Note: Descriptions are shown in the official language in which they were submitted.
CA 02268849 1999-04-12
WO 98/18558 PCT/US97/19062
-1-
TITLE OF THE INVENTION
END-SAMPLING THIEF PROBE
BACKGROUND OF THE INVENTION
Efficient blending and sampling of powders is of critical
importance in the manufacture of a wide variety of pharmaceutical solid
doses such as tablets and capsules. These products are manufactured
from powder blends, granulated powders, and extruded pellets. In a
typical process, powders, granules, and pellets are blended, discharged
to a tote or drum, emptied into a hopper on a press or encapsulator, and
- divided into the final dosage form. Achieving and maintaining
homogeneous and well characterized blends of powders and granules is
of critical importance, especially in formulations involving small
amounts of high potency components, which are a substantial fraction of
all oral dosages. Inadequate mixing somewhere along the production
sequence often results in rejection of finished product due to poor
quality.
In many systems, the components requiring blending are
usually powders of different size, density, shape, and cohesiveness.
Since such materials often display a considerable tendency to segregate,
ultimate mixture homogeneity cannot be taken for granted; quite the
opposite, unless the blending process is properly designed and
controlled, the result is often a mixture with significant composition
fluctuations throughout the powder bed [See, L.T. Fan, Y.-M. Chen, and
F.S. Lai, Powder Technol., 61 (1990) 255. and M. Poux, P. Fayolle, J.
Bertrand, D. Bridoux, and J. Bousquet, Powder Technol., 68 (1991)
213.]. Inhomogeneities in the powder blend can result in increased
variability in the contents of potent components in tablets, leading not
only to decreased therapeutic value but also to direct health risks due to
toxicity in super-potent tablets.
For the reasons mentioned above, a thorough understanding
of blending processes is highly desirable. Unfortunately, blending of
granular materials is largely an art rather than a science, and at the
present time the ability to design and accurately evaluate the
CA 02268849 1999-04-12
WO 98/18558 PCT/US97/19062
__2_
performance of a mixing process for a high potency drug is limited.
Characterization of mixtures in most industrial processes
relies on taking and analyzing discrete samples. Parameters such as
sample size (n), number of samples (N), and location of the sampling
points can affect the measurement values. Guidelines for selecting the
number of samples have been proposed based on theoretical random
mixtures (i.e., >30) (Devore, J.L., Probability and Statistics for
Engineering and the Sciences, Vol., Brooks/Cole Publishing Company,
Monterey, 1982, p. 640] but optimal values of these parameters for real
systems displaying incomplete mixing are often unknown.
In real mixtures, practical considerations and physical
limitations of sampling mechanisms limit the number and size of the
samples that can be obtained. Extensive sampling is often impractical;
commonly, just a few samples (<30) are removed from a blender. The
most common approach is to use a thief probe to withdraw samples
from different locations in stationary powder mixtures.
Of signifcant interest are two essential sampling problems
that cannot be easily solved using currently available commercial
technology:
(i) Disturbances. The most common technique for obtaining
samples is to use a thief probe. Available probes can
introduce large errors in sample composition due to the
massive disturbances that take place during insertion of the
probe.
(ii) Segregation. It is well known among practitioners that
powder mixtures can segregate (unmix) upon handling.
Segregation can be a major sampling problem in any
sampling process involving dry powders because such
powders often segregate during insertion of the thief.
Thief samplers belong to two main classes: side sampling
and end sampling. A typical side sampling probe has one or more
cavities drilled or stamped in an inner cylinder enclosed by an outer
CA 02268849 1999-04-12
WO 98!18558 PCT/US97/19062
-3-
rotating sleeve. The sleeve has holes that align with the cavities,
allowing adjacent powder to flow into the cavities. Rotating the sleeve
to its closed position traps the particles into the cavities. An end
sampling thief has a single cavity at the end of the probe that can be
remotely opened and closed. In both cases, the thief is introduced into
the powder with its cavities closed. Once insertion is complete, the
cavities are opened, allowing the powder to flow into them. The
cavities are then closed, and thief is withdrawn, removing samples from
the mixtures.
Thief sampling is a rather laborious and cumbersome
technique and is rarely practical to take more than 10 or 20 samples. In
any sampling scheme, the experimentally measured variance, sot, is
actually a combination of the true variance resulting from the mixing
process, amt, the variance introduced by sampling error, 6s2 [Fan,
L.T., et al., Powder Technol., 61 (1990) 255], and the variance
resulting from analytical analysis, aa2, i.e.,
6e2 = 6m2 + as2 + 6a2 (1)
In an ideal situation, 6s2 and 6a2 are negligible, and ae2 (the variance
subject to USP rules) is almost identical to amt {the true variance).
