15

chapter

Moisture and Total

Solids Analysis
Lisa J. Mauer ( ) *

Department of Food Science,

Purdue University,
West Lafayette, IN 47907, USA
e-mail: mauerl@purdue.edu
Robert L. Bradley Jr
Department of Food Science,
University of Wisconsin,
Madison, WI 53706, USA
e-mail: rbradley@wisc.edu

15.1 Introduction 15.2.1 Overview
15.1.1 Importance of Moisture Assays 15.2.2 Oven Drying Methods
15.1.2 Water in Foods 15.2.3 Distillation Procedures
15.1.3 Sample Collection and Handling 15.2.4 Chemical Method: Karl Fischer
15.2 Moisture/ Water Content Titration

S. Nielsen (ed.), Food Analysis, Food Science Text Series, 257

DOI 10.1007/978-3-319-45776-5_15, © Springer International Publishing 2017
 15.2.5 Physical Methods 15.4.3 Automated Gravimetric Moisture
15.2.6 Comparison of Moisture Content Sorption Balance
Determination Methods 15.4.4 Phase Diagrams Containing aw,
15.3 Water Activity Moisture Content, and Tg
15.3.1 Overview Relationships
15.3.2 Importance of Water Activity 15.5 Summary
15.3.3 Water Activity Measurement 15.6 Study Questions
15.4 Moisture Sorption Isotherms 15.7 Practice Problems
15.4.1 Overview References
15.4.2 Isopiestic Desiccator Method
Chapter 15 • Moisture and Total Solids Analysis 259

15.1 INTRODUCTION 1 5 . 1 Importance of moisture content in the food


table industry
Moisture assays can be one of the most important
analyses performed on a food product and yet one of Moisture content is
the most difficult from which to obtain accurate and important for Food examples
precise data. Water molecules are small and ubiqui-
Preservation and Dehydrated vegetables, potatoes,
tous in the environment in which foods are produced,
stability and fruits
stored, and used. Moisture exchange between foods
Dried milks and infant formulas
and the environment can lead to under- or overesti-
Powdered eggs, coffees, and teas
mation from moisture assays, and water can be diffi-
cult to completely remove from foods. The first Spices and herbs
sections of this chapter describe various methods for Crispy fried or baked chips and
moisture content analysis – their principles, proce- crackers
dures, applications, cautions, advantages, and disad- Cotton candy
vantages. Water activity measurement also is Quality factors Jams and jellies, to inhibit sugar
described later in this chapter, since it parallels the crystallization
measurement of total moisture as an important stabil- Sugar syrups
ity and quality factor. Determining both the water Prepared cereals: conventional 4–8 %
content and the water activity of a food provides a H2O, puffed 7–8 % H2O
complete moisture analysis. With an understanding Convenience in Concentrated milks and fruit juices
of the techniques described, one can apply appropri- packaging or Liquid cane sugar (67 % solids) and
ate moisture analyses to a wide variety of food shipping liquid corn sweetener (80 % solids)
products. (usually Dehydrated products (which clump or
reduced become sticky and difficult to
moisture) package if moisture content is too
15.1.1 Importance of Moisture Assays
high)
One of the most fundamental and important ana- Meeting Cheddar cheese must be ≤39 % H2O
lytical procedures that can be performed on a food compositional Enriched flour must be ≤15 % H2O
product is an assay for the amount of moisture, standards and Pineapple juice must have soluble
referred to as the moisture or water content of standards of solids of ≥10.5°Brix (conditions
the food [1–4]. In this context, the words “water” identity specified)
and “moisture” are generally used interchange-
Glucose syrup must have ≥70 % total
ably. The dry matter that remains after moisture solids
removal is commonly referred to as total solids.
% H2O in processed meats is
This analytical value is of great economic impor- commonly specified
tance to a food manufacturer, and there are legal
Accurate
limits as to how much water must or can be pres-
computation of
ent in some foods. Some examples in which mois-
nutritional value
ture content is important to the food processor are
Expressing results
provided in Table 15.1.
of other
In addition to quantifying the amount of water in analytical
foods, it is also important to document the energy sta- determinations
tus of the water in the food by determining the water on a uniform
activity [5]. It is the water activity, more so than the basis
moisture content, that influences microbial growth,
physical properties, and chemical and enzymatic reac-
tions in foods. Additionally, it is differences in water
15.1.2 Water in Foods
activity, not moisture content, that drive moisture
migration between different food components (such as The amount, physical state, and location of water in
between a crust and a filling) or between a food and foods will affect the types of analyses best suited to a
the environment. Water molecules move from regions particular food product, the ease and rate of water
of high water activity to regions of low water activity removal, the time required for assay equilibration, and
until equilibrium water activity is reached. the sample handling.
260 L.J. Mauer and R.L. Bradley Jr

15.2
table
 Properties of water

Property Value
Molecular weight 18.0153 g
Melting point (1 atm) 0.000 °C
Boiling point (1 atm) 100.000 °C
Enthalpy of vaporization (1 atm) 40.647 kJ/mol
Glass transition temperature (1 atm) 136 K
Critical temperature 373.99 °C
Density (20 °C) 0.99821 g/cm3
Vapor pressure (20 °C) 2.3388 kPa
Heat capacity (20 °C) 4.1818 J/gK
1 5 . 1 Schematic of the water molecule and hydro-

Thermal conductivity (20 °C) 0.5984 W/mK
figure
gen bonding
Thermal diffusivity (20 °C) 1.4 × 10−7 m2/s
Dielectric constant (20 °C) 80.20
Water activity (20 °C) 1.000
Water activity (−20 °C) 0.82
Adapted from [5, 6, 38]

15.1.2.1 Structure of the Water Molecule

Water molecules are comprised of two hydrogen
atoms covalently bound to an oxygen atom in a dis-
15.1.2.3 W
 ater Interactions with Food
torted tetrahedral arrangement (Fig. 15.1) [6]. There is
Ingredients
Although water does not covalently bond with food
a slight partially negative charge on the oxygen atom
ingredients, it can be difficult to remove from foods.
and a slight partially positive charge on the hydrogen
Having some understanding of the water intermolecu-
atoms. Thus, water molecules are small and highly
lar interactions with food ingredients and the locations
polar, have two hydrogen bond donor sites (the hydro-
of water in foods can improve the application of mois-
gen atoms), and have two hydrogen bond acceptor
ture assays. The hydrogen bond donor and acceptor
sites (the two nonbonding electron pairs on the oxy-
sites on water molecules interact with food ingredients
gen atom). Hydrogen bonds are relatively weak attrac-
via hydrogen bonds, dipole-dipole interactions, ionic
tive interactions with short life spans (picoseconds),
attractions, and van der Waals forces [6]. Water mole-
readily breaking and reforming as water molecules are
cules tend to cluster around ions and charged groups, as
constantly moving, particularly when temperature or
well as other hydrogen bond donor and acceptor groups
humidity fluctuations occur [7, 8]. This poses chal-
such as hydroxyl groups (-OH), carbonyl oxygens (=O),
lenges for accurately quantifying the amount or activ-
and amine groups (-NH2), adopting a more ordered
ity of water in a sample since the target can be moving
structure in these hydration shell clusters than is found
from or escaping to the environment during sample
in bulk water [9]. To remove water from food, enough
handling.
energy must be applied to overcome these intermolecu-
lar interactions. There are also different mechanisms by
15.1.2.2 P
 hysical States and Properties
which water interacts with solids, including surface
of Water
interactions (adsorption), condensed water (capillary
The temperature, pressure, and extent of intermolecu-
condensation and deliquescence), and internalized
lar hydrogen bonding between water molecules result
water (absorption and crystal hydrate formation) [10].
in water existing in different physical states: as a solid
Water also can be physically entrapped in food matrices,
(ice crystals), liquid (water), and gas (water vapor,
such as in gel structures or dense dehydrated or fried
humidity). Determining the total amount of water as
products, making it more difficult to remove than sur-
well as the water activity in foods may involve analyz-
face water. It is nearly impossible to remove all water
ing water in all three of these states. Many direct mois-
from foods, particularly the internalized water for which
ture assays are based on the weight lost from
heat and mass transfer rates and the properties of the
evaporating all water from a sample, often by applica-
food material can complicate complete moisture egress.
tion of heat to overcome the enthalpy of vaporization.
Other properties of water may be exploited for the
indirect determination of moisture contents, including 15.1.3 Sample Collection and Handling
the dielectric constant, density, and freezing point. General procedures for sampling, sample handling
These and other properties of water are summarized and storage, and sample preparation are given in Chap.
in Table 15.2. 5. These procedures are perhaps the greatest potential
Chapter 15 • Moisture and Total Solids Analysis 261

source of error in any analysis. Precautions must be 15.3

taken to minimize inadvertent moisture losses or table
gains that occur during these steps. Obviously, any  Moisture content of selected foods
exposure of a sample to the open atmosphere should Approximate
be as short as possible. Any heating of a sample by fric- percent moisture
tion during grinding should be minimized. Headspace (wet weight
in the sample storage container should be minimal Food item basis)
because moisture transfer between the sample and the
Cereals, bread, and pasta
container environment will likely occur. It is critical to  Wheat flour, whole grain 10.3
control temperature fluctuations since moisture will  White bread, enriched (wheat flour) 13.4
migrate in a sample to the colder part. To control this  Corn flakes cereal 3.5
potential error, the entire sample should be removed  Crackers saltines 4.0
from the container and reblended quickly, and a new  Macaroni, dry, enriched 9.9
test portion should then be sampled [11, 12]. Dairy products
To illustrate the need for optimum efficiency and  Milk, reduced fat, fluid, 2 % 89.3
speed in weighing samples for analysis, Bradley and  Yogurt, plain, low fat 85.1
Vanderwarn [13] showed, using shredded Cheddar  Cottage cheese, low fat or 2 % milk fat 80.7
cheese (2–3 g in a 5.5 cm aluminum foil pan), that  Cheddar cheese 36.8
moisture loss within an analytical balance was a  Ice cream, vanilla 61.0
straight line function. The rate of loss was related to Fats and oils
the relative humidity (RH). At 50 % RH, it required  Margarine, regular, hard, corn, 15.7
hydrogenated
only 5 s to lose 0.01 % moisture. This time doubled
 Butter, with salt 15.9
at 70 % RH or 0.01 % moisture loss in 10 s. While
 Oil – soybean, salad, or cooking 0
one might expect a curvilinear loss, the moisture
Fruits and vegetables
loss was actually linear over a 5 min study interval.
 Watermelon, raw 91.5
Samples with lower aw than the environmental RH,  Oranges, raw, California navels 86.3
such as many powders and crispy fried products,  Apples, raw, with skin 85.6
may pick up moisture during handling leading to  Grapes, American type, raw 81.3
overestimation of moisture contents. These examples  Raisins 15.3
demonstrate the necessity of absolute control during  Cucumbers, with peel, raw 95.2
collection of samples through weighing, before dry-  Potatoes, microwaved, cooked in skin, 72.4
ing or other analysis. flesh and skin
 Snap beans, green, raw 90.3
Meat, poultry, and fish
 Beef, ground, raw, 95 % lean 73.3
15.2 MOISTURE / WATER CONTENT  Chicken, broilers and fryers, light meat, 68.6
meat and skin, raw
15.2.1 Overview  Finfish, flatfish (flounder and sole 79.1
species), raw
The moisture content of foods varies greatly, as  Egg, whole, raw, fresh 75.8
shown in Table 15.3 [3]. Water is a major constituent Nuts
of many food products. The approximate, expected  Walnuts, black, dried 4.6
moisture content of a food can affect the choice of the  Peanuts, all types, dry roasted with salt 1.6
method of measurement. It also can guide the ana-  Peanut butter, smooth style, with salt 1.8
lyst in determining the practical level of accuracy Sweeteners
required when measuring moisture content, relative  Sugar, granulated 0
to other food constituents. The method used for  Sugar, brown 1.3
determining moisture content may measure more or  Honey, strained or extracted 17.1
less of the water present. This is the reason for offi- From the US Department of Agriculture, Agricultural Research
cial methods with stated procedures [14–16]. Service (2016) USDA National Nutrient Database for
However, several official methods may exist for a Standard Reference. Release 28. Nutrient Data Laboratory
particular product. For example, the AOAC Home Page, http://ndb.nal.usda.gov [3]
International methods for cheese include Method
926.08, vacuum oven; 948.12, forced draft oven; The different types of moisture content assays can
977.11, microwave oven; and 969.19, distillation [14]. be generally categorized into direct and indirect meth-
Usually, the first method listed by AOAC ods. Direct methods for moisture content are often
International is preferred over others in any section. done by removing water, although the method for
262 L.J. Mauer and R.L. Bradley Jr

