Fresenius J Anal Chem (2001) 371 : 944–950

DOI 10.1007/s002160101058

O R I G I N A L PA P E R

Michael Krachler

Critical assessment of the performance

of electronic moisture analyzers for small amounts
of environmental samples and biological reference materials

Received: 29 March 2001 / Revised: 11 July 2001 / Accepted: 17 July 2001 / Published online: 19 September 2001
© Springer-Verlag 2001

Abstract Two electronic moisture analyzers were criti-

cally evaluated with regard to their suitability for deter- Introduction
mining moisture in small amounts (≤200 mg) of various
environmental matrices such as leaves, needles, soil, peat, The accurate and precise determination of moisture con-
sediments, and sewage sludge, as well as various biologi- tent can be very important for a wide range of physical
cal reference materials. To this end, several homogeneous and chemical properties. Frequently, the determination of
bulk materials were prepared which were subsequently the moisture content of a specific material is of the utmost
employed for the development and optimization of all an- importance. During the production of tobacco products,
alytical procedures. The key features of the moisture ana- for example, processing of the raw material includes steps
lyzers included a halogen or ceramic heater and an inte- such as threshing, cutting, casing, blending, ageing, ex-
grated balance with a resolution of 0.1 mg, which is an es- pansion, and flavoring [1]. To obtain a satisfactory quality
sential prerequisite for obtaining precise results. Oven dry- of the end product, all these processes must be carried out
ing of the bulk materials in a conventional oven at 105 °C at optimal specific moisture levels, varying between 10
until constant mass served as reference method. A heating and 50% [1]. Sediment samples used in geochemical stud-
temperature of 65 °C was found to provide accurate and ies may contain 60% of water which needs to be accu-
precise results for almost all matrices investigated. To fur- rately assessed before further determinations of specific
ther improve the accuracy and precision, other critical pa- analytes [2]. In food products the moisture content influ-
rameters such as handling of sample pans, standby tem- ences storage conditions and shelf life as well as textural
perature, and measurement delay were optimized. Because and technological qualities, and thus needs to be moni-
of its ponderous heating behavior, the performance of the tored [3].
ceramic radiator was inferior to that of the halogen heater, For analytical chemists, determination of the moisture
which produced moisture results comparable to those ob- content of a reference material is an essential prerequisite
tained by oven drying. The developed drying procedures for any reliable quality control measures. Its knowledge is
were successfully applied to the fast moisture analysis necessary to report exactly the mass of analyte per unit
(1.4–6.3 min) of certified biological reference materials of dry mass of sample. Trace element analysis of a biologi-
similar provenance to the investigated the bulk materials. cal reference material, for example, will generate lower
Moisture results for 200 mg aliquots ranged from 1.4 to results than the certified elemental values if the moisture
7.8% and good agreement was obtained between the rec- content is not correctly considered. Powdered certified
ommended drying procedure for the reference materials reference materials normally contain between 0.3 and 3%
and the electronic moisture analyzers with absolute uncer- of moisture [4]. These low levels of moisture in the refer-
tainties amounting to 0.1% and 0.2–0.3%, respectively. ence materials positively influence the long-term stability
of the powders. However, depending on their hygroscopic
nature and the conditions under which they are stored,
their moisture content may increase to as much as 10% [5].
It should be emphasized that the moisture present in a
material is not equivalent to its water content. The mois-
ture of a material includes all the volatile components such
M. Krachler (✉) as water (by far the most abundant component), but also
University of Heidelberg,
Institute of Environmental Geochemistry,
comprises fat, oil, alcohol, organic solvents, etc. which are
Im Neuenheimer Feld 236, 69120 Heidelberg, Germany evaporated when a sample is heated [6]. Therefore, mois-
e-mail: [email protected] ture is an operationally defined term, which includes mainly
 945

