White Paper

Drying Oven vs. Halogen Moisture Analyzer

A Practical Guide to Compare Methods

This white paper will be of interest to anyone involved in moisture analysis

applications in pharmaceutical, chemical, food and other industries.

Table of Contents

1. Introduction

2. Overview of moisture analysis

2.1. The importance of moisture content
2.2. Loss on drying using the drying oven method
2.3. Halogen moisture analysis: a fast alternative

3. 
Practical guide on how to replace the drying oven by a halogen moisture analyzer
3.1. Evaluation of comparability based on process requirements
3.2. Evaluation of comparability by statistical means

4. Accurate moisture determination with halogen moisture analyzers

4.1. Sound drying method and good sample handling
4.2. Accurate instruments

5. Conclusion

Appendix 1: Exemplary method comparison at a specific moisture content based

on process requirements
Appendix 2: Exemplary method comparison at a specific moisture content by
statistical means
Appendix 3: Exemplary method comparison over a range of moisture content by
statistical means

6. References
 Moisture affects the quality, shelf-life and usability of many products, including pharmaceutical substances,
White Paper
plastics and foodstuffs. Therefore, monitoring and determination of the moisture content in samples is an impor-
tant application. Typically, loss on drying (LOD) using a drying oven is used as reference method, but this can
be slow with many manual steps. Faster determination of moisture content can be achieved with new methods,
such halogen moisture analyzers (HMA), which are easy to use and give a direct result in a fraction of the time.
The challenge has been how to validate the HMA method and prove that the results are comparable with the dry-
ing oven method. This white paper will describe how.

In some industries, such as plastics, the HMA method has already been established as the new standard test
method by ASTM (formerly the American Society for the Testing of Materials, now known as ASTM International)

.
1 Introduction
In the field of moisture determination using loss on drying techniques, a common question is:

"Can the drying oven method be replaced by fast halogen moisture analysis?"

The simple answer is yes, as long as the results obtained by the two methods are comparable. This means that
it is necessary to show evidence that the results are equivalent within specific tolerances, which is not such a
straightforward question to answer.

This white paper guides the analyst through this process. It explains the key decision criteria surrounding the
choice of method and provides practical guidance on how to demonstrate that the two different methods (dry-
ing oven and halogen moisture analyzer) for establishing the moisture content of a sample deliver comparable
results. In addition, two alternative and acceptable comparison approaches are outlined here: the first approach
is based on specific process requirements (tolerances), and the second approach is based on statistical analy-
sis of the data obtained.

2 Overview of moisture analysis

2.1. The importance of moisture content

Moisture content is a key quality parameter in most industries, including food, chemical, and pharmaceutical
industries. Moisture content determines the quality and the cost of raw materials; it affects product quality (e.g.
shelf life) and often influences the financial margin of finished goods. Moisture is also a key parameter for pro-
cess control in many production processes. For these reasons, monitoring of moisture content is very important.
These moisture determinations need to be carried out rapidly and reliably so that any interventions in the produc-
tion process can be made promptly, to avoid costly interruptions.

2.2 Loss on drying using the drying oven method

For many substances the admissible moisture limits and the applicable measurement method have been estab-
lished by government agencies (e.g. USP monographs [1,2]) or industry commissions (e.g. ICUMSA method for
sugar [3]). Therefore, the specific measurement method for a given sample is called reference method. Typically,
loss on drying using the drying oven method is used as reference method. LOD is robust and reliable, provid-
ing good results and only requires standard laboratory equipment (a drying oven and typically an analytical
balance). The LOD method however is slow, usually requiring 2-3 hours or more for a measurement to be con-
cluded and tedious due to many manual steps in the process. LOD is not suited for use on the factory floor, as
the it takes too long to achieve a result and qualified lab personnel are required.

2 White Paper
METTLER TOLEDO
White Paper 2.3 Halogen moisture analysis: a fast alternative
Halogen moisture analyzers also operate on the principle of LOD, but offer a much faster alternative to the drying
oven method. Measuring moisture with an HMA normally takes 5 to 15 minutes. The other important advantage
of the HMA is that they are easy to operate, providing a direct measurement with no calculations necessary. This
makes halogen moisture analyzers well-suited for conducting reliable measurements both in the laboratory envi-
ronment and at the production line by factory staff during shift operations. In some industries the halogen mois-
ture analyzer method has already been established as an accepted method. For example, the ASTM published in
2012 a standard test method for determination of moisture in plastics use of an HMA [4].
A comparison of the two methods, along with advantages and disadvantages is presented in Table 1.

Drying Oven Halogen Moisture Analyzer

Principle Thermogravimetry Thermogravimetry

Heating of sample by convection. Sample is Heating of sample through absorption of IR

dried in the oven for a defined period of time radiation from a halogen radiator.
Measuring at constant temperature. Mass is determined Continual determination of mass during drying
method before and after drying. The moisture content process. The moisture content percentage is
percentage is determined from the difference in determined from the difference in weight before and
weight before and after drying. after drying.