moisture removal may vary. Drying, distillation, and time and a particular temperature at which limited
extraction are commonly used for moisture removal decomposition might be a factor. One major problem
followed by weighing, volumetry, or titration to deter- exists in that the physical process must separate all the
mine moisture content. Indirect methods are based on moisture without decomposing any of the constituents
properties of the food that are related to the presence that could release water. For example, carbohydrates
of water, such as capacitance, specific gravity, density, decompose at elevated temperatures and release
refractive index, freezing point, and electromagnetic water to form dehydrated hydrocarbon compounds.
absorption. The moisture generated in this decomposition leads to
overestimation of the moisture content and is not the
moisture that we want to measure. Certain other chemi-
15.2.2 Oven Drying Methods
cal reactions (e.g., sucrose hydrolysis) can result in uti-
In oven drying methods, which are direct methods, lization of moisture, which would reduce the moisture
the sample is heated under specified conditions, and for measurement. The loss of volatile constituents,
the loss of weight is used to calculate the moisture con- such as acetic, propionic, and butyric acids, and alco-
tent of the sample. The amount of moisture deter- hols, esters, and aldehydes among flavor compounds,
mined is highly dependent on the type of oven used, can also lead to errors. While weight changes in oven
conditions within the oven, and the time and tempera- drying methods are assumed to be due to moisture loss,
ture of drying. Various oven methods are approved by weight gains also can occur due to oxidation of unsatu-
AOAC International for determining the amount of rated fatty acids and certain other compounds.
moisture in many food products. The methods are Nelson and Hulett [17] determined that moisture
simple, and many ovens allow for simultaneous anal- was retained in biological products to at least 365 °C,
ysis of large numbers of samples. The time required which is coincidentally near the critical temperature for
may be from ~1 h to over 24 h. water. Their data indicate that among the decomposi-
tion products at elevated temperatures were CO, CO2,
CH4, and H2O. These were not given off at any one par-
15.2.2.1 General Information ticular temperature but at all temperatures and at dif-
ferent rates at the respective temperature in question.
15.2.2.1.1 Removal of Moisture By plotting moisture liberated against tempera-
Water evaporates more quickly at higher temperatures. ture, curves were obtained that show the amount of
Any oven method used to evaporate moisture has as its moisture liberated at each temperature (Fig. 15.2).
foundation the fact that the boiling point of water is Distinct breaks were shown that indicated the temper-
100 °C; however, this considers only pure water at sea ature at which decomposition became measurable.
level. According to Raoult’s law, if 1 molecular weight
(1 mol) of a solute is dissolved in 1.0 L of water, the
boiling point would be raised by 0.512 °C. This boiling
point elevation continues throughout the drying pro-
cess as solute concentrations increase.
Moisture removal is sometimes best achieved in a
two-stage process. Liquid products (e.g., juices, milk)
are commonly pre-dried over a steam bath before dry-
ing in an oven. Products such as bread and field-dried
grain are often air-dried and then ground and oven
dried, with the moisture content calculated from mois-
ture loss at both air and oven drying steps. Particle
size, particle size distribution, sample sizes, hygro-
scopicity, and surface area during drying influence the
rate and efficiency of moisture removal.

15.2.2.1.2 D
 ecomposition of Other Food
Constituents
Moisture loss from a sample during analysis is a func- 1 5 . 2 Moisture content of several foods held at

figure
various temperatures in an oven. The
tion of time and temperature. Decomposition enters hyphenated line extrapolates data to 275 °C,
the picture when time is extended too much or temper- the true moisture content (Reprinted with
ature is too high. Thus, most methods for food mois- permission from [17]. Copyright 1920,
ture content analysis involve a compromise between American Chemical Society)
Chapter 15 • Moisture and Total Solids Analysis 263

None of these curves showed any break before 15.2.2.1.4 Types of Pans for Oven Drying
184 °C. Generally, proteins decompose at temperatures Methods
somewhat lower than required for starches and cellu- Pans used for moisture content determinations are
loses. Extrapolation of the flat portion of each curve to varied in shape and may or may not have a cover. The
250 °C gave a true moisture content based on the AOAC International [14] moisture pan is about 5.5 cm
assumption that there was no adsorbed water present in diameter with an insert cover. Other pans have cov-
at the temperature in question. ers that slip over the outside edge of the pan. These
pans, while reusable, are expensive, in terms of labor
15.2.2.1.3 Temperature Control costs to clean appropriately to allow reuse.
Drying methods utilize specified drying temperatures Pan covers are necessary to control loss of sample
and times, which must be carefully controlled. by spattering during the heating process. If the cover
Moreover, there may be considerable variability of is metal, it must be slipped to one side during drying
temperature, depending on the type of oven used for to allow for moisture evaporation. However, this slip-
moisture analysis. One should determine the extent of ping of the cover also creates an area where spattering
variation within an oven before relying on data col- will result in product loss. Examine the interior of
lected from its use. most moisture ovens and you will detect odor and
Consider the temperature variation in three deposits of burned-on residue, which, although unde-
types of ovens: convection (atmospheric), forced tected at the time of occurrence, produce erroneous
draft, and vacuum. The greatest temperature varia- results and large standard deviations [13].
tion exists in a convection oven. This is because hot Consider the use of disposable pans whenever
air slowly circulates without the aid of a fan. Air possible; then purchase glass fiber discs for covers. At
movement is obstructed further by pans placed in 5.5 cm in diameter, these covers fit perfectly inside dis-
the oven. When the oven door is closed, the rate of posable aluminum foil pans and prevent spattering
temperature recovery is generally slow. This is while allowing the surface to breathe. Paper filter discs
dependent also upon the load placed in the oven and foul with fat and thus do not breathe effectively. Drying
upon the ambient temperature. A 10 °C temperature studies done on cheese using various pans and covers
differential across a convection oven is not unusual. have shown that fat does spatter from pans with slipped
This must be considered in view of anticipated ana- covers and fiberglass is the most satisfactory cover.
lytical accuracy and precision. A convection oven
should not be used when precise and accurate mea- 15.2.2.1.5 Handling and Preparation of Pans
surements are needed. The preparation and handling of pans before use require
Forced draft ovens have the least temperature dif- consideration. Use only tongs to handle any pan or wear
ferential across the interior of all ovens, usually not gloves. Even the oils in fingerprints have weight. All pans
greater than 1 °C. Air is circulated by a fan that forces must be oven treated to prepare them for use. This is a
air movement throughout the oven cavity. Forced factor of major importance unless disproved by the tech-
draft ovens with air distribution manifolds have the nologist doing moisture determinations with a particular
added benefit of horizontal air movement across type of pan. Disposable aluminum pans must be vacuum
shelving. Thus, no matter whether the oven shelves oven dried for 3 h before use. At 3 and 15 h in either a
are filled completely with moisture pans or only half vacuum or forced draft oven at 100 °C, pans varied in
filled, the result would be the same for a particular their weight within the error of the balance or 0.0001 g,
sample [13]. and therefore longer drying times are not needed [13].
Two features of some vacuum ovens contribute to Store dried moisture pans in a functioning desiccator. The
a wider temperature spread across the oven. One fea- glass fiber covers should be dried for 1 h before use.
ture is a glass panel in the door. Although from an
educational point of view it may be fascinating to 15.2.2.1.6 Control of Surface Crust Formation
observe some samples in the drying mode, the glass is (Sand Pan Technique)
a heat sink. The second feature is the way by which air Some food materials tend to form a semipermeable
is bled into the oven. If the air inlet and discharge are crust or lump together during drying, which will
on opposite sides, conduct of air is virtually straight contribute to erratic and erroneous results. To con-
across the oven. Some newer models have air inlet trol this problem, analysts use the sand pan tech-
and discharge manifolds mounted top and bottom. nique. Clean, dry sand and a short glass stirring
Air movement in this style of vacuum oven is upward rod are pre-weighed into a moisture pan.
from the front and then backward to the discharge in Subsequently, after weighing in a sample, the sand
a broad sweep. The effect is to minimize cold spots as and sample are admixed with the stirring rod left
well as to exhaust moisture in the interior air. in the pan. The remainder of the procedure follows
264 L.J. Mauer and R.L. Bradley Jr

a standardized method if available; otherwise the placed in the oven for an arbitrarily selected time if no
sample is dried to constant weight. The purpose of standardized method exists. Drying time periods for
the sand is twofold: to prevent surface crust from this method are 0.75–24 h (Table 15.4), depending on
forming and to disperse the sample so evaporation the food sample and its pretreatment; some liquid
of moisture is less impeded. The amount of sand samples are dried initially on a steam bath at 100 °C to
used is a function of sample size. Consider 20–30 g minimize spattering. In these cases, drying times are
sand/3 g sample to obtain desired distribution in shortened to 0.75–3 h. A forced draft oven is used with
the pan. Similar to the procedure, applications, or without a steam table pre-drying treatment to deter-
and advantages of using sand, other heat-stable mine the solids content of fluid milks (AOAC Method
inert materials such as diatomaceous earth can be 990.19, 990.20).
used in moisture content determinations, espe- An alternative to selecting a time period for dry-
cially for sticky fruits. ing is to weigh and reweigh the dried sample and pan
The inert matrices such as sand and diatoma- until two successive weighings taken 30 min apart
ceous earth function to disperse the food constitu- agree within a specified limit, for example, 0.1–0.2 mg
ents and minimize retention of moisture in the food for a 5 g sample. The user of this second method must
products. However, the analyst must ascertain that be aware of sample transformation, such as browning
the inert matrix used does not give erroneous which suggests moisture loss of the wrong form. Lipid
results for the assay because of decomposition or oxidation and a resulting sample weight gain can
entrapped moisture loss. Test the sand or other occur at high temperatures in a forced draft oven.
inert matrix for weight loss before using in any Samples high in carbohydrates should not be dried in
method. Add approximately 25 g of sand into a a forced draft oven but rather in a vacuum oven at a
moisture pan and heat at 100 °C for 2 h and weigh temperature no higher than 70 °C.
to 0.1 mg. Add 5 mL water and mix with the matrix
using a glass rod. Heat dish, matrix, cover, and 15.2.2.3 Vacuum Oven
glass rod for at least 4 h at 100 °C and reweigh. The By drying under reduced pressure (25–100 mm Hg),
difference between weighing must be less than the rate of evaporation is faster, and one is able to
0.5 mg for any suitable matrix [18]. obtain a more complete removal of water without
decomposition within a 3–6 h drying time. Vacuum
15.2.2.1.7 Calculations
Moisture (wwb and dwb) and total solids contents of
foods can be calculated as follows using oven drying
procedures:
1 5 . 4 Forced draft oven temperature and times for

