
adsorption
affinity
while
the
hydroxyl
group
in
umbelliferone
strongly
inter•
acts
with
surface
silanol
groups
causing
strong
adsorption
RF
0.8
0.6
0.4
02
4
3
2
5
20
30
40
5
0
#4,
Me
Et
CO
Fig.
3
R
F
values
of
coumarins
plotted
against
the
per
cent
concentration
of
methyl
ethyl
ketone
(MeEtCO);
diluent:
n-heptane.
For
the
coumarins
see
Fig.
1.
0.8
R
F
0.6
0.4
0.2
3
4
5
10
30
S
D
7
0
V.
Et
OAc
Fig.
4
R
F
values
of
coumarins
plotted
against
the
per
cent
concentration
of
ethyl
acetate
(EtOAc);
diluent:
n-heptane.
For
the
coumarins
see
Fig.
1.
308
Chromatographia
Vol.
22,
No.
7-12,
December
1986
Originals
0.8
0.6
01tt
0.2
RF
3
5
5
10
V.Et0Ac
Fig.
5
R
F
values
of
coumarins
plotted
against
the
per
cent
concentration
of
ethyl
acetate
(EtOAcI;
diluent:
chloroform.
For
the
coumarins
see
Fig.
1.
and
low
R
F
values.
The
furocoumarins
are
poorly
separa-
ted;
the
AR
M
(OCH
3
)
value
for
the
extreme
pair
bergaptene-
isopimpinelline
at
30%
CH
2
Cl
2
is
only
0.3
R
M
units.
The
use
of
diisopropyl
ether
as
the
modifier
(Fig.
2)
re-
sulted
in
higher
R
E
values,
however,
the
selectivity
is
worse
than
in
the
former
system:
the
R
F
vs.
%
diisopropyl
ether
lines
cross
indicating
changes
in
the
retention
se-
quence
with
the
composition
of
the
eluent,
which
is
a
drawback
in
gradient
elution.
More
interesting
results
were
obtained
for
methyl
ethyl
ketone
(Fig.
3)
and
ethyl
acetate
(Fig.
4)
used
as
modifiers.
The
R
F
vs.
%
lines
are
approximately
parallel
and
have
higher
slopes
in
comparison
to
Figs.
1
and
2.
The
selec-
tivity
is
satisfactory
although
a
slightly
lower
value
of
AR
M
(OCH
3
)
is
obtained
(0.28
R
M
units).
There
are
some
differences
in
the
sequence
of
the
R
E
values
for
the
two
systems.
The
fifth
eluent
was
composed
of
chloroform
and
ethyl
acetate
(Fig.
5).
A
higher
range
of
R
F
values
was
ob-
tained
and
the
slopes
of
the
R
F
vs.
%
plots
were
lower
since
both
components
of
the
eluent
are
polar.
Satisfac-
tory
selectivity
of
separation
was
obtained.
On
the
whole,
the
range
of
R
F
values
of
the
test
solutes
important
components
of
the
coumarin
and
furocoumarin
fractions
was
rather
narrow
and
did
not
exceed
0.4
R
F
units
so
that
the
narrow
polarity
range
is
suitable
for
isocratic
elution.
However,
the
selectivity
of
separation
of
the
components
is
frequently
poor.
Furthermore,
other
components
of
lower
or
higher
polarity
are
present
in
plant
extracts.
Therefore,
gradient
elution
was
applied
in
the
second
series
of
experiments.
As
demonstrated
in
an
earlier
paper
[20],
the
program
of
gradient
elution
can
be
selected
from
the
isocratic
data.
It
should
start
from
that
composition
of
the
eluent
at
which
the
less
polar
fraction
is
well
separated
and
should
terminate
at
an
eluent
com-
position
which
separates
the
strongly
retained
components.
Therefore,
two
gradient
programs
were
selected
from
the
first
series
of
(isocratic)
experiments:
heptane
+
methyl
ethyl
ketone
(Fig.