Unfortunately, thief probes bias measurements to the point that sampling
uncertainty is expected to be a large fraction of the measurement
[Ashton, M.D., et al., Trans. Instn. Chem. Engrs., 44 (1966) T166;
Schofield, C., Powder Technol., 15 (1976) 169; Yip, C.W., et al.,
Powder Technol., 16 (1977) 189; Lai, F., et al., Chem. Eng. Sci., 36
(1981) 1133]. As mentioned earlier, two type of errors are often
introduced by thief probes: (i) the mixture is extensively disturbed
when the thief probe is inserted into the powder bed, and (ii) particles
of different sizes often flow unevenly into the thief cavities. Side-
sampling probes often have an additional problem: cohesive powders do
not flow easily into thief cavities, sometimes resulting in samples that
are smaller than desired.
Only a few studies have attempted to quantify the errors
introduces by thief probes mainly focusing on side-sampling thief
CA 02268849 1999-04-12
WO 98/18558 PCT/US17/19062
-4-
probes [Carley-Macauly, K.W., et al., Chem. Eng. Sci., 17 (1962) 493;
Schofield, C., Powder Technol., 15 (1976) I69; Poole, K.R., et al.,
Trans. Instn. Chem. Engrs., 42 (1964) T305; Masiuk, S., Powder
Technol., 51 (1987) 217; Gopinath, S., 27 f-i982) 321; Gayle, J.B. et al.,
50 (1958) 1279]. Carley-Macauley and Donald [Carley-Macauly, K.W.,
et al., Chem. Eng. Sci., 17 (1961) 493] performed a comparison study
between two types of side sampling probes. One was a conventional
probe (as described above); the other probe had cavities that were closed
using a longitudinal slit. They took samples from a system composed of
sand of two colors arranged in a layered structure. The conventional
probe gave samples of a mixed color within a region of two aperture
diameters (0.13") from the layer boundary. In the longitudinal slit
probe, sand particles tended to run down the slit, causing errors greater
than the conventional side sampling probe. In both cases, errors occur
because the probe is sampling locations that have already been disturbed
by the insertion of the probe itself. Other studies have reached similar
conclusions. For example, in a study conducted by Williams and Khan
[Williams, J.C., et al., Chem. Eng., (1973) 19], although no quantitative
data was reported, the authors concluded that a side sampling thief gave
totally misleading results in a segregating system. Instead, they used a
sampler that removed a core of powder from the bed. The sampler was
divided into sections in order to divide the core into samples. Orr and
Shotton [Orr, N.A., et al., Chem. Eng., London, January (1973)12]
determined that perturbations of the mixture structure are caused by
friction along the length of the probe, and are independent of the profile
of the tip. This result suggests that an end sampling probe will perform
better than a side sampling probe because for an end-sampling probe the
sample is taken from a relatively undisturbed region of powder beneath
the tip of the probe. They developed such a probe for use with cohesive
powders. Components sampled at depths of 1, 2, 4 and 6 cm through a
2 cm layer of charcoal powder qualitatively showed little contamination
of the samples with charcoal powder. In a quantitative analysis, 20
samples of cohesive calcium carbonate were taken through a layer of
cohesive lactose at a depth of 2 cm below an interface. The maximum
CA 02268849 1999-04-12
WO 98/18558 PCT/US97/19062
-5-
amount of lactose found in the samples was 0.07% [Orr, N.A., et al.,
Chem. Eng., London, January ( 1973) 12].
SUMMARY OF THE INVENTION
This invention relates to an end-sampling thief probe and a
method of using this probe. The probe is useful in extracting a sample
with minimal disturbance.
BRIEF DESCRIPTION OF THE FIGURES
The file of this patent contains at least one drawing
executed in color. Copies of this patent with color drawings) will be
provided by the Patent and Trademark Office upon request and payment
of the necessary fee.
Figure 1.
Schematic of the inner hollow rod with a inner conical tip mounted in
the outer hollow rod with a outer conical tip of the thief probe of the
invention (front view).
Figure 2.
Schematic of the inner hollow rod with a inner conical tip mounted in
the outer hollow rod with a outer conical tip of the thief probe of the
invention in the open position (side view).
Figure 3.
