International
Journal
of
Pharmaceutics
425 (2012) 73–
74
Contents
lists
available
at
SciVerse
ScienceDirect
International
Journal
of
Pharmaceutics
jo
ur
nal
homep
a
ge:
www.elsevier.com/locate/ijpharm
Letter
to
the
Editor
Filling
hard
gelatin
capsules
by
the
dosator
nozzle
system
–
Is
it
possible
to
predict
where
the
powder
goes?”
The
transfer
of
powder
into
a
capsule
to
provide
a
unit
dose
ini-
tially
used
the
capsule
shell
as
the
measuring
volume.
The
need
to
provide
unit
doses
on
an
industrial
scale
necessitated
the
replace-
ment
of
such
systems
with
various
procedures
where
a
sample
is
removed
from
a
bed
of
powder
and
transferred
into
the
capsule
body
to
provide
the
unit
dose.
As
with
industrial
tablet
production,
the
systems
for
selecting
a
sample
of
powder
can
be
manufactured
and
adjusted
with
greater
flexibility
than
a
capsule
shell.
The
vari-
ous
methods
available
to
provide
such
a
process
are
well
described
by
Podczeck
(2004).
The
ability
to
relate
the
properties
of
capsule
formulations
to
their
ability
to
be
filled
into
capsules
by
the
various
manufacturing
processes,
has
been
the
subject
of
several
publi-
cations
and
the
recent
publications
of
Khawam
(2011),
Khawam
and
Schultz
(2011),
which
claim
to
provide
a
theoretical
insight
and
experimental
support
for
their
theoretical
approach,
are
recent
additions
to
these.
There
are
certain
problems
with
both
theory
and
practice
presented
in
these
papers,
which
give
me
cause
for
con-
cern,
and
before
formulators
rush
to
use
the
approach
as
an
answer
to
their
problems,
I
should
like
to
outline
my
concerns.
First,
I
should
like
to
consider
the
theoretical
aspects
of
the
work.
The
author
claims
that
there
are
three
settings
that
control
the
encapsulation
process
and
illustrates
these
with
the
diagrams
in
Fig.
3.
I
cannot
agree
that
these
three
dimensions
are
the
only
factors
which
govern
the
quantity
of
powder
that
can
be
removed
from
the
powder
bed
in
an
industrial
capsule
filling
machine
of
the
dosator
nozzle
type.
I
agree
that
the
height
of
the
powder
bed
is
vital
but
how
can
this
be
controlled
and
just
what
size
of
sample
is
removed
is
far
from
as
simple
as
the
author
suggests,
as
I
and
my
co-
authors
have
shown
in
a
series
of
publications
on
the
topic
(Jolliffe
et
al.,
1982;
Jolliffe
and
Newton,
1982,
1983;
Tan
and
Newton,
1990a,b,c).
While
the
author
quotes
the
concept
of
the
formation
of
a
stable
arch
which
we
proposed
(Jolliffe
et
al.,
1980)
to
explain
how
a
sample
can
be
retained
within
a
nozzle,
even
we
realised
that
it
was
not
the
only
factor
involved
in
sampling
and
transfer.
The
author
appears
to
be
unaware
of
the
reality
of
the
machine,
and
the
process.
Their
Fig.
3a
represents
an
ideal
situation.
The
nozzle
is
placed
into
the
powder
bed
and
a
quantity
of
powder
is
confined
within
the
dosator.
Under
ideal
conditions,
it
should
be
possible
to
remove
this
volume
of
powder
and
if
the
bed
is
uniform
in
density,
then
a
constant
weight
can
be
removed.
We
did
in
fact
use
such
an
approach
to
measure
the
bulk
density
of
the
powder
beds
in
our
capsule
filling
simulator
(Jolliffe
et
al.,
1982).
The
author
suggests
that
it
is
always
necessary
to
apply
a
compression
force
to
the
top
of
the
powder
bed
to
be
able
to
remove
a
sample.
It
is
clear
from
our
work
(Tan
and
Newton,
1990c)
that
this
is
not
the
case.
A
powder
that
could
be
sampled
without
applying
compression
was
Microcrystalline
Cellulose,
one
of
the
powders
used
by
Khawam
and
Schultz
(2011).