3)
and
chloroform
+
ethyl
acetate
(Fig.
5).
60
50
40
-
30
20
10
%
Me
Et
CO
eluent
fraction
N.
2
3
4
5
6
Fig.
6
Stepwise
gradient
program
10-50%
methyl
ethyl
ketone;
diluent:
n-heptane.
V.
Me
EtC0
7
p
.
60
SO
40
30
eluent
fraction
Na
1
2
3
4
5
6
Fig.
7
Stepwise
gradient
program
30-70%
methyl
ethyl
ketone;
diluent:
n-heptane.
15.0
12.5
10,0
7.5
5.0
2$
V.
Et
OAc
eluent
fraction
N.
3
4
5
6
7
Fig.
8
Stepwise
eluent
program
2.5-15.0%
ethyl
acetate;
diluent:
chloro-
form.
Chromatographia
Vol.
22,
No.
7-12,
December
1986
Originals
309
cc
,
•=*-
cp
o
c:)
0
cz)
cz)
(=>
g;
0
PH
S
L
M
1
2
3
4
5
6
•.
*.
0cz)
-C3
0
ci-•
-
c3-
-
00
-
-
e
-
o
a3
0.
11
0
0
O
‘=
2,
0
P
I
-
1
5
LM
1
2
3
4
56
0
o
Q
.Jczycn
g
cDoc=,
C;)
61
.
•ez=5
cD
crp
o
c0
m
o
PH
SL
I1
1
2
3
4
5
6
-
-------
----0
Fig.
9
Chromatograms
of
plant
extracts
and
reference
coumarins
corresponding
to
the
programs
given
in
Figures
6-8.
Fig.
9a
vvas
obtained
by
using
the
program
of
Fig.
6,
Fig.
9b
by
using
the
program
given
in
Fig.
7,
and
Fig.
9c
by
using
the
program
given
in
Fig.
8.
Plant
extracts:
P
Pastinaca
sativa;
H
Heracleum
sphondylium;
S
Sium
sisarum;
L
Libanotis
intermedia.
Reference
coumarins:
1
osthol;
2
isopimpinellin;
3
imperatorin;
4
bergapten;
5
xanthotoxin;
6
umbelliferone.
Figs.
6
and
7
illustrate
the
stepwise
programs
for
the
mix-
tures
of
heptane
and
methyl
ethyl
ketone,
and
Figures
9/a
—b
show
the
corresponding
chromatograms.
The
separation
is
better
than
in
the
case
of
isocratic
elution,
especially
for
the
second
gradient
program
(six
stages,
four
steps,
Fig.
7)
with
a
higher
average
eluent
strength;
however,
the
furo-
coumarins
formed
two
pairs,
bergapten
+
imperatorin,
and
isopimpinelline
+
xanthotoxine.
Best
results
were
obtained
when
using
the
eluent
composed
of
chloroform
and
ethyl
acetate
(Fig.
8,
and
Fig.
9/c).
All
test
solutes
are
separated
into
narrow
zones
(the
R
F
values
are:
osthol,
0.76;
bergaptene,
0.62;
imperatorin,
0.57;
xan-
thotoxine,
0.48;
isopimpinelline,
0.44;
and
umbelliferon,
0.17).
The
chromatogram
in
Fig.
9/c
gives
detailed
information
on
the
presence
of
the
coumarins
investigated
in
the
plant
extracts.
Thus,
all
six
test
solutes
are
present
in
Pastinaca
Sativa
and
Heracleum
sphondylium
and
in
the
less
known
Libano-
tis
intermedia
Rupr.
;
higher
concentrations
of
umbelliferone
are
present
in
Heracleum
sphondylium
and
higher
concen-
trations
of
xanthotoxine
are
present
in
Libanotis
intermedia
and
in
Heracleum
sphondylium.