table selected foods
% Moisture ( wwb )
Dry on Oven
wt of wet sample - wt of dry sample steam temperature Time in
= ´ 1000
wt of wet sample Product bath (°C ± 2) oven (h)
wt H 2 O in sample Buttermilk, liquid Xa 100 3
= ´ 100
wt of wet sample (15.1) Cheese, natural type only 100 16.5 ± 0.5
Chocolate and cocoa 100 3
Cottage cheese 100 3
wt H 2 O in sample Cream, liquid and frozen X 100 3
% Moisture ( dwb ) = ´ 100 (15.2) Egg albumin, liquid X 130 0.75
wt of dry sample Egg albumin, dried X 100 0.75
Ice cream and frozen X 100 3.5
% Total solids ( wt / wt ) desserts
wt of dry sample Milk X 100 3
= ´ 100  Whole, low fat, and skim 100 3
wt of wet sample  Condensed skim 100 3
= 100 - % Moisture ( wwb ) (15.3) Nuts: almonds, peanuts, 130 3
walnuts
Fruit, dried 70 6
15.2.2.2 Forced Draft Oven Coffee, roasted 70 16
When using a forced draft oven, the sample is rapidly From Wehr and Frank [15] p. 492, with permission
weighed into a pre-dried moisture pan, covered, and a
X = samples must be partially dried on steam bath before
being placed in oven
Chapter 15 • Moisture and Total Solids Analysis 265

ovens need a dry air purge in addition to temperature composition of a product in-process helps food manu-
and vacuum controls to operate within method defini- facturers reduce production costs, meet regulatory
tion. In older methods, a vacuum flask is used, par- requirements, and ensure product consistency. Such
tially filled with concentrated sulfuric acid as the control could effectively pay for the microwave ana-
desiccant. One or two air bubbles per second are lyzer within a few months.
passed through the acid. Recent changes now stipulate A particular microwave moisture/solids analyzer
an air trap that is filled with calcium sulfate containing (CEM Corporation, Matthews, NC), or equivalent, is
an indicator to show water saturation (such as specified in the AOAC International procedures for
DrieRite™). Between the trap and the vacuum oven is total solids analysis of processed tomato products
an appropriately sized rotameter to measure air flow (AOAC Method 985.26) and moisture content analysis
(100–120 ml/min) into the oven. of meat and poultry products (AOAC Method 985.14).
The following are important points in the use of a The general procedure for use of a microwave
vacuum drying oven: moisture/solids analyzer has been to set the micro-
processor controller to a percentage of full power to
1. Temperature used depends on the product, control the microwave output. Power settings are
such as 95–102 °C for some foods and lower dependent upon the type of sample and the recommen-
temperatures (60–70 °C) for fruits and other dations of the manufacturer of the microwave moisture
high-sugar products. Even with reduced tem- analyzer. Next, the internal balance is tared with two
perature, there can be some decomposition. sample pads on the balance. As rapidly as possible, a
2. If the product to be assayed has a high concen- sample is placed between the two pads, and then pads
tration of volatiles, you should consider the use are centered on the pedestal and weighed against the
of a correction factor to compensate for the loss. tare weight. Time for the drying operation is set by the
3. Analysts should remember that in a vacuum, heat operator and “start” is activated. The microprocessor
is not conducted well. Thus pans must be placed controls the drying procedure, with percentage mois-
directly on the metal shelves to conduct heat. ture indicated in the controller window. Some newer
4. Evaporation is an endothermic process; thus, a models of microwave moisture analyzers have a tem-
pronounced cooling is observed. Because of the perature control feature to precisely control the drying
cooling effect of evaporation, when several process, removing the need to guess appropriate time
samples are placed in an oven of this type, you and power settings for specific applications. These new
will note that the temperature will drop. Do not models also have a smaller cavity that allows the micro-
attempt to compensate for the cooling effect by wave energy to be focused directly on the sample.
increasing the temperature; otherwise samples There are some considerations when using a
during the last stages of drying will be microwave analyzer for moisture determination: (1)
overheated. the sample must be of a uniform, appropriate size to
5. The drying time is a function of the total mois- provide for complete drying under the conditions
ture present, nature of the food, surface area per specified; (2) the sample must be centrally located and
unit weight of sample, whether sand is used as evenly distributed, so some portions are not burned
a dispersant, and the relative concentration of and other areas under-­processed; and (3) the amount
sugars and other substances capable of retain- of time used to place an appropriate sample weight
ing moisture or decomposing. The drying inter- between the pads must be minimized to prevent mois-
val is determined experimentally to give ture loss or gain before weight determination. Sample
reproducible results. pads also should be considered. There are several dif-
ferent types, including fiberglass and quartz fiber
15.2.2.4 Microwave Analyzer pads. For optimum results, the pads should not absorb
Determination of moisture contents in food products microwave energy, as this can cause the sample to
has traditionally been done using a standard oven, burn, nor should they fray easily, as this causes them
which, though accurate, can take many hours to dry a to lose weight and can affect the analysis. In addition,
sample. Microwave moisture analysis, often called they should absorb liquids well.
microwave drying, was the first precise and rapid Another style of microwave oven that includes a
technique that allowed some segments of the food vacuum system is used in some food plants. This vac-
industry to make in-process adjustment of the mois- uum microwave oven will accommodate one sample
ture content in food products before final packaging. in triplicate or three different samples at one time. In
For example, processed cheese could be analyzed, and 10 min, the results are reported to be similar to 5 h in a
the composition adjusted before the blend was vacuum oven at 100 °C. The vacuum microwave oven
dumped from the cooker. The ability to adjust the
266 L.J. Mauer and R.L. Bradley Jr

is not nearly as widely used as conventional microwave and laboratory assays with results comparable to ref-
analyzers, but can be beneficial in some applications. erence methods.
Microwave drying provides a fast (4–8 min),
accurate method to analyze many foods for mois- 15.2.2.7 Thermogravimetric Analyzer
ture content. The method is sufficiently accurate for In a thermogravimetric analyzer (TGA), the mass of a
routine assay. The distinct advantage of rapid analy- sample is continuously measured as it is heated at a
sis far outweighs its limitation of testing only single controlled rate in a controlled atmosphere. The sample
samples [19]. (often 10–50 mg) is loaded into a pan that is then
placed into the TGA instrument that contains a fur-
nace and a precision balance. Foods can lose mass dur-
15.2.2.5 Infrared Drying ing heating through release of adsorbed compounds
Infrared drying involves penetration of heat into the (such as water), chemical reactions, and decomposi-
sample being dried, as compared to heat conductivity tion. To determine moisture content, the sample cham-
and convection with conventional ovens. Such heat ber is purged with an inert gas (e.g., nitrogen) so the
penetration to evaporate moisture from the sample can sample only reacts with temperature during heating
significantly shorten the required drying time to and mass changes due to oxidation are minimized.
10–25 min. The infrared lamp used to supply heat to the The furnaces in TGA instruments cover a wider tem-
sample results in a filament temperature of 2000– perature range than many other moisture content
2500 K. Factors that must be controlled include distance determination techniques and may range from sub-­
of the infrared source from the dried material and thick- ambient temperatures to >1500 °C. This enables not
ness of the sample. The analyst must be careful that the only moisture assessment (moisture content, tempera-
sample does not burn or case harden while drying. ture of dehydration, stoichiometry of a hydrate, and
Infrared drying ovens may be equipped with forced dehydration kinetics) but also at higher temperatures
ventilation to remove moist air and an analytical bal- measurement of pyrolysis, decomposition, and weight
ance to read moisture content directly. No infrared dry- % ash. More detailed information on TGA is provided
ing moisture analysis techniques are approved by in the chapter on thermal analysis (Chap. 30). To use
AOAC International currently. However, because of the TGA, a survey scan is often run on a sample from
speed of analysis, this technique is suited for qualitative approximately 100 °C below to 100 °C above the tran-
in-process use. sition of interest at a heating rate of 20 °C per minute.
For water content determination of many foods, scans
from ambient to 200 °C are generally sufficient, and
15.2.2.6 R
 apid Moisture Analyzer water content is determined by the mass lost up to
Technology ~100 °C. A derivative plot of the rate of mass loss will
Many rapid moisture/solids analyzers based on generate a peak showing the onset, midpoint, and
thermogravimetric principles are used by the food endpoint temperatures for water loss. The precision
industry. In addition to those based on infrared balance and in situ mass monitoring result in precise
and microwave drying as described previously, and accurate moisture content determinations; how-
compact instruments that depend on high heat cre- ever, decomposition or volatile loss at temperatures
ated by different types of heaters are available. Two overlapping water evaporation can lead to errors in
main categories of heating elements include halogen moisture content determination.
heaters (e.g., Halogen Moisture Analyzers, Mettler
Toledo, Columbus, OH) and ceramic heaters (e.g.,
Computrac®, Arizona Instrument LLC, Chandler, 15.2.3 Distillation Procedures
AZ). These analyzers detect moisture levels from
50 ppm to 100 % using sample weights of 150 mg to 15.2.3.1 Overview
54 g. Smaller samples tend to dry more quickly, but it Distillation is used as another direct measure of mois-
is important to use enough sample to be representa- ture content. Distillation techniques involve codistill-
tive of the product. The test sample is placed on an ing the moisture in a food sample with a high boiling
aluminum pan or filter paper, and the heat control point solvent that is immiscible in water, collecting the
program (with a heating range of 25 °C to >200 °C) mixture that distills off, and then measuring the vol-
elevates the test sample to a constant temperature. ume of water. Two distillation procedures are in use
As the moisture is driven from the sample, the today: direct and reflux distillations, with a variety
instrument automatically weighs and calculates the of solvents. For example, in direct distillation with
percentage moisture or solids. The samples are not immiscible solvents of higher boiling point than water,
removed from the oven which minimizes weighing the sample is heated in mineral oil or liquid with a flash
errors, and accurate results are obtained within min- point well above the boiling point for water. Other
utes. These analyzers are utilized for both production immiscible liquids with boiling point only slightly
Chapter 15 • Moisture and Total Solids Analysis 267

above water can be used (e.g., toluene, xylene, and

benzene). However, reflux distillation with the immis-
cible solvent toluene is the most widely used method.
Distillation techniques were originally developed
as rapid methods for quality control work, but they are
not adaptable to routine testing. The distillation
method is an AOAC-approved technique for moisture
content analysis of spices (AOAC Method 986.21),
cheese (AOAC Method 969.19), and animal feeds
(AOAC Method 925.04). It also can give good accuracy
and precision for nuts, oils, soaps, and waxes.
Distillation methods cause less thermal decom-
position of some foods than oven drying at high
temperatures. Adverse chemical reactions are not
eliminated but can be minimized by using a solvent
with a lower boiling point. This, however, will increase
distillation times. Water is measured directly in the
distillation procedure (rather than by weight loss), but
reading the volume of water in a receiving tube may
be less accurate than using a weight measurement.