What
the
quantity
within
the
nozzle
does
represent
is
the
maximum
quantity
that
could
be
sampled.
If
the
quantity
of
powder
transferred
is
less
than
this,
then
there
must
be
a
loss
of
powder
somewhere
in
the
process.
The
possibility
that
the
powder
may
not
have
the
ability
to
from
an
arch
across
the
bottom
of
the
nozzle
is
one
possibility.
In
this
case,
powder
would
be
lost
as
the
nozzle
is
removed
from
the
bed.
Dosator
nozzle
capsule
filling
machines
were
designed
to
apply
a
compression
force
to
the
powder
within
the
nozzle
and
our
original
theoretical
paper
(Jolliffe
et
al.,
1980)
demonstrated
the
consequences
of
the
application
of
a
force
at
the
top
of
the
powder
bed
and
the
factors
involved
with
providing
a
stable
arch
such
that
the
powder
would
not
fall
from
the
nozzle
when
removed
from
the
powder
bed.
Our
subsequent
papers
however
illustrated
that
it
was
not
quite
as
simple
as
this
and
the
quantity
of
powder
removed
was
often
less
than
that
which
was
defined
by
the
volume
provided
by
the
dosator
dimensions.
Increasing
the
amount
of
compression
often
resulted
in
a
decrease
in
the
amount
of
powder
transferred
(Jolliffe
and
Newton,
1982,
1983;
Tan
and
Newton,
1990a,b,c).
At
the
time
we
proposed
that
this
was
due
to
powder
adhering
to
the
wall
of
the
nozzle,
providing
a
reduced
diameter
and/or
loss
of
powder
behind
the
piston
tip.
In
hindsight
and
looking
at
the
theoretical
approach
of
Khawam
(2011),
there
may
also
be
a
further
reason
for
the
failure
of
the
system
to
sample
the
quantity
of
powder
available.
Khawam
(2011)
illustrates
the
two
ways
in
which
compression
can
be
applied
(Fig.
3b
and
c)
and
proposes
equations
to
be
able
to
quantify
the
amount
of
the
powder
which
can
be
sampled,
by
relating
to
the
heights
of
the
powder
bed
and
the
piston.
They
also
suggest
how
the
height
of
the
piston
can
be
calculated
for
various
stages
of
the
process.
The
equations
they
propose
for
the
calcula-
tions
cannot
be
correct
for
they
do
not
take
into
account
the
fact
that
the
powder
cannot
be
reduced
in
volume
to
the
level
they
illustrate
because
to
do
so,
the
powder
bed
would
have
to
have
a
greater
density
than
the
powder
itself,
which
is
clearly
impossible.
In
our
work
we
never
used
a
compression
ratio
greater
than
0.5.
This
would
imply
a
reduction
in
volume
in
the
nozzle
of
50%
of
the
orig-
inal,
which
translates
to
a
doubling
of
the
density
of
the
sample.
For
the
system
we
studied,
this
would
not
exceed
the
apparent
particle
density
of
the
powder
within
the
nozzle.
Even
here,
we
often
found
that
increasing
the
compression
ratio
lead
to
a
reduction
in
cap-
sule
fill
weight.
The
equations
proposed
by
Khawam
(2011)
need
to
provide
a
limiting
value
beyond
which
further
densification
can-
not
occur.
In
terms
of
the
dosage
form
itself,
the
whole
objective
of
making
capsules
is
not
to
provide
a
unit
dose
that
is
a
tablet,
but
a
powder
structure
that
will
readily
disintegrate
when
swallowed.
What
must
also
be
considered
is
that
the
dosator
nozzle
system
is
not
a
‘stiff’
machine,
i.e.,
the
piston
movement
is
controlled
only
by
the
movement
of
the
machine.
The
nozzle
must
be
able
to
move
into
the
bed
and
yet
it
must
not
touch
the
metal
tray
holding
the
powder
bed.
This
provides
a
system,
which
cannot
prevent
that
a
loss
of
powder
between
the
bottom
of
the
feed
hopper
and
the
noz-
zle
will
occur.
In
addition,
the
system
is
fitted
with
a
spring
which
0378-5173/$
–
see
front
matter ©
2012 Elsevier B.V. All rights reserved.
doi:10.1016/j.ijpharm.2012.01.008