water but also other volatile substances present in the drying procedure exactly specifying the needs of the in-
samples which are evaporated under given experimental vestigation. Therefore, the aim of the present study is to
drying conditions. In this context it should be noted that develop a reliable and fast method for the determination
many terms describing the different physical states of wa- of the moisture content in biological matrices, providing a
ter binding exist [3]: bound water, sorbed water, solid-like basis for important measures to correct the results of trace
water, water of crystallization, imbibed water, water of element analysis for the moisture of the samples investi-
hydration, immobilized water, etc. The various terms have gated. Giving adequate consideration to this important is-
different definitions or methods of determination, result- sue of moisture analysis will facilitate reliable comparison
ing in different quantitative values [3]. As drying methods of analytical results from different laboratories if they are
in general do not distinguish between water and other based on dry matter.
volatile substances, experimental parameters have to be
chosen very carefully to obtain correct results. A recent
inter-laboratory study [7] on the determination of the Experimental
residual water content in candidate reference material
BCR 679 cabbage powder determined by various drying Instrumentation
methods revealed that the results ranged from 0.24% to Two identical electronic moisture analyzers (MA100, Sartorius,
13%. Göttingen, Germany), equipped with a motorized ceramic or a halo-
Basically, two kinds of methods are available to deter- gen heater generating infrared (IR) radiation, were used for sample
mine the moisture content of a substance. Absolute meth- drying. The resolution of the balance, which is an integral part of
the moisture analyzers, can be set to 0.1 or 1 mg. The internal
ods such as oven drying, infrared drying, vacuum drying, printer provides user-definable printouts that can be customized
and microwave drying are based on thermogravimetry, i.e. before moisture analysis runs.
recording the loss of mass of a substance that occurs when As a reference method, aliquots of samples were weighed to
it is heated. Deductive methods determine the moisture 0.1 mg on an analytical balance (Sartorius) and subsequently oven
dried at 105 °C in a conventional laboratory oven (type T12,
content indirectly by measuring a physical property – for Heraeus, Hanau, Germany) to constant mass. Reference materials
example absorption of electromagnetic radiation – which were dried according to the “Instructions for Drying” of the manu-
is related to the moisture content of the substance. The lat- facturers.
ter methods include Karl-Fischer titration, and infrared
and microwave spectroscopy. Preparation of bulk materials
Among the absolute methods, conventional oven dry-
ing is the most popular. To this end, normally an aliquot After collection, the various samples (lake sediments, soil, sewage
(0.5 to 1.2 g) of a material is oven dried at 105 °C for 3–4 h, sludge, peat, leaves, needles) were air-dried in the laboratory at
22 °C. Subsequently, specimens were milled in an ultracentrifugal
depending on the matrix and analytes of interest, before mill (ZM 1000, Retsch, Haan, Germany). The dry, coarse materi-
cooling the dried samples in a desiccator for reweighing als were passed through 200 µm and 100 µm sieves (Retsch, Haan,
[7]. This approach has several disadvantages including the Germany), respectively. The obtained fine powders (200–300 g)
potential loss of specific volatile analytes, possible decom- were transferred into 1 L glass bottles and were homogenized by
position/oxidation of the analytes and/or the matrix, as well manual shaking of the bottles for at least 10 min. The homoge-
nized powders were subsequently left to stand for another two
as the long time required to obtain the results. weeks in the closed bottles to allow the moisture to distribute ho-
The humidity of the working environment affects the mogeneously. Because the sample containers were not completely
moisture content of the reference material in the bottle, filled, the powders can exchange their moisture with the air remain-
necessitating repetitive analysis from time to time [7, 8]. ing in the container. Therefore, potentially condensed moisture
was mixed back in with the powders by manual shaking prior to
Sometimes only a very limited amount of “precious” ma- use.
terial is available for analysis, e.g. ancient samples from
geological archives such as sediments or herbarium mate-
rials. However, the moisture content needs to be determined Method development
also in these specimens to correct for a possible underes- The temperature settings of the heating devices of both moisture
timation of the concentration of specific analytes because analyzers were varied between 40 and 140 °C in 5° increments for
of dilution by residual moisture in the sample. In this con- various matrices at a given sample mass to search for a drying tem-
text, electronic moisture analyzers offer an attractive al- perature resulting in comparable moisture values to those of the
standard drying procedure in a conventional oven. Additionally,
ternative to conventional oven drying as regards both the the moisture content was investigated as a function of the sample
required amount of sample and the analysis time [8]. mass (0.1 to 5.0 g) for selected specimens. Various other instru-
The purpose of this study is to critically assess the suit- mental parameters (measurement delay, standby temperature, proper
ability and performance of electronic moisture analyzers handling of sample pans (aluminium, 92 mm i.d.), behavior of the
two different heating devices) were critically evaluated and opti-
equipped with various heating devices to determine the mized.
moisture content in small amounts (≤200 mg) of a variety
of environmental samples. Electronic moisture analyzers
are not designed to determine the absolute water content Procedures
of a material, as, for example, is the case with Karl Fischer Prior to use, the sample pans were heated in the moisture analyzers
titration. As a relative approach, electronic moisture ana- to 65 °C for at least 1.4 min without any sample to remove residual
lyzers have to be calibrated against another well established moisture. Sample pans were only handled with a pair of tweezers,
946