Schematic of the inner hollow rod with a inner conical tip mounted in
the outer hollow rod with a outer conical tip of the thief probe of the
invention in the closed position (side view).
Figure 4.
Schematic of the inner hollow rod with a inner conical tip mounted in
the outer hollow rod with a outer conical tip of the thief probe of the
invention with a stopper (front view).
Figure 5.
CA 02268849 1999-04-12
WO 98/18558 PCT/US97/19062
-6-
Schematic of the inner hollow rod with a inner conical tip mounted in
the outer hollow rod with a outer conical tip of the thief probe of the
invention with a stopper in the open position (side view).
S Figure 6.
Schematic of the inner hollow rod with a inner conical tip mounted in
the outer hollow rod with a outer conical tip of the thief probe of the
invention with a stopper in the closed position (side view).
Figure 7.
Performance of the Globe Pharma thief probe. (a) Disturbances caused
by insertion of the probe into particle bed (b) comparison of theoretical
(---) and experimental (~,~ ) results for 60 p,m particles over 200 ~.m
particle and (c) 200 ~.m particles over 60 ~.m particles.
Figure 8.
Performance of the thief probe of the invention. (a) Minimum
disturbances caused by insertion of the probe into particle bed (b)
comparison of theoretical (---} and experimental (~,~, 0) results for 60
~.m particles over 200 ~.m particle and (c) 200 ~.m particles over 60 ~m
particles.
DETAILED DESCRIPTION OF THE INVENTION
A thief probe comprising an outer hollow rod with a
hollow conical tip attached at one end of the outer hollow rod and an
inner hollow rod with a conical tip attached at one end of the inner
hollow rod;
said outer hollow rod having an inner diameter of about 1/4 in. to
about 63/32 in. and an outer diameter of about 9/32 in. to
about 2 in.;
the hollow conical tip of the outer hollow rod having an aperture;
the aperture of the hollow conical tip of the outer hollow rod._
being up to about 1/2 of the surface of the hollow conical
tip;
CA 02268849 1999-04-12
WO 98/18558 PCT/US97/19062
_7_
said inner hollow rod having an inner diameter of about 7/32 in.
and an outer diameter of about 62/32 in.;
the hollow conical tip of the inner hollow rod having an aperture;
the aperture of the hollow conical tip of the inner hollow rod
being up to about 1 /2 of the surface of the hollow conical
tip;
the outer hollow rod being shorter in length than the inner hollow
rod by about 3 in.;
the inner hollow rod being mounted in the outer hollow rod and
rotatable about the axis of the inner and outer hollow rods;
said inner and outer hollow rods being rotatable to an open
position and a closed position;
the open position being defined as the point where the apertures
of the hollow conical tips of the inner and outer hollow
rods are aligned so as to expose the cavity in the hollow
conical tip of the inner hollow rod; and
the closed position is the point where the inner and outer
apertures of the inner and outer hollow conical tips are
aligned so as not to expose the cavity in the hollow conical
tip of the inner hollow rod.
An embodiment of the invention is a thief probe comprising
an outer hollow rod with a hollow conical tip attached at one end of the
outer hollow rod and an inner hollow rod with a inner hollow conical
tip attached at one end of the inner hollow rod;
said outer hollow rod having an inner diameter of about 1/4 in. to
about 63/32 in. and an outer diameter of about 9/32 in. to
about 2 in.;
the hollow conical tip of the outer hollow rod having an aperture;
the aperture of the hollow conical tip of the outer hollow rod
being up to about 1/2 of the surface of the hollow conical
tip;
CA 02268849 1999-04-12
WO 98!18558 PCT/US97/19062
said inner hollow rod having an inner diameter of about 7/32 in.
to about 61/32 and an outer diameter of about 8/32 in. to
about 62/32;
the hollow conical tip of the inner hollow rod having an aperture;
the aperture of the hollow conical tip of the inner hollow rod
being up to about 1/2 of the surface of the hollow conical
tip;
the outer hollow rod being shorter in length than the inner hollow
rod by about 3 in.;
an inner solid rod having a diameter of 6/32 in. to about 60/32
in.;
said inner solid rod being mounted in the inner hollow rod and
being adjustable in height, so as to define the size of the
cavity of the inner hollow rod;
the inner hollow rod with the mounted inner solid hollow rod
being mounted in the outer hollow rod and rotatable about
the axis of the inner and outer hollow rods;
said inner and outer hollow rods being rotatable to an open
position and a closed position;
the open position being defined as the point where the apertures
of the hollow conical tips of the inner and outer hollow
rods are aligned so as to expose the cavity in the hollow
conical tip of the inner hollow rod; and
the closed position is the point where the inner and outer
apertures of the inner and outer hollow conical tips are
aligned so as not to expose the cavity in the hollow conical
tip of the inner hollow rod.
