16-6-22

PF Online

A Discussion of Net Water Gain for Water Vapor Transmission Rate Determinations
S. Yoon, Eli Lilly and Company, D. Sparks, Eli Lilly and Company, S. Selke, Michigan State Universitya
ABSTRACT This Stimuli article discusses the USP General Chapter ContainersPermeation 671 method to
determine net water gain. This article introduces a proposal to change the procedure based on an analysis of the
steady-state conditions of the plastic walls of containers. The analysis discussed in this paper shows that the
plastic walls of controls have a higher water concentration than do the plastic walls of test containers and that
the current Water Vapor Transmission Rate calculation in USP 671 is not theoretically correct. Although this
proposal does not have an important impact on test results, this method change advances the scientific basis by
providing more accurate test results.

INTRODUCTION
USP General Chapter ContainersPermeation 671 (1) provides a method to determine the net water gain of
anhydrous calcium chloride or other desiccant in a plastic bottle containerclosure system. The net water gain is
used to calculate water vapor transmission rate (WVTR) of the sealed containerclosure system. The current
method employs test containerclosure systems filled with desiccant (test containers) and containerclosure
systems filled with glass beads (controls). The net water gain is determined by subtracting the weight gain of
controls from the weight gain of test containers. This is then used as the weight gain of desiccant only. This
method is based on two assumptions: 1) the glass beads do not absorb water, and 2) the test container walls
and control walls absorb the same amount of water.
However, the net water gain calculated in this manner is not theoretically correct because the water content of
the plastic walls of the controls is not the same as the water content of the plastic walls of test containers.
Theoretically, the plastic walls of controls achieve a higher water concentration than do the plastic walls of test
containers at the steady-state conditions of the test. Therefore, the subtraction of a higher water concentration in
controls from the lower concentration in test containers results in a net water gain that is lower than the actual
net water gain. Further, this would result in a WVTR that is lower than the actual WVTR.
Plastic (e.g., polyethylene and polypropylene) container walls reach a steady state of water concentration. The
steady-state condition is a gradient that depends on the storage conditions and the conditions of the headspace
in the container. This gradient is maintained as constant during testing if the conditions of storage and the
container headspace are constant. This Stimuli article proposes an alternative approach for determining net water
gain and calculating WVTR of plastic containerclosure systems (bottles and blisters).

DISCUSSION
1. Water Vapor Transmission into Plastic Container Walls
TheoryWater vapor from a high vapor pressure environment permeates plastic container walls of a sealed
container that has a low water vapor pressure internal environment. This is described by Equation (1), Fick's
Second Law of Diffusion:

where C is the concentration of water in the plastic container wall, t is time, D is the diffusion coefficient, and x is
the spatial coordinate in the direction of transfer (2,3).
Figure 1 describes Fick's Second Law of Diffusion graphically and shows a graphical representation of water
concentration across the plastic container wall as a function of time (t). The total change in water concentration
across the plastic container wall with time is directly proportional to the change in concentration gradient with the
www.usppf.com/pf/pub/index.html

1/6

16-6-22

PF Online

permeant penetration depth. At a steady-state rate of water transmission, the concentration gradient remains
constant across the plastic wall.

Figure 1. Graphical representation of water concentration across the plastic container wall as a function of time
(t).
ExampleIn the standard test, test containers with desiccant and controls with glass beads are prepared at
ambient conditions [e.g., 21 C, 30% relative humidity (RH)] with the following assumptions:
The container itself reaches equilibrium at 21 C/30% RH.
The desiccant is completely dried.
A sufficient amount of desiccant is inserted in the test containers to achieve and maintain an atmosphere inside
the container close to 0% RH during the test. RH is used to express the water concentration in the headspace
and in the walls of the container. Figure 2 shows the initial water concentration across the plastic wall of a test
container and a control. Because the plastic wall does not provide a good temperature barrier, the temperature
across the plastic wall, inside and outside, is 21 C.
Desiccant is assumed to absorb the water immediately in the headspace of test containers. Therefore, the RH of
test container headspace with desiccant is close to 0% and the RH of the plastic wall of test containers is close
to 30%. It is assumed that glass beads do not absorb water in controls. Therefore, the RH of the control
headspace with glass beads is close to 30% and the RH of the plastic wall of the controls is close to 30%.