Results and discussion

Basic considerations

In a conventional drying oven, circulating hot air warms

the sample from the outside to the inside. Efficiency is
lost during drying because as the moisture evaporates it
cools the sample surface. In contrast, infrared (IR) rays
penetrate a sample without being impeded. On reaching
the interior of the sample, they are converted into heat en-
ergy, which promotes evaporation, thus drying the sam-
ple. Depending mainly on the color of the sample, a small
part of the IR radiation is reflected from the sample sur-
face. Light substances reflect IR radiation to a greater ex-
Fig. 1 Termination of moisture analysis using the “Fully Automatic
Mode” with the MA 100 moisture analyzer
tent than dark samples. This is the drying principle of both
the ceramic and the halogen heater used in the moisture
analyzers employed in this study. The ceramic heater
for reasons given in the Discussion. On completion of a measure- mainly emits radiation in the middle IR range between
ment, the powder on the sample pan was removed by overturning wavelengths of 2.8 and 5.0 µm, whereas the halogen
the pan. To avoid potential errors arising from use of a warm sam- heater emits a broader range of radiation. For water, the
ple pan, two sample pans were employed alternately. After every
measurement, the sample pan was gently cleaned with a soft tissue optimal absorption of IR radiation, however, is found be-
(igefa, Berlin, Germany) and conditioned as described above. tween 3 and 5 µm. Because of the higher drying effi-
All optimization experiments and measurements were performed ciency, the temperature setting selected during IR drying
with the maximum resolution setting of 0.1 mg of the balance of is usually lower than the temperature used when working
the moisture analyzers.
To determine when analysis is finished, the fully automatic with a drying oven [6, 8].
mode of the moisture analyzers was chosen because this easy-to- Beside the appropriate choice of the drying tempera-
operate mode provided largely reproducible results. The principal ture, the resolution of the integrated balance of the mois-
criterion for completing the moisture analyses is based upon the ture analyzer is of fundamental importance for the preci-
calculation of the point of inflection on the analysis curve as illus-
trated in Fig. 1. Applying this termination mode, the moisture analy- sion of the obtained results. The reason for this can be
sis was finished when the mass of a sample aliquot changed by less illustrated as follows: if the readability of the balance is
than 0.2 mg/24 s during the drying procedure. The instrument re- 1 mg, the sample mass amounts to 200 mg, and the mois-
ports the calculated moisture value as well as the time needed for ture content of the sample is 4%, the absolute mass differ-
analysis on a LCD and the optional internal printer. ence will be 8 mg. Considering 3-times the readability of
Glass beakers were conditioned in a drying oven at 105 °C for
2 h and also only handled with a pair of tweezers. After heating, the balance (3 mg) the technical limitation of the balance
the beakers were allowed to cool down to room temperature in a alone contributes an error of about 40% to the final result.
desiccator containing silica gel. Two-hundred-mg aliquots of pow- Consequently, repetitive analysis of the moisture content
ders were weighed to 0.1 mg on an analytical balance and dried at of aliquots of the sample will yield a value of approxi-
105 °C until constant mass, typically for 2 h, before reweighing
and subsequent calculation of the moisture content from the mass mately 4.0±1.5%. Moisture analyzers which are conven-
difference. tionally designed for sample amounts of 5–15 g obviously
provide superior repeatability for such high sample masses.
Fortunately, the resolution of the balance of the moisture
Samples
analyzers employed in this study can be set to 0.1 mg.
Beside the prepared bulk materials, certified biological reference Therefore, the limitation of the precision of repetitive
materials were also analyzed for their moisture content using the measurements caused by the balance will be 0.3 mg, cor-
developed drying procedures. However, it should be noted that the responding to a moisture content of 4.0±0.15% in our ex-
reference materials are certified for specific elemental concentra- ample. This calculation clearly demonstrates that a resolu-
tions and not for their moisture content. The recommended drying
procedure of the supplier of the particular reference material was tion of the balance of 0.1 mg is absolutely necessary to
chosen as the reference drying method. The following reference obtain acceptable results when small sample masses and
materials were considered: (a) Lake Sediments, LKSD-4, Canada low moisture levels are expected. In addition to these tech-
Centre for Mineral and Energy Technology (CANMET), Ottawa, nical prerequisites, also the structure of the sample, atmos-
Ont., Canada; (b) SRM 1632b, Trace Elements in Coal, National
Institute of Standards & Technology (NIST), Gaithersburg, MD, pheric pressure, room temperature and humidity may sig-
USA; (c) SRM 1575 Pine Needles (NIST); (d) SRM 1547 Peach nificantly influence the moisture results [8].
Leaves (NIST); (e) BCR 281 Rye Grass, Institute for Reference All specimens investigated were sieved to fine pow-
Materials and Measurements (IRMM), Geel, Belgium; (f) OGS ders of < 100 µm because this particle size represents the
1878P, a peat reference material that has been made available by
the Geoscience Laboratories, Sudbury, Ont., Canada. consistency of most certified reference materials and nor-
mally assures a sufficiently homogenous material. An-
other reason for investigating matrices with a small parti-
cle size was the increased specific surface of such samples
 947