An embodiment of the invention is a thief probe comprising
an outer hollow rod with a hollow conical tip attached at one end of the
outer hollow rod and an inner hollow rod with a inner hollow conical
tip attached at one end of the inner hollow rod;
said outer hollow rod having an inner diameter of about 7/16 in.
and an outer diameter of about 8/16 in.;
CA 02268849 1999-04-12
WO 98/18558 PCT/US97I19062
-9-
the hollow conical tip of the outer hollow rod having an aperture;
the aperture of the hollow conical tip of the outer hollow rod
being up to about 1/2 of the surface of the hollow conical
tip;
said inner hollow rod having an inner diameter of about 11/32 in.
and an outer diameter of about 13/32 in.;
the hollow conical tip of the inner hollow rod having an aperture;
the aperture of the hollow conical tip of the inner hollow rod
being up to about 1/2 of the surface of the hollow conical
tip;
the outer hollow rod being shorter in length than the inner hollow
rod by about 3 in.;
the inner hollow rod being mounted in the outer hollow rod and
rotatable about the axis of the inner and outer hollow rods;
said inner and outer hollow rods being rotatable to an open
position and a closed position;
the open position being defined as the point where the apertures
of the hollow conical tips of the inner and outer hollow
rods are aligned so as to expose the cavity in the hollow
conical tip of the inner hollow rod; and
the closed position is the point where the inner and outer
apertures of the inner and outer hollow conical tips are
aligned so as not to expose the cavity in the hollow conical
tip of the inner hollow rod.
A thief probe as described above, wherein the inner and
outer hollow rod and their respective hollow conical tips are constructed
from an non-reactive material.
A thief probe as described above, wherein the non-reactive
material is selected from the group consisting of: aluminum, copper,
steel and bronze.
CA 02268849 1999-04-12
WO 98/18558 PCT/IJS97119062
- 10-
A thief probe as described above, wherein the non-reactive
material is aluminum.
A method for improved ending=sampling of a solid mixture
comprising the steps of:
(a) inserting a thief probe, as described above, in the closed
position into the solid mixture to a certain depth;
(b) rotating the inner hollow rod to the open position which
allows a sample of the solid mixture to fill the cavity;
(c) inserting the thief probe in the open position to a certain
depth so as to obtain the desired sample size;
(d) rotating the inner hollow rod to the closed position which
allows a sample of the solid mixture to be trapped in the
cavity; and
(e) removing the thief probe from the solid mixture.
Several experiments were recently performed to assess the
performance of two recently developed probes: a newly-released
commercially available side sampling probe (Globe-Pharma,
Piscataway, NJ), and an end-sampling thief described in this application.
The disturbances introduced by these probes were determined using two
procedures:
1) A qualitative assessment of the extent of perturbation of
the granular structure was performed by inserting the thief probes into
a system which consisted of several alternate one-inch layers of white
(1500 ~.) and red (600 ~,) glass beads. The granular beds were then
solidified by infiltration with gelatin without removing the probes. The
solidified beds were cut along the path of the probe and photographed.
2) Quantitative assessment of the errors introduced by each
thief probe was carried out by sampling structures which consisted of
CA 02268849 1999-04-12
WO 98/18558 PCTIUS97119062
-11-
. two layers of differently-sized beads. The layered structure consisted of
a 3 in. top layer over a 4 in. bottom layer. The beads were contained in
cylindrical b in. diameter cans. The number of samples taken during
each experiment was determined by the diameter of the probe being
tested. A sampling grid was made for each probe, with each sampling
location being at least two probe diameters away from any other
sampling location. Two systems were considered: 60 ~ beads on top of
200 p beads, and 200 ~ beads on top of 60 ~. beads. Samples were taken
at known positions above and below the interface between layers, and
the composition of such samples was compared with the theoretical
composition of the material at the sampling depth. In each case, samples
were taken at locations far enough apart that the disturbances caused by
previous insertions of the probe would not affect subsequent sampling.