Figure 2. Diagram showing initial RH (i.e., water concentration) outside the container, across the plastic
container wall, and inside the plastic container: initial state at sample preparation.
In the standard test, test containers and controls are placed in a storage chamber (e.g., 23 C/75% RH) to allow
study of the weight gain of desiccant due to water transmission into the containers and absorption of water by
the desiccant. Figure 3 shows the water concentration across the plastic wall of the test containers and controls
after the containers are placed in the 23 C/75% RH chamber. The water in the storage chamber first adsorbs on
the surface of the plastic container wall. Therefore, the RH of the surface will change immediately to that of the
storage RH condition (75% RH). Because the plastic walls do not provide a good temperature barrier, the
temperature across the plastic wall, inside and outside, reaches 23 C essentially immediately.

www.usppf.com/pf/pub/index.html

2/6

16-6-22

PF Online

Figure 3. Diagram showing RH (i.e., water concentration) outside the container, across the plastic container wall,
and inside the plastic container immediately after containers are placed in the 23 C/75% RH chamber: initial
state upon storage.
Water will transmit through the plastic wall because the vapor pressure outside the container is greater than the
vapor pressure of the wall itself and of that inside the container. Figure 4 shows the water transmission and the
resulting water concentration across the plastic wall. The arrows show a progression with increasing time. For
test containers, the plastic material has a lower water concentration than the storage condition and has a higher
water concentration than the container headspace. Therefore, the plastic wall absorbs water from the 75% RH
chamber environment and desorbs water to the 0% RH headspace. The absorbed water diffuses through the
plastic wall. The diffused water desorbs from the plastic wall to the 0% RH container headspace, where it is
immediately absorbed by the desiccant, maintaining the container headspace at 0% RH. This results in the
water concentration gradient shown in Figure 4. This water concentration gradient across the plastic walls may
reach a steady-state condition (equilibrium) within a few days for typical pharmaceutical packages such as highdensity polyethylene (HDPE) bottles and polychlorotrifluoroethylene (PCTFE) blisters. See Appendix for more
information about the equilibrium time.
For controls, the plastic material has a lower water concentration than the storage condition and has the same
water concentration as the container headspace. Therefore, the plastic wall absorbs water from the 75% RH
chamber environment, and initially there is no transfer between the plastic wall and package headspace. The
absorbed water diffuses through the plastic wall. The diffused water desorbs from the plastic wall to the container
headspace, where it is not absorbed by the glass beads, resulting in the water concentration gradient shown in
Figure 4. The water concentration gradient across the plastic wall reaches a steady-state condition (equilibrium).
The time to reach equilibrium for controls is longer than for test containers because the water concentration
gradient changes dynamically as the inside vapor pressure increases. It is assumed that the equilibrium time for
controls is not more than 2 weeks. At the equilibrium condition, the water concentration across the plastic wall is
constant at 75% RH.

Figure 4. Graphical representation of water concentration across the plastic container wall as a function of time
(t) after containers are placed in the 23 C/75% RH chamber.
Conclusion, Part 1USP 671 uses 14 days or 28 days to determine the net weight gain. After 14 or 28 days
of storage, the water concentration gradient in the plastic material itself reaches steady-state condition. The
water concentration (hence, the weight gain) of the plastic wall of controls is greater than the concentration
(hence, the weight gain) of the plastic wall of test containers. Practically, for small or thin-walled containers, the
moisture gain of container walls may not be significant and may not affect the accuracy of experimental results
provided by USP

671 .

2. Correction of Current Net Water Weight Gain Equation

The analysis of the water weight gain of plastic bottle containerclosure systems has been important to the
pharmaceutical industry for many years. USP 671 prescribes this analysis, currently, through the use of
Equation (2) to calculate the net water gain of containers. The weight gain of controls (C) filled with glass beads
is subtracted from the weight gain of test containers (T) filled with desiccant:
www.usppf.com/pf/pub/index.html

3/6

16-6-22

PF Online

W net = (TF

TI )

(CF

CI )

(2)

where W net is the net weight gain determined by subtracting the difference between the final (CF) and initial (CI )
weights of the controls from the difference between the final (TF) and initial (TI ) weights of test containers.
Equation (2) assumes that the water concentrations in the plastic walls of test containers and controls are equal
at the same time of storage. However, the water concentration is not equal, as explained in the previous section.
The plastic wall of controls has a higher water concentration than the plastic wall of test containers. Therefore, it
is not correct to subtract the water weight gain of controls from the water weight gain of test containers to
determine the net water gain of the test containers. That is, the water gain of containers calculated by Equation
(2) results in a value that is less than actual. Hence, Equation (2) is not theoretically correct.
Instead, the steady-state (equilibrium) condition of test containers described in the previous section can be
utilized in this determination without using the water weight gain of controls to determine the net water gain of
containers. Test containers are placed at the desired condition for preconditioning to achieve a steady-state
concentration gradient. (See Test container in Figure 4.) Typical plastic containers [e.g., HDPE, PP,
polyethylene terephthalate (PET), and PCTFE] used in the pharmaceutical industry can achieve the steady-state
concentration gradient within 1 or 2 days. When a sufficient amount of completely dried desiccant is used in test
containers, the preconditioning does not affect the water vapor pressure of the container headspace. The
constant water vapor pressure difference can be achieved during the test. Therefore, the net water gain can be
measured by Equation (3) after preconditioning.
W net = (TF

TIC)

(3)

where W net is the net weight gain determined from the difference between the final (TF) and initial (TIC) weights
of test containers. The initial weight (TIC) is determined after preconditioning.