15.2.3.2  eflux Distillation with Immiscible

R
Solvent
Reflux distillation uses either a solvent less dense than
water (e.g., toluene, with a boiling point of 110.6 °C, or
xylene, with a boiling range of 137–140 °C) or a solvent 1 5 . 3 Apparatus for reflux distillation of moisture

from a food. Key to this setup is the Bidwell-
more dense than water (e.g., tetrachlorethylene, with a figure
Sterling moisture trap. This style can be used
boiling point of 121 °C). The advantage of using this only where the solvent is less dense than water
last solvent is that the material to be dried floats; there-
fore it will not char or burn. In addition, there is no fire 15.2.4 C
 hemical Method: Karl Fischer
hazard with this solvent. Titration
A Bidwell-Sterling moisture trap (Fig. 15.3) is
commonly used as part of the apparatus for reflux dis- The Karl Fischer titration, a direct measure of
tillation with a solvent less dense than water. The distil- moisture content, is particularly adaptable to food
lation procedure using such a trap requires about 1 h products that show erratic results when heated or
and involves using a brush to dislodge adhering water submitted to a vacuum. This is the method of choice
drops from the glassware, thereby minimizing error. for determination of water content in many low-
Three potential sources of error with distillation moisture foods such as dried fruits and vegetables
should be eliminated if observed: (AOAC Method 967.19 E-G), candies, chocolate
(AOAC Method 977.10), roasted coffee, oils and fats
1. Formation of water-solvent emulsions that will (AOAC Method 984.20), or any low-moisture food
not break. Usually this can be controlled by high in sugar or protein. The method is quite rapid
allowing the apparatus to cool after distillation and accurate and uses no heat. This method is based
is completed and before reading the amount of on the fundamental reaction described by Bunsen in
moisture in the trap. 1853 [20] involving the reduction of iodine by SO2 in
2. Clinging of water droplets to dirty apparatus. the presence of water:
Clean glassware is essential, but even with this,
a burette brush is needed to dislodge water 2H 2 O + SO 2 + I 2 ® H 2SO 4 + 2HI (15.4)
droplets. This was modified to include methanol and pyridine
3. Decomposition of the sample with production in a four-component system to dissolve the iodine and
of water. If this is a measurable problem, dis- SO2:
continue method use and find an alternative
procedure. C 5H 5 N • I 2 + C 5H 5 N • SO 2 + C 5H 5 N + H 2 O
® 2C 5H 5 N • HI + C 5H 5 N • SO 3 (15.5)

C 5H 5 N • SO 3 + CH 3 OH
® C 5H 5 N ( H ) SO 4 • CH 3 (15.6)
268 L.J. Mauer and R.L. Bradley Jr

These reactions show that for each mole of water,

1 mol of iodine, 1 mol of SO2, 3 mol of pyridine, and
1 mol of methanol are used. For general work, a
methanolic solution is used that contains these com-
ponents in the ratio of 1 iodine:3 SO2:10 pyridine, and
at a concentration so that 3.5 mg of water = 1 ml of
reagent. A procedure for standardizing this reagent
is given below.
In a volumetric titration procedure (Fig. 15.4 is
manual titration unit; Fig. 15.5 is an example of auto-
mated titration unit), iodine and SO2 in the appropri-
ate form are added to the sample in a closed chamber
protected from atmospheric moisture. The excess of I2
that cannot react with the water can be determined
visually. The endpoint color is dark red-brown. Some
instrumental systems are improved by the inclusion of
a potentiometer (i.e., conductometric method) to elec-
tronically determine the endpoint, which increases the
sensitivity and accuracy. The automated volumetric
titration, units (used for 100 ppm water to very high
concentrations) use a pump for mechanical addition of
titrant, and use the conductometric method for end-
point, determination (i.e., detection of excess iodine is
by applying a current and measuring the potential). 1 5 . 5 Automated Karl Fischer volumetric titration

The volumetric titration procedure described figure
unit (Courtesy of Mettler Toledo, Columbus,
above is appropriate for samples with a moisture con- OH)
tent greater than ~0.03 %. A second type of titration,
referred to as coulometric titration, is ideal for prod-
ucts with very low levels of moisture, from 0.03 %
down to parts per million (ppm) levels. In this method,
iodine is electrolytically generated (2I → I2 + 2e−) to
titrate the water. The amount of iodine required to
titrate the water is determined by the current needed
to generate the iodine. Just like for volumetric titra-
tion, automated coulometric titration units are avail-
able commercially.
In a Karl Fischer volumetric titration, the Karl
Fischer reagent (KFR) is added directly as the titrant if
the moisture in the sample is accessible. However, if
moisture in a solid sample is inaccessible to the
reagent, the moisture is extracted from the food with
an appropriate solvent (e.g., methanol). (Particle size
affects efficiency of extraction directly.) Then the meth-
anol extract is titrated with KFR.
The noxious odor of pyridine makes it an unde-
sirable reagent. Therefore, researchers have experi-
mented with other amines capable of dissolving
iodine and sulfur dioxide. Some aliphatic amines and
several other heterocyclic compounds were found
suitable. On the basis of these amines, one-compo-
nent reagents (solvent and titrant components
together) and two-component reagents (solvent and
titrant components separate) have been prepared. The
one-component reagent may be more convenient to
1 5 . 4 Manual Karl Fischer titration unit
 use, but the two-component reagent has greater stor-
figure
(Courtesy of Lab Industries, Inc., Berkeley, age stability.
CA)
Chapter 15 • Moisture and Total Solids Analysis 269

Before the amount of water found in a food sam- 3. Moisture adhering to walls of unit. All glass-
ple can be determined, a KFR water (moisture) equiv- ware and utensils must be carefully dried.
alence (KFReq) must be determined. The KFReq value 4. Interferences from certain food constituents.
represents the equivalent amount of water that reacts Ascorbic acid is oxidized by KFR to dehydro-
with 1 ml of KFR. Standardization must be checked ascorbic acid to overestimate moisture content;
before each use because the KFReq will change with carbonyl compounds react with methanol to
time. form acetals and release water, to overestimate
The KFReq can be established with pure water, a moisture content (this reaction also may result in
water-in-methanol standard, or sodium tartrate fading endpoints); unsaturated fatty acids will
dihydrate. Pure water is a difficult standard to use react with iodine, so moisture content will be
because of inaccuracy in measuring the small overestimated.
amounts required. The water-in-methanol standard
is premixed by the manufacturer and generally con- 15.2.5 Physical Methods
tains 1 mg of water/mL of solution. This standard
Most physical methods are indirect measures of mois-
can change over prolonged storage periods by
ture content and do not separate the water from the
absorbing atmospheric moisture. Sodium tartrate
sample for analysis. These techniques can be rapid
dihydrate (Na2C4H4O6 · 2H2O) is a primary standard
and nondestructive, which has led to their wide-
for determining KFReq. This compound is very sta-
spread use in food production and quality control;
ble, contains 15.66 % water under all conditions
however, they must be calibrated against data col-
expected in the laboratory, and is the material of
lected by a direct method to quantify the amount of
choice to use.
water in samples.
The KFReq is calculated as follows using sodium
tartrate dihydrate:
15.2.5.1 Dielectric Method
KFReq ( mg H 2 O / mL ) The electrical properties of water are used in the
dielectric method to determine the moisture content
36 g H 2 O / mol Na 2 C 4 H 4 O 6 × 2H 2 O ´ S ´ 1000
= of certain foods, by measuring the change in capaci-
230.08 g / mol ´ A (15.7) tance or resistance to an electric current passed
where: through a sample. These instruments require calibra-
tion against samples of known moisture content as
KFReq = Karl Fischer reagent water (moisture) determined by standard direct methods. Sample den-
equivalence sity or weight/volume relationships and sample tem-
S=w eight of sodium tartrate dihydrate (g) perature are important factors to control in making
A = mL of KFR required for titration of sodium reliable and repeatable measurements by dielectric
tartrate dihydrate methods. These techniques can be very useful for pro-
cess control measurement applications, where con-
Once the KFReq is known, the moisture content of tinuous measurement is required. These methods are
the sample is determined as follows: limited to food systems that contain no more than
30–35 % moisture.
KFReq ´ Ks The moisture content determination in dielec-
% H2O = ´ 100 (15.8)
S tric-type meters is based on the fact that the dielec-
where: tric constant of water (80.37 at 20 °C) is higher than
that of most solvents. The dielectric constant is
KFReq = Karl Fischer reagent water (moisture) measured as an index of capacitance. As an exam-
equivalence ple, the dielectric method is used widely for cereal
Ks = mL of KFR used to titrate sample grains. Its use is based on the fact that water has a
S = weight of sample (mg) dielectric constant of 80.37, whereas starches and
proteins found in cereals have dielectric constants of
10. By determining this property on samples in stan-
The major difficulties and sources of error in the
dard metal condensers, dial readings may be
Karl Fischer titration methods are as follows:
obtained and the percentage of moisture determined
1. Incomplete moisture extraction. For this rea- from a previously constructed standard curve for a
son, fineness of grind (i.e., particle size) is particular cereal grain.
important in preparation of cereal grains and
some foods. 15.2.5.2 Hydrometry
2. Atmospheric moisture. External air must not Hydrometry is the science of measuring specific grav-
be allowed to infiltrate the reaction chamber. ity or density, which can be done using several differ-
270 L.J. Mauer and R.L. Bradley Jr

ent principles and instruments. While hydrometry is racy of specific gravity measurements can be improved
considered archaic in some analytical circles, it is still by using a hydrometer calibrated in the desired range
widely used and, with proper technique, is highly of specific gravities.
accurate. Specific gravity measurements with various The rudimentary but surprisingly accurate
types of hydrometers or with a pycnometer are com- hydrometer comes equipped with various modifica-
monly used for routine testing of moisture (or solids) tions depending on the fluid to be measured:
content of numerous food products. These include
beverages, salt brines, and sugar solutions. Specific 1. The Quevenne and New York Board of Health
gravity measurements are best applied to the analysis lactometer is used to determine the density of
of solutions consisting of only one solute in a medium milk. The Quevenne lactometer reads from 15
of water. to 40 lactometer units and corresponds to 1.015–
1.040 specific gravity. For every degree above
15.2.5.2.1 Hydrometer 60 °F, 0.1 lactometer unit is added to the read-
One approach to measuring specific gravity is based ing, and 0.1 lactometer unit is subtracted for
on Archimedes’ principle, which states that a solid every degree below 60 °F.
suspended in a liquid will be buoyed by a force equal 2. The Baumé hydrometer was used originally to
to the weight of the liquid displaced. The weight per determine the density of salt solutions (origi-
unit volume of a liquid is determined by measuring nally 10 % salt), but it has come into much wider
the volume displaced by an object of standard weight. use. From the value obtained in the Baumé
A hydrometer is a standard weight on the end of a scale, you can convert to specific gravity of
spindle, and it displaces a weight of liquid equal to its liquids heavier than water. For example, it is
own weight (Fig. 15.6). For example, in a liquid of low used to determine the specific gravity of milk
density, the hydrometer will sink to a greater depth, being condensed in a vacuum pan.
whereas in a liquid of high density, the hydrometer 3. The Brix hydrometer is a type of saccharometer
will not sink as far. Hydrometers are available in nar- used for sugar solutions such as fruit juices and
row and wide ranges of specific gravity. The spindle of syrups, and one usually reads directly the per-
the hydrometer is calibrated to read specific gravity centage of sucrose at 20 °C. Balling saccharom-
directly at 15.5 °C or 20 °C. A hydrometer is not as eters are graduated to indicate percentage of
accurate as a pycnometer, but the speed with which sugar by weight at 60 °F. The terms Brix and
you can do an analysis is a decisive factor. The accu- Balling are interpreted as the weight percent-
age of pure sucrose.
4. Alcoholometers are used to estimate the alco-
hol content of beverages. Such hydrometers are
calibrated in 0.1° or 0.2° proof to determine the
percentage of alcohol in distilled liquors (AOAC
Method 957.03).
5. The Twaddell hydrometer is only for liquids
heavier than water.