The Globe Pharma thief probe has a side-sampling design
as described above. The thief probe has two cavities, and removable
dies fit into the cavities in order to control sample volume. In all of the
tests reported here, the lower cavity contained a 0.2 ml die while the
upper one was filled with a solid die. Experiments showed that
insertion of the Globe Pharma thief creates significant disturbances in
the mixture (Fig. 7a). Particles from upper layers are dragged deeply
into lower layers as the thief penetrates the granular bed. Once the thief
is opened, the sample flowing into the thief will be contaminated with
particles from positions along the path of insertion and will not
necessarily reflect the true composition of the system at the sampling
location before the thief was inserted.
Other types of errors are also possible. Data comparing
actual sample composition to theoretical composition expected from
sampling location are shown in Fig. 7b and 7c for the Globe Pharma
probe. The graphs show the percentage of the particles in the top layer
that are contained in each sample. Figure 7b shows results obtained for
60 p. beads on top of 200 ~. beads, and Figure 7c corresponds to 200 ~.
beads on top of 60 ~ beads. If the probe accurately samples the desired
location, the percentage of particles from the top layer should be zero
once the opening crosses the interface between the top and bottom
CA 02268849 1999-04-12
WO 98/18558 PCT/US97/19062
- 12-
layers. However, as shown in Figs. 7b and 7c, large sampling errors
are incurred, and the samples are composed entirely of particles from
the top layer regardless of sampling location. In order to identify the
source of these errors, additional experiments were performed, in
which the probe was introduced into the granular system and then
removed without ever opening the cavity. Upon removal of the thief, it
was observed that the cavity was filled with particles from the top layer.
These experiments show that free-flowing materials can enter the cavity
even before the Globe Pharma thief is opened. These errors could be
magnified in industrial applications. In our experiments the thief was
only inserted 3 to 5 inches below the upper surface of the granular bed,
while in industrial sampling depths may be as much as several feet. If
particles can flow into the thief while the cavity is closed, the resulting
sample will be a composite of the system along the path of penetration
of the thief, rather than the true composition of the undisturbed system.
The second thief tested is an end-sampling thief. The thief
probe consists of two concentric hollow pipes, both ending in a pointed
cone (an inner and an outer rod each having a hollow conical tip) (Fig.
$a). Half of each cone has been removed (each hollow conical tip
having an aperture) so that the sampling cavity of the thief can be set to
an open position (when the aperture ,of the inner and outer hollow rods
align so as to expose the cavity) and closed position (when the aperture
of the inner and outer hollow rods align so as not to expose the cavity,
the apertures are at 1$0 degrees to each other) by rotating the inner
hollow pipe (inner hollow rod). The outer hollow rod having an inner
diameter of about 7/16 in. and an outer diameter of about 8/16 in. and
the inner hollow rod having an inner diameter of about 11/32 in. and an
outer diameter of about 13/32 in. The length of the rods can be varied
according to the depth of the vessel being sampled, however the outer
hollow rod is shorter in length then the inner hollow rod by about 3 in.
The thief can be constructed from any material which will not react
with the samples being taken. Examples of such materials are
aluminium, copper, steel and bronze. The material used to construct the
thief used in the experiments described is aluminium. Additionally, it
CA 02268849 1999-04-12
WO 98118558 PCT/iJS97/19062
-13-
should be noted that the clearance between the inner and outer hollow
rods should be minimized. However, if the clearance allows for
particles to flow into the thief when in the closed position, a gasket can
be employed to fill in this clearance and prevent this from occurring.
The thief is used as follows: the thief is inserted in the
closed position into a vessel containing a powder mixture to be sampled
to the desired depth, the inner hollow rod is rotated to the open position
exposing the cavity of the inner hollow conical tip, the thief is inserted
further into the powder bed, the inner hollow rod is rotated to the
closed position closing the cavity of the inner hollow conical tip, and
then the thief containing a sample of the powder mixture is removed
from the mixture. Samples of consistent size were obtained by
controlling the depth of insertion after the cavity of the thief probe is
opened, as well as the diameter of the thief probe. Additionally, a
stopper may be used to set the sample size, as shown in Fig. 4-6. As
shown in Fig. 8a, the pointed cone design of this thief introduces much
smaller disturbances of the granular structure than the previous device.
Disturbances are minimal near the tip of the probe, where the sampling
takes place. Quantitative comparison of theoretical and experimental
sampling data is shown in Fig. 8b for 60 p, particles on top of 200 p
particles, and in Fig. 8c for 200 p particles on top of 60 p particles.
The data indicates that the thief probe of the invention performs much
better than the Globe Pharma thief probe.