• Often reference procedure (for historical • Quick measurement (typically 5 – 15 min.)

reasons this procedure often forms part of • Simple handling, no calculations
legislation) • Compact instrument. No balance or
Advantages
• Several samples can be determined at the dessicator required
same time • Suitable for at-line use
•  Large sample volumes possible

•  Very long determination period (hours)

•  Substances other than water may evaporate
• Prone to errors because of the high level of
Disadvantages •  Substances other than water may evaporate
handling and calculations involved
• Unsuitable for at-line use - requires analytical
balance and dessicator

Table 1: Comparison of drying oven and halogen moisture analyzer methods for determining moisture content of a sample

3 Practical guide on how to replace the drying oven by a

halogen moisture analyzer
The drying oven method can be replaced by the halogen moisture analyzer, if the results of the two methods
are comparable. This chapter describes how to verify comparability. Two approaches are described to establish
that the drying oven and HMA provide equivalent results: the first evaluates the comparability based on process
requirements; the second is based on statistical data comparison.

In practice, the first approach is typically applied, as the acceptance criteria for comparability take the specific
process context into account.

3.1 Evaluation of comparability based on process requirements

Broadly accepted guidelines for comparability are the pharmaceutical industry guidelines. For example, the
United States Pharmacopeia (USP) Chapter <1010> "Analytical Data – Interpretation and Treatment" [5], states
that an alternative method (in this case the HMA method) is comparable, if its results do not differ from the refer-
ence method (the drying oven method) by more than “an amount deemed important” [6]. To evaluate equiva-
lency of the methods, their precision [7] and accuracy [8] should be compared. The decision on whether differ-

White Paper
METTLER TOLEDO
3
 ences found between the two methods are within an acceptable range must be taken within the specific context
White Paper
of the application. This is based on the accepted tolerances in moisture content (%MC tolerances) of a produc-
tion process, e.g. ‘Statistical Process Control’ [9]).

The typical and well-proven approach for comparison of drying oven versus HMA methods is to apply a range of
acceptance to the mean value and standard deviation of the drying oven results, and then to verify that the HMA
results are within this range (see example in Table 2).

Parameter Unit Acceptance criteria Exemplary acceptance criteria

(formula / value)

Accuracy %MC Δ%MC (DO – HMA) = ¦ %MCDO - %MCHMA ¦ Δ%MC (DO – HMA) ≤ 0.1%MC: excellent
(where ‘¦ ¦’ is the absolute value) Δ%MC (DO – HMA) ≤ 0.2%MC: good
Δ%MC (DO – HMA) ≤ 0.4%MC: acceptable
Δ%MC ((DO – HMA) > 0.4%MC: failed

Precision SD Q = SDHMA / SDDO Q ≤ 1.5 : good

Table 2: Exemplary tolerances applied as acceptance criteria for samples within a moisture range of ~2%MC to ~15%MC.

Note: these values are exemplary and it is the responsibility of the operator to verify their suitability for a specific process. For samples out-
side this moisture range other values may become applicable.

Definitions:
%MCDO = mean of at least 6 measurements utilizing the drying oven method.
%MCHMA = mean of at least 6 measurements utilizing the HMA method
Q = the quotient of SDHMA and SDDO
SDDO = the standard deviation of at least 6 measurements utilizing the drying oven method
SDHMA = the standard deviation of at least 6 measurements utilizing the HMA method

As demonstrated in the method collection for Pharma Excipients [10], a precision of the HMA method that is
equal or less than 1.5 times the precision of the DO method is typically achievable (see section 4.1 below for
further information).

Please refer to Appendix 1 for an exemplary method comparison for Ethyl Cellulose based on process compara-
bility. If the comparability of drying oven and HMA methods shall be verified over a moisture range (e.g. between
1.00%MC and 8.00%MC), it is recommended to verify accuracy and precision at multiple (typically three, e.g. at
1.00%MC, 4.50%MC, 8.00%MC) moisture values representing the moisture range of interest.

Analysts, however, will often decide to focus on accuracy and precision at the critical moisture content only.

3.2 Evaluation of comparability by statistical means

Statistical methods may be applied to evaluate comparability of drying oven and HMA methods of LOD, as dis-
cussed in USP <1010> [5]. A proven statistical method for method comparison at a specific moisture content is
to apply the generic statistical tool Student t-test [11, 12] which tests the statistical significance of the differences
between the drying oven and HMA methods. If the differences are not found to be significant, the methods are
considered equivalent. For method comparison over a moisture range, the linear regression analysis is often
applied.

Other than the process requirement based approach described above (chapter 3.1), the statistical methods
(e.g. Student t-test, linear regression analysis) compare two sets of data (results), testing if they are statistically
equivalent. Statistical methods only look at the population of the samples and cannot take acceptable differences
between the drying oven and HMA methods into account (typically process requirements allow for some differ-
ences in the results of the methods). The use of statistical methods may therefore lead to unnecessary restrictive
acceptance criteria for the HMA method; subsequently the HMA method may be unnecessarily rejected, leading

4 White Paper
METTLER TOLEDO