15.2.5.2.2 Pycnometer
Another approach to measuring specific gravity is a
comparison of the weights of equal volumes of a liq-
uid and water in standardized glassware, a pycnome-
ter (Fig. 15.7). This will yield density of the liquid
compared to water. In some texts and reference books,
20/20 is given after the specific gravity number. This
indicates that the temperature of both fluids was 20 °C
when the weights were measured. Using a clean, dry
pycnometer at 20 °C, the analyst weighs it empty, fills
it to the full point with distilled water at 20 °C, inserts
the thermometer to seal the fill opening, and then
touches off the last drops of water and puts on the cap
for the overflow tube. The pycnometer is wiped dry in
case of any spillage from filling and is reweighed. The
1 5 . 6 Hydrometers (Courtesy of Cole-Parmer

density of the sample is calculated as follows:
figure
Instrument Company, Vernon Hills, IL)
Chapter 15 • Moisture and Total Solids Analysis 271

installed in a liquid processing line to monitor the

°Brix of products such as carbonated soft drinks, dis-
solved solids in orange juice, and the percentage of
solids in milk [21].
When a beam of light is passed from one medium
to another and the density of the two differs, then the
beam of light is bent or refracted. Bending of the
light beam is a function of the media and the sines of
the angles of incidence and refraction at any given
temperature and pressure, and is thus a constant
(Fig. 15.8). The (RI) (η) is a ratio of the sines of the
angles:
sine incident ray angle
h= (15.10)
sine refracted ray angle
15.7
Pycnometer
All chemical compounds have an index of refrac-
figure
tion. Therefore, this measurement can be used for the
qualitative identification of an unknown compound
by comparing its RI with literature values. RI varies
weight of sample - filled pycnometer with concentration of the compound, temperature,
and wavelength of light. Instruments are designed
- weight of empty pycnometer
to give a reading by passing a light beam of a spe-
weight of water - filled pycnometer cific wavelength through a glass prism into a liquid,
- weight of empty pycnometer the sample. Benchtop or handheld units use Amici
prisms to obtain the D line of the sodium spectrum or
= density of sample (15.9) 589 nm from white light. Whenever refractive indices
of standard fluids are given, these are prefaced with
This method is used for determining alcohol con-
hD20 = a value from 1.3000 to 1.7000 . The Greek letter η
tent in alcoholic beverages (e.g., distilled liquor, AOAC
is the symbol for RI; the 20 refers to temperature in °C;
Method 930.17), solids in sugar syrups (AOAC Method
and D is the wavelength of the light beam, the D line
932.14B), and solids in milk (AOAC Method 925.22).
of the sodium spectrum.
Benchtop instruments are more accurate com-
15.2.5.3 Refractometry
pared to handheld units mainly because of tempera-
Moisture in liquid sugar products and condensed
milks can be determined using a Baumé hydrometer
(solids), a Brix hydrometer (sugar content), gravimet-
ric means, or a refractometer. If it is performed cor-
rectly and no crystalline solids are evident, the
refractometer procedure is rapid and surprisingly
accurate (AOAC Method 932.14C, for solids in syr-
ups). The refractometer has been valuable in deter-
mining the soluble solids in fruits and fruit products
(AOAC Method 932.12; 976.20; 983.17).
The refractive index (RI) of an oil, syrup, or other
liquid is a dimensionless constant that can be used to
describe the nature of the food. While some refractom-
eters are designed only to provide results as refractive
indices, others, particularly handheld, quick-to-use
units, are equipped with scales calibrated to read the
percentage of solids, percentage of sugars, and the
like, depending on the products for which they are
intended. Tables are provided with the instruments to
convert values and adjust for temperature differences.
Refractometers are used not just on the laboratory 1 5 . 8 Reflection and refraction concepts of

figure
refractometry
bench or as handheld units. Refractometers can be
272 L.J. Mauer and R.L. Bradley Jr

15.2.5.4 Infrared Analysis

Infrared spectroscopy (see Chap. 8) has attained a
primary position in monitoring the composition of
food products before, during, and following process-
ing [22]. It has a wide range of food applications and
has proven successful in the laboratory and online.
Similar to the use of ultraviolet (UV) or visible (Vis)
light in UV-Vis spectroscopy, in infrared spectroscopy
a sample is exposed to IR radiation (near IR 700–
2400 nm or mid IR 2500–25,000 nm), specific wave-
lengths are absorbed, and an IR spectrum is measured
by calculating the intensity of the IR radiation before
and after passing through the sample. The absorbance
peaks are related to the type and amount of functional
groups present. However, infrared spectrometers
must be calibrated for each analyte to be measured,
and the analyte must be uniformly distributed in the
sample.
For water, near-infrared (NIR) bands (1400–1450;
1920–1950 nm) are characteristic of the -OH stretch of
the water molecule and can be used to determine the
moisture content of a food. NIR has been applied to
moisture content analysis of a wide variety of food
commodities and is an official method for moisture
content determination in dried vegetables (AOAC
Method 967.19).
The use of mid-infrared milk analyzers to deter-
mine fat, protein, lactose, and total solids in milk
1 5 . 9 Rhino Brix handheld refractometer, R2 mini

(AOAC Method 972.16) is covered in Chap. 8 of this
figure
digital handheld refractometer, and Mark III
Abbe refractometer (Courtesy of Reichert text. The instrument must be calibrated using a mini-
Analytical Instruments, Depew, NY) mum of eight milk samples that were previously ana-
lyzed for fat (F), protein (P), lactose (L), and total solids
(TS) by standard methods. Then, a mean difference
value, a, is calculated for all samples used in
ture control (Fig. 15.9). These former units have
calibration:
provisions for water circulation through the head
where the prism and sample meet. Digital and Abbe a = å ( TS - F - P - L ) / n (15.11)
refractometers are common for laboratory use. Care
must be taken when cleaning the prism surface fol- where:
lowing use. The contact surface should be wiped
clean with lens paper and rinsed with distilled water a = solids not measurable by the F, P, and L
and then ethanol. The prism chamber should be methods
closed and the instrument covered with a bag when n = number of samples
not in use to protect the delicate prism surface from F = fat percentage
dust or other debris that might lead to scratches and P = protein percentage
inaccuracy. L = lactose percentage
The fact that the RI of a solution increases with TS = total solids percentage
concentration has been exploited in the analysis of
total soluble solids of carbohydrate-based foods such Total solids then can be determined from any
as sugar syrups, fruit products, and tomato products. infrared milk analyzer results by using the formula
Because of this use, these refractometers are calibrated
TS = a + F + P + L (15.12)
in °Brix (g of sucrose/100 g of sample), which is equiva-
lent to percentage sucrose on a wt/wt basis. Refractive The a value is thus a standard value mathematically
index measurements are used widely to approximate derived. Newer instruments have the algorithm in
sugar concentration in foods, even though values are their computer software to ascertain this value auto-
accurate only for pure sucrose solutions. matically. Although not directly measured, the mois-
Chapter 15 • Moisture and Total Solids Analysis 273

ture content can be calculated by subtracting the total cise as possible. The thermistor used can sense tem-
solids content from 100. perature change to 0.001 °C. The general technique is
to supercool the solution and then induce crystalliza-
15.2.5.5 Microwave Absorption tion by a vibrating reed. The temperature will rise
The absorption of microwaves, which are electromag- rapidly to the freezing point or eutectic temperature
netic waves with wavelength of 0.001–1 m and fre- as the water freezes. In the case of pure water, the
quency of 0.3–300 GHz, can be used to determine the temperature remains constant until all the water is
moisture contents of a wide variety of food products. frozen. In the case of milk, the temperature is read
The approach is based on dielectric constant or per- when there is no further temperature rise. Time
mittivity value differences between water and dry required for the automated instruments is 1–2 min
materials. The permittivity of most dry materials is per prechilled sample.
much lower than that of water, and small changes in
the amount of water in a sample lead to measurable
15.2.6 C
 omparison of Moisture Content
changes in permittivity. As microwave energy is
Determination Methods
passed through a sample (<40 mm thick) placed
between a microwave transmitter and receiver, the 15.2.6.1 Principles
absorption at 2.450 GHz (the most widely used fre- Characteristics of the various moisture content analy-
quency in microwaves) is linearly related to moisture sis methods described in Sects. 15.2.2, 15.2.3, 15.2.4,
content [23]. After calibration against laboratory stan- and 15.2.5 are summarized in Table 15.5. Direct mois-
dards, moisture contents can be determined in ~2 s ture content determination methods often remove
using the microwave absorption method. Power water from a sample and determine moisture contents
absorption and attenuation may be affected by the by mass, volumetry, or titration. Oven drying methods
dimensions, surface area, temperature, and dielectric involve the removal of moisture from the sample and
properties of some food ingredients; however, the then a weight determination of the solids remaining to
technique is rapid, nondestructive, and useful for calculate the moisture content (and consequently also
some heterogeneous, powdered, multilayered, and the total solids content). Non-­water volatiles can be
frozen foods [4, 24]. lost during drying, but their loss is generally a negli-
gible percentage of the amount of water lost.
15.2.5.6 Freezing Point Distillation procedures also involve a separation of the
When water is added to a food product, many of the moisture from the solids, and the moisture is quanti-
physical constants are altered. Some properties of tated directly by volume. Karl Fischer titration is based
solutions depend on the number of solute particles as on chemical reactions of the moisture present, reflected
ions or molecules present. These properties are vapor as the amount of titrant used.
pressure, freezing point, boiling point, and osmotic Indirect moisture content determination methods
pressure. Measurement of any of these properties can analyze some property of the food that is related to the
be used to determine the concentration of solutes in a presence of water. The dielectric method is based on
solution. However, the most commonly practiced electrical properties of water. Hydrometric methods
assay for milk is the change of the freezing point value. are based on the relationship between specific gravity
It has economic importance with regard to both raw and moisture content. The refractive index method is
and pasteurized milk. The freezing point of milk is its based on how water in a sample affects the refraction of
most constant physical property. While termed a phys- light. Near-­infrared analysis of water in foods is based
ical constant, the freezing point varies within narrow on measuring the absorption at wavelengths character-
limits, and the vast majority of samples from individ- istic of the molecular vibration in water. Microwave
ual cows fall between −0.503 and −0.541 °C, with the absorption is based on the dielectric properties of
average very close to −0.521 °C. Herd or bulk milk will water. Freezing point is a physical property of milk that
exhibit a narrower range unless the supply was is changed by a change in solute concentration.
watered intentionally or accidentally or if the milk is
from an area where severe drought has existed. 15.2.6.2 Nature of Sample
The AOAC Method 961.07 for water added to While many foods will tolerate oven drying at high
milk uses a cryoscope to test for freezing points, and temperatures, some foods contain volatiles that are lost
assumes a freezing point for normal milk of at such temperatures. Some foods have constituents
−0.527 °C. The Food and Drug Administration will that undergo chemical reactions at high temperatures to
reject all milk with freezing points above generate or utilize water or other compounds, and these
−0.507 °C. Since the difference between the freezing reactions affect the calculated moisture content. Vacuum
points of milk and water is slight and since the freez- oven drying at reduced temperatures may overcome
ing point can be used to calculate the amount of such problems for some foods. However, a distillation
water added, it is essential that the method be as pre- technique is necessary for some food to minimize vola-
15.5
274

table
 Comparison of moisture analysis methods

How is water
What is actually removed/reacted/
Method Principle measured? identified/etc.? Cautions/things to control Advantages Disadvantages Typical applications