The
extract
of
Slum
sisarum
contained
osthol,
bergapten
and
imperatorin
and
probably
small
amounts
of
xanthotoxine.
Conclusions
Stepwise
gradient
TLC
greatly
increases
the
efficiency
of
separation
which
is
especially
evident
in
the
case
of
com-
plex
samples
such
as
plant
extracts.
Due
to
displacement
effects
under
conditions
of
gradually
increasing
eluent
strength
of
the
mobile
phase,
the
separa-
tion
of
solutes
of
similar
R
F
values
is
also
improved
which
facilitates
the
comparison
of
compositions
of
various
plant
extracts.
Relatively
larger
sample
sizes
can
be
applied
in
gradient
TLC
which
is
favourable
for
the
detection
of
components
present
in
smaller
amounts.
References
[1]
K.
Wierzchowska-Renke,
Acta
Polon.
Pharm.,
31,
225
(1974).
[2]
B.
Kaminski,
K.
Glowniak,
A.
Majewska,
J.
Petkowicz,
D.
Szaniawska-Dekundy,
Farm.
Pol.,
34,
25
(1978).
[3]
H.
E.
Nordby,
S.
Nagy,
J.
Chromatogr.,
207,
21
(1981).
[4]
R.
V.
Tamma,
G.
C.
Miller,
R.
Everett,
J.
Chromatogr.,
322,
236
(1985).
[5]
M.
Kowalska,
J.
Skrzypczakowa,
Dissert.
Pharm.
(Cracow),
16,
225
(1964).
[6]
J.
Karlsen,
L.
E.
J.
Boomsma,
A.
B.
Svendsen,
J.
Chroma-
togr.,
42,
550
(19691.
[7]
J.
Mendez,
J.
Rubido,
Planta
Med.,
36,
219
(1979).
[8]
E.
H.
Oertli,
R.
C.
Beier,
G.
W.
!vie,
L.
D.
Rowe,
Photo-
chemistry
23,
439
(1984).
[9]
K.
Glowniak,
M.
Bieganowska,
J.
Liq.
Chromatogr.,
8,
2927
(1985).
[10]
K.
Glowniak,
M.
Bieganowska,
in
press.
[11]
T.
M.
Swager,
J.
H.
Cardellina
II.,
Phytochemistry
24,
805
(1985).
[12]
F.
R.
Stermitz,
R.
D.
Thomas,
J.
Chromatogr.,
77,
431
(1973).
[13]
E.
Soczewinski,
in
Planar
Chromatography
Vol.
I
(R.
E.
Kaiser,
Editor)
Hilthig
Verlag,
Heidelberg
1986;
p.
79.
[14]
E.
Soczewinski,
G.
Matysik,
J.
Liq.
Chromatogr.,
8,
1225
(19851.
[15]
E.
Soczewinski,
G.
Matysik,
K.
Glowniak,
in
Instrumental
High
Performance
Thin
-Layer
Chromatography
IR.
E.
Kaiser,
Ed.)
Proc.
Third
Intern.
Symp.
Wiirzburg
IV
1985.
Institute
for
Chromatography,
Bad
Diirkheim
1985;
p.
413.
[16]
W.
Cisowski,
K.
Glowniak,
G.
Matysik.
E.
Soczewinski,
Herba
Polon,
in
press.
[17]
T.
Beyrich,
Pharmazie
23,
336
(1968).
[18]
N.
P.
Maksyutina,
Khim.
Prir.
Soedin.,
3,
213
(1967).
[19]
D.
Motl,
B.
Droidi,
Collect.
Czech.
Chem.
Commun.,
42,
2815
(1977).
[20]
E.
Soczewinski,
J.
Chromatogr.,
in
press.
[21]
G.
Matysik,
E.
Soczewinski,
J.
Chromatogr.,
in
press.
Received:
July
23,
1986
Accepted:
Aug.18,
1986
A
310
Chromatographia
Vol.
22,
No.
7-12,
December
1986
Originals