compared to coarse powdered material, thus enhancing ers at the recommended temperature settings of 80–100 °C
the moisture binding ability of the fine powders. yielded distinctly higher values compared to results ob-
As a reference method, drying of sample aliquots in tained after oven drying at 105 °C (Fig. 2). Therefore, op-
glass beakers in a conventional oven at 105 °C for 2 h or timal temperature settings of the electronic moisture ana-
until constant mass was chosen. This procedure has proved lyzers for all investigated matrices were sought. As can
to give accurate and precise results and is recommend for be clearly seen from Table 1, two temperatures yielding
drying many certified reference materials of various prove- optimal results were identified. Matrices such as soil and
nances before use by all major suppliers of reference ma- sewage sludge require a temperature setting of 45 °C,
terials, such as BCR and NIST [9, 10, 11, 12]. whereas for all other investigated specimens a tempera-
ture of 65 °C proved to give best results.
Before determination of the moisture content in a sam-
Optimization of the drying procedures ple, both electronic moisture analyzers need to be heated
to the selected temperature for about 10 min. Otherwise,
Initial experiments involving various matrices revealed the first 2–3 measurements underestimate the true mois-
that results obtained with the electronic moisture analyz- ture values. The behavior of the two kinds of heaters used
in the electronic moisture analyzers was quite different,
with the halogen heater giving higher results at tempera-
tures of 45 to 70 °C (Fig. 2b). This effect will be discussed
in detail later. Part of the observed discrepancies, i.e. the
higher results obtained with the electronic moisture analyz-
ers, can be attributed to the low sample mass (ca 200 mg)
compared to 1 to 15 g aliquots normally used with elec-
tronic moisture analyzers, as indicated in Fig. 3a. This
graph clearly shows that increasing the sample mass from
0.2 to 2 g results in about 10% lower moisture values at
the same temperature setting. Moreover, the time needed
until moisture analysis is finished increases almost expo-
nentially with increasing sample mass (Fig. 3a). The above
arguments underline the fact that the use of electronic
moisture analyzers represents a relative method that has to
calibrated against another well established drying proce-
dure. The experimental data displayed in Fig. 3a were ob-
tained when drying 200 mg aliquots of sewage sludge at
80 °C, which is at the low end of the temperature range of
80 to 100 °C recommended by the manufacturer. How-
ever, comparing these results with those from drying in
the heating oven using recommended drying procedures

Table 1 Comparison of the moisture content (%) in various envi-

ronmental matrices using the optimized procedures for the investi-
gated electronic moisture analyzers and conventional oven drying
Matrixa Drying oven Electronic moisture analyzer
2 h, 105 °C
Tempera- Halogen Ceramic
ture, °C
Soil 1.1±0.1 45 1.3±0.1 1.2±0.2
Sewage sludge 1.7±0.1 45 1.6±0.3 1.2±0.2
Lake sediment 1.3±0.1 65 1.5±0.3 1.3±0.3
Forest soil 1 4.7±0.1 65 4.6±0.2 4.9±0.6
Forest soil 2 7.0±0.2 65 6.9±0.2 7.5±0.2
Fig. 2 Dependence of the determined moisture content on the se-
lected temperature of the heater of the moisture analyzer. Matrix: Spruce needles 1 4.7±0.1 65 4.9±0.3 4.0±0.3
sewage sludge, sample mass ca 200 mg; hatched areas represent Spruce needles 2 5.2±0.1 65 5.4±0.4 5.9±0.2
the moisture content determined by drying at 105 °C for 2 h in a Leaves 1 5.0±0.2 65 5.2±0.2 6.0±0.2
conventional oven. Each point represents at least six measure- Leaves 2 5.3±0.2 65 5.2±0.2 6.4±0.3
ments. (a) Comparison of the behavior of halogen and ceramic
heater between 45 °C and 140 °C using the median. (b) Mean and aAll matrices were homogenized and sieved to <100 µm, at least