Forced draft Sample is heated in oven Weight change Heat evaporates Control time, Easy to handle Takes long time to get Official method for many
oven to evaporate water. water when it boils temperature; control many samples at results. High temperature types of samples. Not
Weight loss equals at 100 °C sample particle size. one time can cause loss of suitable of rapid quality
moisture content Must pre-dry some volatiles, lipid oxidation, control results. Not
samples to avoid Maillard browning, suitable for samples
splattering sucrose hydrolysis, so not subject to loss of
suitable for some types volatiles, lipid oxidation,
of foods Maillard browning, or
sucrose hydrolysis

Vacuum oven Sample is heated in oven Weight change Heat sample under Control time and Easy to handle Takes long time to get An official method for
under reduced pressure, reduced pressure temperature. Pull and many samples at results (though usually many types of products.
so water evaporates at a to evaporate water release vacuum slowly one time. Lower less time than with forced Not suitable for rapid
lower temperature. at ~70 °C temperature for draft oven). More quality control. Not
Weight loss equals evaporating water expensive than forced suitable for powdered
moisture content reduced problems draft oven products, since they can
with high-sugar blow around when
products vacuum is pulled and
released

Microwave Sample is heated with Weight change Heat from Control power and time Rapid More expensive than Suitable for rapid quality
drying oven microwave energy to microwave energy to prevent sample other drying methods control, especially for
evaporate water. Weight causes water decomposition. Spread listed. Can only run one liquid products, since use
loss equals moisture evaporation sample evenly. Check sample at a time of pads avoids
content calibration of analytical splattering
balance

Infrared drying Infrared lamp supplies Weight change Heat from infrared Control time and Rapid Expensive. Can only run Suitable for rapid quality
oven heat that penetrates lamp evaporates temperature. Spread one sample at a time control, but not for high
sample to evaporate water sample evenly moisture products (would
water. Weight loss equals get splattering)
moisture content

Rapid moisture Sample is heated with Weight change Heat evaporates Control time and Rapid Expensive. Can only run Suitable for rapid quality
analyzer heating elements to water when it boils temperature. Spread one sample at a time control, but not for high
evaporate water. Weight at 100 °C sample evenly. Regular moisture products (would
loss equals moisture calibration of analytical get splattering)
content balance

(continued)
L.J. Mauer and R.L. Bradley Jr
 Chapter 15
•

How is water
What is actually removed/reacted/
Method Principle measured? identified/etc.? Cautions/things to control Advantages Disadvantages Typical applications

Reflux When sample is heated to Volume of water Co-distill water Any emulsion formed Causes less thermal Can only run one sample AOAC method for spices
distillation toluene (an immiscible from sample from sample with must break to read decomposition of at a time. Solvent is likely
(with toluene) liquid), the toluene and collected after toluene. Collect volume of water. Need some foods than flammable and toxic.
water are co-distilled. distillation and water and very clean glassware oven drying. Reading volume of water
Collected moisture distills condensation measure with no water. Use Solvent protects in receiving tube may be
off, is condensed and caution with solvents sample from losing less accurate than by
collected, and volume of (fire hazards; toxic) volatiles and gravimetric method
water is measured minimized
oxidation. Water is
Moisture and Total Solids Analysis

measured directly

Karl Fischer In titration of sample with Volume of Karl Water in sample Control particle size of No heat, so no Can only run one sample Method of choice for
Karl Fischer reagent, Fischer Reagent reacts with iodine sample and humidity of thermal at a time. Expensive, if many low-moisture foods
water in sample reacts titrated and sulfur dioxide room. Prevent any water decomposition; using automated unit (e.g., dried fruits and
with sulfur dioxide to to cause reduction in glassware. Must rapid. Higher vegetables, candies,
cause reduction of iodine. of iodine standardize KRF. accuracy than chocolate, roasted
Endpoint of titration is Choose another method many other coffee, oils and fats, and
detected when excess if interferences from methods for many low-moisture foods
iodine cannot react with certain food constituents low-moisture foods that are high in sugar or
water. Volume of titrant is (e.g., ascorbic acid, protein). Good method
used to calculate % carbonyl compounds, to try if method with
moisture unsaturated fatty acids) heating and/or vacuum
gives erratic results

Hydrometer Archimedes’ principle. Volume displaced Based on solids Control temperature. Rapid. Easy. Limited applications. Commonly used as rapid
Compare relative density by hydrometer. content of the Need clean hydrometer Enexpensive Measures only solids method to measure solids
(specific gravity) of Read specific solution, to content content of beverages,
sample to that of water at gravity directly determine specific salt brines, and sugar
same temperature from hydrometer. gravity compared solutions. Best applied to
Measuring solids to pure water solutions with only one
content solute in a medium of
water

(continued)
275
 276

15.5
table (continued)

How is water
What is actually removed/reacted/
Method Principle measured? identified/etc.? Cautions/things to control Advantages Disadvantages Typical applications

Refractometer Based on bending of light Refractive index. Based on solids Control temperature. Rapid. Easy. Limited applications; Commonly used as rapid
(i.e., refraction. Measure Measuring solids content of solution, Need clean contact inexpensive measures only solids method to measure solids
refractive index) as it hits content. to determine surface content content of beverages and
surface of product. Commonly refractive index milk, and soluble solids
Refractive index can be calibrated in of fruits and fruit products
used to determine degree Brix (g of and of tomato products
concentration of sucrose/100 g
compound of interest if sample)
nature of compound,
temperature of sample,
and wavelength of light
are constant

Infrared Measure absorption of Amount of NIR Molecular Must calibrate Rapid. Easy. Can Expensive. Can only run Has wide range of food
analyzer infrared radiation at light reflected vibration of instrument for each type be used to estimate one sample at a time. applications, in the
wavelength characteristic from sample functional groups of product and each content of various Value obtained are only laboratory, at-line, and
of the -OH stretch of the of water analyte being analyzed. food constituents estimates/predictions. on-line. NIR is heavily
water molecule. determines Control sample particle Must calibrate the used in the grain/cereal
Concentration of water is absorption, which size. Prevent scratches instrument for each industry for moisture,
determined by energy that is inversely related on glass container for analyte for each type of protein, and fat. NIR is
is reflected or transmitted, to reflected light sample; remember that sample heavily used in the dairy
which is inversely what is measured values obtained are industry for total solids,
proportional to energy only a prediction fat, protein, and lactose
absorbed in milk
L.J. Mauer and R.L. Bradley Jr
Chapter 15 • Moisture and Total Solids Analysis 277

15.6
table
 Common food industry uses of moisture content and water activity assays
Where used:
Production, Product
Technique quality control development Basic research
Water content measurement
   Direct methods Forced draft oven drying X X X
Vacuum oven drying X X X
Microwave analyzer X X
Infrared drying X X
Rapid moisture analyzer technology X X
Thermogravimetric analyzer X X
Lyophilization X X
Chemical desiccation X X
Karl Fischer titration X X X
   Indirect methods Dielectric capacitance X
Hydrometer X
Pycnometer X
Refractometer X X X
NIR spectroscopy (absorbance or X X X
reflectance)
Freezing point, cryoscope X X
Microwave absorption X X X
Conductivity X X
Water activity measurement Dewpoint analyzer X X X
Electric (capacitance or electrolyte) X X X
hygrometer
Freezing point depression X
Tunable diode laser sensor X X

tilization and decomposition. For foods very low in nies for production, product development, and basic
moisture or high in fats and sugars, Karl Fischer titra- research applications are identified.
tion is often the method of choice. The use of a pycnom-
eter, hydrometer, and refractometer requires liquid
samples, ideally with limited constituents. 15.3 WATER ACTIVITY

15.2.6.3 Intended Purposes 15.3.1 Overview

Moisture content analysis data may be needed quickly Water content alone is not a reliable indicator of food
for quality control purposes, in which high accuracy stability, since foods with the same water content dif-
may not be necessary. Of the oven drying methods, fer in their perishability [6]. It is the water activity (aw)
microwave drying, infrared drying, and the rapid mois- of foods that has been correlated to microbial growth,
ture analyzer techniques are fastest. Some forced draft physical properties, and chemical and enzymatic
oven procedures require less than 1 h drying, but most reactions.
forced draft oven and vacuum oven procedures require The water activity (aw) of foods varies greatly, as
a much longer time. The electrical, hydrometric, refrac- shown in Table 15.7. For referring to water activity,
tive index, and microwave absorption methods are the a is lower case and the w is a subscript because an
very rapid but often require correlation to less empiri- activity coefficient of water is being used to describe
cal methods. Oven drying procedures are official meth- its energy state. Generally, foods with higher mois-
ods for a variety of food products. Reflux distillation ture contents also have higher aws, although the rela-
is an AOAC method for chocolate, dried vegetables, tionship between moisture content and aw is not
dried milk, and oils and fats. Such official methods are linear. Water activity is a thermodynamic property
used for regulatory and nutrition labeling purposes. A of water in foods defined as the ratio of the fugacity
survey of food industry usage of the moisture content (or escaping tendency) of water in the food to the
assays is summarized in Table 15.6, wherein the most fugacity of pure water at the same temperature and
commonly used techniques across a variety of compa- pressure [4, 6, 25]. Because fugacity cannot be
directly measured, aw is more commonly determined
278 L.J. Mauer and R.L. Bradley Jr

15.7
Ta b l e
 Water activity (25 °C) of foods and saturated solutions of food ingredients and salts
Saturated solutions
Foods Food ingredients aw control salts
Type aw Type aw Type aw
Potato chips 0.07 Single ingredients LiCl 0.11
Hard candy 0.12 Malic acid 0.58 CH3CO2K 0.23
Crisp crackers 0.13–0.20 Fructose 0.62 MgCl2 0.33
Sugar-free hard candy 0.25 Sorbitol 0.67 K2CO3 0.43
Crisp cookie 0.25 Glucose 0.74 Mg(NO3)2 0.53
Chewy cookie 0.55 Citric acid 0.78 CoCl2 0.65
Honey 0.56 Xylitol 0.79 NaCl 0.75
Beef jerky 0.61 Sucrose 0.85 KCl 0.84
Gummy candy 0.66 Lactose 0.97 K2SO4 0.97
High fructose corn syrup 0.75 Maltose 0.97
Condensed milk, strawberry preserves 0.84 Ingredient blends
Soy sauce 0.87 NaCl +fructose 0.45
Salted butter 0.90 Fructose + citric acid 0.50
Bread 0.94 NaCl+sucrose 0.65
Reduced sugar ketchup, unsalted butter 0.97 NaCl+glucose 0.71
Juices, milk, fruits, vegetables 0.98–0.99
All measurements were done using an AquaLab 4TE instrument (Decagon Devices, Pullman, WA)