CONCLUSION
This article has reviewed the use of controls to determine the net water gain for plastic containerclosure
systems provided in USP 671 and has shown that the procedure is not theoretically correct. Theoretically,
the net water weight gain would be underestimated by subtracting the weight gain of controls from the weight
gain of test containers. Therefore, it may result in a WVTR that is lower than the actual WVTR. The newly
proposed simple equation with preconditioning of plastic containers determines net water gain more accurately,
theoretically, and offers an improvement to the precision of the WVTR method.

ACKNOWLEDGMENTS
The authors would like to acknowledge Drs. Maria Rubino and Rafael Auras at Michigan State University for the
WVTR experiments using the moisture analyzer equipped with a microbalance as referenced in this paper. The
authors also would like to thank Randy Thackrey (Eli Lilly and Company) for constructive technical comments
that have been incorporated in this paper.

REFERENCES

1. USP. USP 30NF 25, ContainersPermeation

671 . Rockville, MD: USP; 2007:267269.

2. Crank J. The Mathematics of Diffusion. 2nd ed. Bristol, UK: Clarendon Press; 1979.
3. Selke S. Plastic Pack aging. 2nd ed. Munich, Germany: Hanser Publishing; 2004.

APPENDIX: Moisture Permeation of 2-mil PCTFE Blisters

www.usppf.com/pf/pub/index.html

4/6

16-6-22

PF Online

Moisture permeation of 2-mil PCTFE blisters has been determined with a VTI moisture analyzer. The analyzer
can maintain constant environmental conditions and weigh the sample during the storage time. If the sample
package contains adequate desiccant to maintain the inside vapor pressure consistently at 0, then a constant
vapor pressure difference between inside and outside of the package can be achieved. This is identical with the
approach in USP 671 and ASTM E96 gravimetric WVTR methods. The benefit of this method is that the
sample can be weighed without taking the sample out of the chamber. Therefore, the analyzer can show the
weight change immediately after the sample is mounted on the balance. It shows the weight change from a
dynamic (unsteady) state to the steady state. The time to reach the steady-state condition can be estimated
from this experiment.
Materials and Equipment
The 2-mil PCTFE film (supplied by TekniPlex,) is structured as 2-mil PCTFE/2-mil PE/7.5-mil PVC. The lidding
foil (supplied by Alcan) is structured as paper/adhesive/ polyester/adhesive/1-mil aluminum foil/heat-seal coating.
The size of the individual blister cavity is 11 mm diameter 5 mm height. A 4(supplied by SudChemie) was used.

molecular sieve desiccant

A symmetrical gravimetric analyzer (SGA 100) was newly equipped with a CAHN 200 microbalance at Michigan
State University. The moisture analyzer and balance can measure the weight change of the sample with 5 g
accuracy, control the temperature with 0.5 C accuracy, and control the RH with 1% RH accuracy.
Methods
Blister samples were prepared with a Klockner EAS blistering machine. The 2-mil PCTFE film was
thermoformed, and completely dried molecular sieve desiccant tablets were inserted manually into the blister
cavities. The desiccant-filled blisters were heat-sealed immediately with aluminum foil lidding.
Moisture permeation of 2-mil PCTFE blisters was determined gravimetrically by using the VTI moisture analyzer
equipped with a microbalance. The blister sample (6 cavities, approximately 4 g weight) was mounted on the
weighing dish in the chamber, which is connected to the microbalance. The system was set up to maintain the
relative humidity at 60% and the temperature at 30 C for 2000 minutes (1.4 days), and data were collected
every 2 minutes. The test was replicated with 2 different samples.
Results
The system maintained the desired RH and temperature consistently during the entire testing period. The
temperature was maintained between 29.83 C and 29.91 C, and the RH was maintained between 59.6% and
60.3% for 2000 minutes. Figure 5 shows the weight change of samples stored at 30 C/60% RH. Initially, the
weight change rate was in an unsteady state, but the weight change rate reached a steady state before 0.5 day
for both samples.

Figure 5. Recorded sample weight change from the VTI system.

Conclusions
A steady-state rate of weight change for 2-mil PCTFE blisters was obtained before 0.5 day at 30 C/60% RH. If
the storage condition is changed, the equilibrium time may be increased or decreased. In addition, if the material
www.usppf.com/pf/pub/index.html

5/6

16-6-22

PF Online

thickness and barrier property increase, the equilibrium time will be increased. The equilibrium time should be
determined experimentally but probably will not require more than a few days with typical plastic package
materials (e.g., HDPE, PP, PET, and PCTFE) at typical stability conditions such as 25 C/60% RH, 30 C/65%
RH, and 40 C/75% RH.
a Correspondence should be addressed to: Desmond G. Hunt, PhD, Scientist, Documentary Standards Division, US
Pharmacopeia, 12601 Twinbrook Parkway, Rockville, MD 20852-1790; tel. 301.816.8341; e-mail dgh@usp.org.

www.usppf.com/pf/pub/index.html

6/6