standard deviation of the results obtained with the two electronic six sample aliquots of 200 mg±10% each, weighed to 0.1 mg, were
moisture analyzers between 45 °C and 70 °C considered for measurements
948

Fig. 4 Typical profiles of loss of moisture during drying at 65 °C

with the halogen heated moisture analyzer for selected matrices,
i.e. bulk materials prepared for this study

be allowed to stabilize for several seconds before the mea-

surement starts automatically. However, this option is not
very helpful because a significant amount of moisture is
already given off in the first seconds of the drying process
and approximately 50% in the first 30 s of the measure-
ment (Fig. 4). Using no delay, the accuracy and precision
of the measurements could be substantially improved.
Another crucial point is the proper handling of the dis-
posable sample pans. The performance of the moisture an-
Fig. 3 Dependence of the moisture content and of the measure- alyzers can be distinctly improved by avoiding direct con-
ment time on the sample mass using the halogen heater. (a) Ma- tact with the sample pans (plastic gloves or tweezers are
trix: sewage sludge; temperature 80 °C. (b) Matrix: leaves; temper-
ature 65 °C; hatched area: results of oven drying
recommended). To remove residual moisture, it is essen-
tial to condition the sample pan prior to use by heating it
in the moisture analyzer without a sample. This procedure
becomes particularly important when the moisture content
[11], it is obvious that moisture values obtained with the of the sample is very low. The absolute error introduced
electronic moisture analyzer at 80 °C are too high. A sim- by residual moisture on the sample pan can be up to 0.3%.
ilar mass-dependent decrease in moisture results was also Moreover, re-used, not well cleaned sample pans nega-
established for other electronic moisture analyzers [8]. In tively influence the results of the next moisture analysis,
contrastthe contrary, when performing such an experiment leading to increased uncertainty. Although the manufac-
(moisture content vs. sample mass) with the leaves matrix turer recommends using the sample pans only once, they
at the optimized temperature setting of 65 °C – which is can be easily re-used when handled carefully. All measure-
valid for almost all matrices – the moisture results are no ments in this study were carried out with only four sample
longer mass-dependent (Fig. 3b). Consequently, no sys- pans. Additional use of new sample pans improved nei-
tematic decrease of the moisture content with increasing ther accuracy nor precision of the developed drying pro-
sample mass can be observed between sample masses of cedures. During cleaning of the sample pans by wiping
0.2 and 5 g, respectively. Only the standard deviations of with soft dry tissues, scratches and grooves have to be
the mean values of repetitive measurements improve with avoided. Otherwise the sample pans will provide additional
increasing sample mass, however, as expected. Moreover, surfaces on which the rising hot air produced during the
these results agree well with the moisture content deter- drying process can act, resulting in a more pronounced
mined after drying in a conventional oven at 105 °C. Sam- buoyancy leading to incorrect results.
ple masses should not be lower than 180 mg because then
precision and accuracy of the results worsen dramatically,
most probably due to pronounced buoyancy and due to Ceramic versus halogen heater
the small absolute changes in sample mass.
After closing the motorized lid containing the heater of In comparison to the ceramic heater, the halogen radiator
each of the electronic moisture analyzers, the balance can needs a shorter time to reach the pre-selected temperature.
 949