as the ratio of the vapor pressure of water in a food ences all aspects of microbial growth, and each microor-
(p) to the vapor pressure of water (po) at the same ganism has a threshold aw below which it will not grow,
temperature and barometric pressure, as shown in although microorganisms may not be dead below this
Eq. 15.13. threshold aw [27]. Understanding which spoilage or
pathogenic microorganisms are a concern for a specific
p
aw = (15.13) food and their aw limits for growth can provide a foun-
po dation for formulation efforts to reduce the aw of the
This concept is similar to how the relative humidity food below that level. Because of the direct correlation
(RH) of the atmosphere is determined. The relation- between aw and microbial growth, there are many food
safety regulations that incorporate aw guidelines (21CFR
ship between the aw and equilibrium relative humidity
110, 113, and 114 [26]; ANSI/NSF Standard 75 [28]; ISO
(ERH) can be expressed as aw = ERH /100 . Water
activity is a dimensionless number between 0 (abso- 21807:2004(E) [29]; AOAC Method 978.18 [14]), and aw
can be a critical control point in HACCP plans.
lute no water) and 1 (pure water).
Food stability maps have been created to display
There are fewer official methods for determining
the relationships between water activity, microbial
the aw of foods than there are for determining the
moisture content [4, 14]. AOAC Method 978.18 growth, chemical and biochemical reaction rates, and
describes techniques to determine the aw of canned physical properties [30]. An example of a general map
vegetables: in the regulation of acidified foods, an aw for an amorphous food is provided in Fig. 15.10,
of 0.85 is used as the cutoff for pathogen growth. In although the reaction rates and properties will vary
21CFR Part 114 [26], low-­acid foods are defined as across different aw ranges for different foods. In addition
to controlling microbial growth, maintaining the desired
foods with an equilibrium pH of >4.6 and aw > 0.85,
and acidified foods are defined as low-­acid foods to aw of a product is a critical aspect in maintaining its tex-
ture and quality throughout shelf-life, and there will be
which acid has been added to create a finished equi-
critical aws beyond which unwanted physical and/or
librium pH of ≤4.6 and that have an aw > 0.85. Foods
chemical changes will occur in most products. If a crispy
with aw < 0.85 are not covered by 21CFR Part 108, 113,
or 114 [26]; however, it is important to recognize that fried product (potato chip) is exposed to an environ-
lowering the aw is not a kill step. ment with a higher RH than its aw, moisture will migrate
into the potato chip, increasing its aw and leading to a
15.3.2 Importance of Water Activity softening of its texture once the critical aw is exceeded.
Likewise, if a food product contains multiple compo-
The aw of foods affects important quality and safety fac- nents with different aws (cheese and cracker, pizza sauce
tors, and therefore aw measurements are commonly on crust, pastry with filling, etc.), water migrates from
used in food development and production. The aw influ-
Chapter 15 • Moisture and Total Solids Analysis 279

1 5 . 1 0 Food stability map (Adapted from Labuza et al. [30] and overlaid with data from Table 15.7)

figure

the region of high aw to the region of low aw until aw

equilibration is attained, which can result in a harden-
ing of one phase and a softening of the other. If these
changes are unwanted, then reformulation of the phases
to the same aw (if possible) or physical separation of the
phases in barrier packaging could extend shelf-life.
Increasing the aw above a critical value can lead to
clumping and caking of powders, recrystallization of
sugars or salts in some products, a glassy-to-rubbery
transition in amorphous solids, or even deliquescence
of water-soluble deliquescent crystalline compounds.

15.3.3 Water Activity Measurement

15.3.3.1 Principles
Since aw = p/po, determining the aw often involves the 1
5 . 11 Dewpoint water activity meter (Courtesy of
figure
Decagon Devices, Pullman, WA)
direct measurement of vapor pressures, although there
are also indirect measurements for aw. For the majority
of aw measurement techniques, a representative food equal) (Fig. 15.11). Enough sample must be used to
sample is sealed in a container and allowed to reach assure that moisture migration during equilibration
equilibrium (i.e., the water will migrate out of the sam- will not dry the sample but not so much sample that
ple into the container headspace, or vice versa, until the sensor is compromised. A general recommenda-
the aw of the food and the RH of the headspace are tion is to fill half of the container with sample. More
280 L.J. Mauer and R.L. Bradley Jr

sample might be warranted for dense, dry products, point leads to an aw error of 0.06 in dewpoint instru-
while less sample could be used for wet products or ments [4, 32]. The majority of reported aw values were
liquids (AOAC Method 978.18 recommends canned measured at 25 °C (as stipulated in AOAC Method
vegetables fill >1/20 of the sample container volume). 978.18) or 20 °C, although the temperatures of interest
The time required to attain headspace equilibrium for a particular product may vary (particularly for
can vary between samples and between assay types. refrigerated or frozen products).
Slow water-emitting samples (dense, dry, dehydrated, It is important to use a calibrated sensor for aw mea-
high fat, high viscosity) may require several hours for surements. Calibration is often done using salt solution
precise aw measurement, while samples c­ontaining standards of known aw (Table 15.7). It is important to
more water or those with water as the continuous comply with instrument manufacturer guidelines for
phase may equilibrate much faster. For establishing the cleaning, calibrating, and maintaining aw instruments.
equilibrium criterion for a sample, the aw could be mon- Calibration may involve at least three measurement
itored over time, such as at 15, 30, 60, 120 min (AOAC points (ISO 21807) or ≥5 salts (AOAC Method 978.18)
Method 978.18), extending out in 60 min intervals until that encompass the aw range of interest for the sample
consecutive readings vary by <0.01 aw (or other des- and sensor to generate a standard curve and/or deter-
ignated equilibrium criteria for a particular product). mine an offset adjustment. Once the instrument is cali-
While some samples may require 24 h for equilibra- brated and temperature and headspace equilibrium
tion, many foods (barring the slow water emitters) may are reached, the headspace is analyzed to determine
reach equilibrium within 5 min for dewpoint analyses the aw. There are a variety of sensors and methods to
or 30–90 min for electric hygrometers [31]. directly or indirectly determine the equilibrium water
Temperature is an important factor in equilibra- vapor pressure in the headspace, although only the
tion and aw – higher temperatures often lead to higher most common used by the food industry (Tables 15.6
aw readings for many foods if the moisture content is and 15.8) will be further described. Examples of aw
kept constant (water molecules move faster as tem- methods include dewpoint measurement, electric
perature is increased). This is why temperature is hygrometer sensors (including capacitance and elec-
reported along with aw values. Additionally, inequali- trolyte sensors), direct measurement of manometric
ties in temperature between samples and containers or pressure, change in length of a hair/thread, increase
equipment can lead to unwanted water condensation in sorbent mass in an isopiestic method, tunable diode
and/or errors in determining the aw. For example, a laser sensor, thermocouple psychrometry, and freez-
1 °C temperature difference between sample and dew ing point determination [5, 33].

15.8
table
 Comparison of water activity measurement methods

Internal
temperature Volatile
Method Instrument aw range Accuracy Repeatability Resolution control Read time interference
Dewpoint AquaLab4 0.030–1.000 0.003 0.001 0.0001 15–50 °C 2–5 min Yes
(Decagon
Devices)
Electric (resistive LabMaster-aw 0.03–1.000 0.003 0.001 0.001 0–50 °C 10–15 min Yesa
electrolytic (Novasina)
based on
electrolyte)
Electric HygroLab C1 0.0–1.000 0.008 0.001 0.001 No 4–6 min Minimalb
(capacitance) (Rotronic)
Electric Aw Therm 0.005–1.000 0.005 0.001 0.001 10–60 °C 4–6 min Minimalb
(capacitance) (Rotronic)
Tunable diode TDL (Decagon 0.03–1.000 0.005 0.001 0.0001 15–50 °C 2–5 min No
laser Devices)
Adapted from Fontana and Campbell [4] and personal communications with Brady Carter (Decagon Devices), Markus Bernasconi
(Novasina), and Harry Trainor and Rico Hasler (Rotronic)
a
For the LabMaster-aw, filters can be used to block out low to medium levels of volatiles, or use the CM-3 sensor that is resilient to
alcohol
b
The sensors in the HygroLab C1 and Aw Therm instruments are affected by volatiles when subjected to high concentrations and
long-term exposure. A single aw measurement in 4–6 min is not enough exposure in most cases for volatiles to affect the analysis
Chapter 15 • Moisture and Total Solids Analysis 281

1 5 . 1 2 Schematic of water activity measurement chambers for the two most common techniques used by the food

figure
industry: dewpoint and capacitance/electrolyte sensors

15.3.3.2 Chilled Mirror Dew Point perature [31]. Some models have internal temperature
In a chilled mirror dewpoint instrument, such as a control features, while others may require an incuba-
Decagon Devices AquaLab (Fig. 15.12), a sample is tor or water-jacketed sample chamber for temperature
equilibrated in a temperature-controlled sealed control. The sensors monitor the electrical conductiv-
chamber that contains a fan to circulate air in the head- ity of their immobilized salt solutions or a hygroscopic
space, an infrared thermometer to measure sample tem- polymer (the sensor composition is proprietary) that
perature, a temperature-­controlled mirror, and a sensor sorb or desorb water in response to the headspace
that detects condensation on the mirror [5]. The mirror RH. Changes in electric conductivity are calibrated
temperature is controlled by a thermoelectric Peltier to reflect changes in the headspace RH, and aw is cal-
cooler, and a photodetector monitors the reflectance of culated as aw = ERH / 100 . Volatile compounds may
light off of the mirror. The aw of the sample is the same as interfere with aw measurements by electric hygrom-
the RH of the headspace at equilibrium (aw = ERH/100), eters, and tunable diode laser techniques may be a bet-
and it is the headspace RH that is analyzed. The tem- ter option for accurately determining the aw of samples
perature of the mirror is cooled until condensation from that contain alcohols, acetic acid, or propylene glycol.
the headspace first appears, which is determined by a
change in the reflectance. When condensation occurs,
15.4 MOISTURE SORPTION ISOTHERMS
the temperatures of the sample and the mirror are
recorded. The temperature of the sample is used to 15.4.1 Overview
determine the water vapor pressure (po), and the dew-
point temperature is used to determine the vapor pres- A moisture sorption isotherm is a plot of the equilib-
sure of water in the headspace (p). The aw of the sample rium relationship between a sample’s moisture con-
is determined by applying Eq. 15.13, aw = p/po. Volatiles tent and aw at a constant temperature and pressure.
that may condense on the mirror, such as alcohols, ace- The hygrocapacity, or water-holding capacity, of a
tic acid, or propylene glycol, interfere with accurate aw sample is the amount of water it contains when equili-
measurements using dewpoint techniques. brated at a set aw. Hygrocapacity depends on the affin-
ity between water and the sample, temperature, RH,
15.3.3.3 Electric Hygrometer surface area, and water-­solid interaction mechanism
In an electric hygrometer, which for our purposes [34]. A key to establishing moisture sorption isotherms
will include capacitance and electrolyte sensors such is to ensure that the samples have equilibrated at the
as Novasina and Rotronic instruments, a sample con- set aws. Different samples exhibit different kinetics of
tainer is equilibrated in a chamber with a potentiom- moisture sorption and desorption. In the absence of
eter and a sensor (such as an electrolytic moisture equilibrium, a plot of moisture content versus aw
sensor) that measures the headspace RH and tem-
282 L.J. Mauer and R.L. Bradley Jr