This results in shorter drying times of about 1 min. Due to erence materials, certified for concentrations of selected
the selected shut-off principle (change of sample mass trace elements, were investigated. Aliquots of all refer-
less than 0.2 mg/24 s) and the slow temperature increase ence materials (200 mg) were dried according to the man-
of the ceramic heater, it frequently happens that the maxi- ufacturer’s recommendations, when such recommenda-
mum pre-selected heating temperature is not achieved tions were available. Additionally, the reference materials
during the measurement cycle. On the contrary, once the were also dried at a given temperature for 2 h and until
ceramic heater has reached its working temperature, it constant mass to investigate the potential changes in the
needs several minutes to drop down again. The heater is humidity. Independently, the moisture in 200 mg sub-sam-
often even observed to overshoot the maximum pre-se- ples of the same reference materials was determined with
lected temperature by as much as 3 to 4° when a new mea- the electronic halogen dryer. The results are summarized
surement is started before the temperature arrives again at in Table 2.
the desired temperature within about one minute. As can Generally, moisture results agree quite well between
be seen from Fig. 2, the appropriate choice of temperature the conventional drying method and the developed proce-
is essential for obtaining correct results. Relatively small dures using the halogen moisture analyzer (Table 2). How-
variations of the drying temperature have a significant in- ever, following the recommended drying procedure for
fluence the result of moisture analysis. Use of the ceramic SRM 1575 Pine Needles (2 h at 85 °C, drying oven) led to
heater offers only insufficient temperature control and the slightly lower moisture values than when drying the mate-
halogen heater thus provides better results (Table 1). These rial at 85 °C until constant mass. Drying of this particular
problems can be partly overcome by activating the “standby reference material at 105 °C in an oven causes large losses
temperature” feature of the moisture analyzer. When the of volatiles other than water and should not be used [12].
standby temperature of the ceramic heater was set to 62 °C, However, this could be possible also for other materials.
results for matrices dried at 65 °C improved. However, the Therefore, prolonged drying, even at 85 °C may lead to
halogen heater provided faster and better results – even higher moisture values than expected. Results obtained
without activating the “standby temperature” option – and with the halogen moisture analyzer for SRM 1575 are be-
was therefore the only heater considered for further mea- tween the values achieved by oven drying at 85 °C for 2 h
surements. and until constant mass, respectively. Similar observa-
tions apply to LKSD-4 Lake Sediment. This reference
material should be analyzed as bottled and thus no in-
Application to certified reference materials structions for drying are available. Oven drying at 105 °C
for 2 h again yielded lower values than drying until con-
To demonstrate the applicability of the developed drying stant mass with the results of the electronic moisture ana-
procedures to commercially available matrices, several ref- lyzer lying in between (Table 2). Drying of the peat refer-

Table 2 Moisture content of 200 mg-aliquots of selected reference materials certified for elemental concentrations as determined by
conventional oven drying and the halogen heated electric moisture analyzer using the developed drying procedures
Matrixb Conventional drying procedurea Halogen moisture analyzerb

Moisture, % Moisture, % Time, min c

Peach Leaves, SRM 1547 Desiccator at room temperature for 3.4±0.1 3.7±0.2 3.4
120 h over magnesium perchlorate
Pine Needles, SRM 1575 85 °C, 2 h 4.1±0.1 4.3±0.3 4.3
85 °C, until constant mass 4.7±0.1
Rye grass, BCR 281 105 °C, until constant mass 5.0±0.1 4.7d 3.8
Lake Sediment LKSD-4, CCRMP 105 °C, 2 h 3.6±0.1 4.1±0.2 3.8
105 °C, until constant mass 4.5±0.1
Coal, SRM 1632b 105 °C, until constant mass 0.9±0.1 1.4±0.1 1.4
Peat, OGS 1878P e 85 °C, 2 h 7.2±0.4 7.8±0.4 6.3
85 °C, until constant mass 8.5±0.1
105 °C, 2 h 7.8±0.1
105 °C, until constant mass 8.5±0.1

All results are based on the analysis of at least 4 separate sample cAverage time needed for one measurement
aliquots dOnly one measurement was performed because the bottle with the
aRecommended drying procedures by the manufacturers for the reference material was finished
reference materials are given in italics eIn-house reference material, Geosciences Laboratories, Sudbury,
bThe temperature of the halogen heater was set to 65 °C except for Ont., Canada (not certified)
NIST 1632b Coal the moisture of which was determined at 45 °C
950

ence material, which has been characterized for trace ele- tinctly from each other. Needles, for example, should be
ments by an inter-laboratory comparison program in the oven dried at 85 °C for 2 h (explicit statement that a tem-
EUROPEAT project [13], is recommended at 105 °C until perature of 105 °C be avoided) [13] or should be dried at
constant mass. From the data given in Table 2 it can be 105 °C until constant mass [14]. These inconsistent recom-
clearly seen that the drying temperature and the drying mendations produce different analytical results and should
time distinctly influence the final result for the peat refer- therefore be reconsidered and harmonized on a global scale.
ence material. Interestingly, drying until constant mass at
85 °C and at 105 °C yielded exactly the same results. The Acknowledgments M.K. is grateful to W. Shotyk for linguistic
relatively high standard deviation and the slightly lower revision of this manuscript and to H. Emons for fruitful discus-
sions. Thanks also go to Sartorius, Göttingen, Germany for the
values obtained with the electronic moisture analyzer can loan of the two moisture analyzers MA100 with the halogen and
be ascribed to the automatic shutoff criterion selected for ceramic heater, respectively.
all measurements. As the peat reference material gives off
its moisture very slowly, sometimes the moisture analyzer
terminates the analysis before the material is completely References
dried. Accuracy and precision of moisture measurements
of peat could be thus improved by selecting an adequate 1. Dane AD, Rea GJ, Walmsley AD, Haswell SJ (2001) Anal
Chim Acta 429:185
heating period, rather than using the automatic shut-off fea- 2. Tanner PA, Leong LS (1997) Anal Chim Acta 342:247
ture of the moisture analyzer. This argument is further sub- 3. Rückold S, Grobecker KH, Isengard HD (2000) Fresenius J
stantiated by comparing the moisture values (7.5; 7.6; 8.1; Anal Chem 368:522
8.3 %) with the measurement times (5.4; 5.8; 6.6; 7.8 min) 4. Kramer GN, Muntau H, Maier E, Pauwels J (1998) Fresenius J
Anal Chem 360:299
obtained for 4 sample aliquots. 5. Krachler M, Shotyk W, Emons H (2001) Anal Chim Acta
432:307
6. Operating Instructions MA 100 Electronic Moisture Analyzer,
Conclusions Satorius, Göttingen, Germany
7. Korhammer S, Isengard HD, Langer N, Denzel E, Markert B,
Muntau H, Herzig R, Quevauviller P (2000) Fresenius J Anal
Electronic moisture analyzers represent an attractive alter- Chem 368:664
native to conventional oven drying for the determination 8. Ressing Th (1999) Materialfeuchtebestimmung in Biomatrizes
of moisture in powdered environmental specimens. Mois- sowie Ultraspurenanalytik von Platin und Rhodium in Umwelt-
ture results are available within a few minutes – depend- proben, Thesis, FH Aachen, Germany
9. National Institute of Standards and Technology (NIST), Cer-
ing on the matrix and the moisture content – with accu- tificate of Analysis for SRM 1632b, Trace elements in coal,
racy and precision comparable with those obtained by Gaithersburg, MD, USA, 1997
drying sample aliquots in an oven. As a relative method, 10. Community Bureau of Reference (BCR), Certificate of Analy-
drying procedures involving electronic moisture analyzers sis for CRM 281 Trace elements rye grass, Brussels, Belgium,
1988
have to be calibrated against another well established dry- 11. National Institute of Standards and Technology (NIST), Cer-
ing procedure. Because of its ability to reach the pre-se- tificate of Analysis for SRM 2782 Industrial Sludge, Gaithers-
lected heating temperature very fast, a halogen radiator burg, MD, USA, 1998
provides superior performance to a ceramic heater. 12. National Institute of Standards and Technology (NIST), Cer-
The developed drying procedures using the halogen tificate of Analysis for SRM 1575 Pine needles, Gaithersburg,
MD, USA, 1993
heater have been successfully applied to the moisture 13. Barbante C, Shotyk W, Biester H, Cheburkin A, Emons H,
analysis of environmentally relevant reference materials Farmer J, Hoffman E, Cortizas AM, Matschullat J, Norton S,
and can thus be easily adopted for routine daily work. Re- Roos F, Schweyer J, Steinnes E, lecture presented at the con-
liable quantification of the moisture content of reference ference Heavy Metals in the Environment, Silver Anniversary
Meeting, 6–10 August 2000, Ann Arbor, Michigan, USA
materials and unknown specimens will certainly aid in 14. United Nations Economic Commission for Europe–Convention
improving the quality of the analytical data derived from on Long-range Transboundary Air Pollution, Manual on meth-
those materials. ods and criteria for harmonized sampling, assessment, monitor-
In general, drying procedures recommended for similar ing, and analysis of the effects of air pollution on forests, Part
IV: Sampling and Analysis of Needles and Leaves, May 2000
or even identical matrices for subsequent trace element
analysis using conventional drying methods differ dis-