would simply be called a sorption profile (not an tion <1 °C. Samples should be stored and analyzed in
isotherm). triplicate over time, weighing to ±0.0001 g until some
Generally as aw increases, moisture content also equilibrium criterion is reached. For example, equilib-
increases although the relationship is most certainly rium could be defined as <0.01 % mass change across
not linear. Different types of solids and different foods three consecutive days of weighings. The amount of
have different shapes of moisture sorption profiles, time it takes for samples to reach equilibrium varies by
and these shapes have been organized into six major sample type and RH condition, often from 10 to
types (I–IV) [35]. Types II (sigmoid) and III (_/-shaped) 21 days. Desiccator methods are time and labor inten-
are most common for foods. The inflection points and sive, there is potential for RH fluctuation as samples
slopes of the different isotherms provide useful infor- are removed from the desiccators for weighing, and
mation about hygroscopicity, potential for reformulat- challenges with chemical change or microbial growth
ing or drying to alter aw, how a sample might respond in samples during extended equilibration may be
to different environments or formulations, and stability encountered. There are a limited number of RHs gen-
(microbial, physical, and chemical). Inflection points in erated by saturated salt solutions, and therefore equa-
sorption profiles occur during phase changes, such as a tions (such as the GAB and BET models) must be used
glassy-to-rubbery transition, hydrate formation, or del- to calculate sorption profiles from the measured data
iquescence. Recrystallization events are often accom- points [37]. However, the relative simplicity of the
panied by expulsion of water from the matrix and a technique and ability to store multiple samples are
corresponding weight loss. Overlaying isotherms from attractive features of using the desiccator method to
different food types is a useful practice when consider- generate moisture sorption isotherms.
ing co-formulation and multicomponent foods, keeping
in mind that water does not move if the aws of the dif- 15.4.3 A
 utomated Gravimetric Moisture
ferent components are equal. If differences in aw are ini- Sorption Balance
tially present between the different food components,
water will migrate until aw equilibrium is reached. In automated gravimetric moisture sorption balances,
There are several approaches and instruments sometimes referred to as static dynamic vapor sorp-
used to generate moisture sorption isotherms (or pro- tion (DVS) instruments, the mass of a sample is moni-
files), although there are no government regulations or tored as a function of time, while the sample is exposed
AOAC official methods dictating techniques and to a series of tightly controlled RH step changes at con-
parameters. A description of two common approaches stant temperature. An example of data collected using
for moisture sorption isotherm measurement is pro- this technique is provided in Fig. 15.13. The initial
vided below. moisture content of the sample must be known, or a
sample can be dried in the instrument prior to analy-
sis. The parameters used in generating automated
15.4.2 Isopiestic Desiccator Method moisture sorption isotherms (or profiles), including
In desiccator methods, samples are equilibrated in equilibrium criterion (% weight change in a set time
RH-controlled desiccators at set temperatures, and, period, e.g., 0.001 % or 0.01 %), step time (how long
upon equilibrium, their moisture contents are deter- will a set RH be held before the next programmed RH
mined by either a mass change from a known starting step is initiated, in the absence of attaining the equilib-
amount or any of the moisture content methods rium criterion), and RH step size (1 % RH, 5 % RH,
described above. The RH in the desiccators is most
often controlled by saturated salt solutions (Table 15.7)
with the exception of the 0%RH condition, which is
created using calcium sulfate (DrieRite™) or phospho-
rous pentoxide (P2O5), the latter of which is more effec-
tive [36]. Enough saturated salt solution and headspace
must be present to maintain a constant RH as moisture
is exchanged with stored samples. General recommen-
dations are to use ≥10 % of the container volume filled
with salt solution, >10:1 ratio of salt solution surface
area to sample surface area, and a 20:1 ratio of head-
space volume to sample volume. Vacuum may or may
not be applied to the desiccators. Six to nine different
RH chambers are recommended to generate a mois-
ture sorption isotherm for a product [37]. Temperature
is controlled by placing the desiccators in temperature-­ 1 5 . 1 3 Moisture sorption isotherm (type II) of an

figure
amorphous powder
controlled incubators, ideally with temperature varia-
Chapter 15 • Moisture and Total Solids Analysis 283

1 5 . 1 4 Moisture sorption data for an amorphous powder collected on a DVS instrument (SPS, ProUmid GmbH,

figure
Germany) and used to generate the moisture sorption isotherm shown in Fig. 15.13

10 % RH, etc.), have a profound influence on the curves until a constant RH is reached (on average 5–6 min),
generated (Fig. 15.14). and at this point both the mass and the water activity
When a DVS instrument moves to a programmed of the sample (based on the chilled mirror dewpoint
RH step, air with the programmed RH is moved into technique) are recorded. While it is not possible to
the sample chamber, and the mass of the sample is program RH steps or timing between measurements
monitored as it sorbs (or desorbs) water. Usually the into the DDI technique, the DDI sorption profiles are
initial sorption is more rapid than later stages of sorp- very high resolution with numerous data points col-
tion, as seen in Fig. 15.14. If true equilibrium is reached lected and may be collected in less time than the static
between the sample and the surrounding environmen- DVS moisture sorption isotherms.
tal RH, then no more moisture uptake would occur and
the equilibrium criterion would be 0.00 % mass change
15.4.4 P
 hase Diagrams Containing aw,
over any time scale. However, due to the labile nature
Moisture Content, and Tg
of some foods and the need to collect data for foods
Relationships
over a period of days not months, an equilibrium cri-
terion of 0.01 % weight change over some time scale In addition to routine moisture analyses for quality
(1–5 h) is more common. It is important to recognize control, the food industry is increasingly taking a
that the shorter the step time, the more likely the fundamental materials science approach to designing
sample is not in equilibrium with the surrounding RH foods and processes and controlling food quality. An
resulting in sorption profiles and not isotherms. important practice here is to establish not only mois-
Careful control and reporting of test parameters are ture sorption isotherm relationships for foods but
important for reproducibility of moisture sorption iso- also to establish the relationship between moisture
therms. Advantages of DVS methods compared to des- content and glass transition temperature (Tg) as well
iccator methods include precise RH control and mass as the relationship between water activity and glass
determinations, a smaller amount of sample transition temperature (Tg). More detailed informa-
(5 mg–1.5 g), smaller RH step sizes, and possibly a tion about Tg measurements is provided in Chap. 31.
reduced time to generate a moisture sorption isotherm
for a single sample. Disadvantages of DVS methods
include high instrumentation costs, the upper RH limit 15.5 SUMMARY
being 95 % (or lower RH at higher temperatures), and a
more limited sample size (many DVS instruments only
The moisture content and water activity of foods are
analyze one sample at a time, although there is cur-
important to food processors and consumers for a
rently one that analyzes up to 23 samples simultane-
variety of reasons. While moisture determination may
ously (SPS, ProUmid GmbH & Co., Germany)).
seem simplistic, it is often one of the most difficult
A modified approach to the DVS instrumentation
assays in obtaining accurate and precise results. Direct
is found in the AquaLab Vapor Sorption Analyzer
moisture content analysis methods involve a separa-
(Decagon Devices, Pullman, WA) operated in DDI
tion of moisture in the sample from the solids and
mode. In the DDI mode, a small amount of saturated
then quantitation by weight or volume. Indirect meth-
(100 % RH) or dry (0 % RH) air is introduced into the
ods do not involve such a separation but instead are
sample chamber, the chamber is sealed, the RH is
based on some physical or chemical property of the
continuously monitored by a capacitance RH sensor
284 L.J. Mauer and R.L. Bradley Jr

sample that varies with respect to its moisture con- • Food high in vitamin C
tent. A major difficulty with many methods is attempt- • Food high in unsaturated fatty acids
ing to remove or otherwise quantitate all water
5. The procedure for an analysis for moisture in a
present. This often is complicated by decomposition
liquid food product requires the addition of
or interference by other food constituents. For each
1–2 ml of deionized water to the weighed sam-
moisture analysis method, there are factors that must
ple in the moisture pan. Why should you add
be controlled or precautions that must be taken to
water to an analysis in which moisture is being
ensure accurate and precise results. Careful sample
determined?
collection and handling procedures are extremely
6. A new instrument based on near-infrared prin-
important and cannot be overemphasized. The choice
ciples has been received in your laboratory to be
of moisture analysis method is often determined by
used in moisture analysis. Briefly describe the
the expected moisture content, nature of other food
way you would ascertain if the new instrument
constituents (e.g., highly volatile, heat sensitive),
would meet your satisfaction and company
equipment available, speed necessary, accuracy and
standards.
precision required, and intended purpose (e.g., regu-
7. A technician you supervise is to determine the
latory or in-plant quality control).
moisture content of a food product by the Karl
Fischer method. Your technician wants to know
15.6 STUDY QUESTIONS what is this “Karl Fischer reagent water equiva-
lence” that is used in the equation to calculate
percentage of moisture in the sample, why is it
1. Identify five factors that one would need to con-
necessary, and how it is determined. Give the
sider when choosing a moisture content analy-
technician your answer.
sis method for a specific food product.
8. You are fortunate to have available in your labo-
2. Why is standardized methodology needed for
ratory the equipment for doing moisture con-
moisture content determinations?
tent analysis by essentially all methods – both
3. What are the potential advantages of using a
official and rapid quality control methods. For
vacuum oven rather than a forced draft oven for
each of the food products listed below (with the
moisture content determination?
purpose specified as rapid quality control or
4. In each case specified below, would you likely
official), indicate (a) the name of the method
overestimate or underestimate the moisture
you would use, (b) the principle (not procedure)
content of a food product being tested? Explain
for the method, (c) a justification for use of that
your answer.
method (as compared to using a hot air-drying
(a) Forced draft oven: oven), and (d) two cautions in use of the method
• Particle size too large to ensure accurate results:
• High concentration of volatile flavor
compounds present (a) Ice cream mix (liquid) – quality control
• Lipid oxidation (b) Milk chocolate – official
• Sample very hygroscopic (c) Spices – official
• Alteration of carbohydrates (e.g., (d) Syrup for canned peaches – quality control
Maillard browning) (e) Oat flour – quality control
• Sucrose hydrolysis
9. You are a manufacturer of processed cheese.
• Surface crust formation
The maximum allowed moisture content for
• Splattering
your product is 40 %. Your current product has
• Desiccator containing DrieRite™ or P2O5
a mean moisture content of 38 %, with a stan-
desiccant (~0%RH) with dried sample
dard deviation of 0.7. It would be possible to
not sealed properly
increase your mean moisture content to 39.5 %
(b) Toluene distillation: if you could reduce your standard deviation to
• Emulsion between water in sample and 0.25. This would result in a saving of $3.4 mil-
solvent not broken lion per year. You can accomplish this by rap-
• Water clinging to condenser idly analyzing the moisture content of the
cheese blend prior to the cooking step of manu-
(c) Karl Fischer:
facture. The cheese blend is prepared in a batch
• Very humid day when weighing original
process, and you have 10 min to adjust the
samples
moisture content of each batch:
• Glassware not dry
• Sample ground coarsely
Chapter 15 • Moisture and Total Solids Analysis 285

(a) 
Describe the rapid moisture analysis 15.7 PRACTICE PROBLEMS
method you would use. Include your ratio-
nale for selecting the method.
(b) How would you ensure the accuracy and 1. As an analyst, you are given a sample of con-
precision of this method (you need to be densed soup to analyze to determine if it is
sure your standard deviation is below 0.25)? reduced to the correct concentration. By gravi-
metric means, you find that the concentration is
10. You work in a milk drying plant. As part of the 26.54 % solids. The company standard reads
production process, you need to rapidly ana- 28.63 %. If the starting volume were 1000 gal at
lyze the moisture content of condensed milk: 8.67 % solids and the weight is 8.5 lb per gallon,
(a) What rapid secondary method would you how much more water must be removed?
use, and what primary method would you 2. Your laboratory just received several sample
use to calibrate the secondary method? containers of peas to analyze for moisture con-
Additionally, how would you ensure the tent. There is a visible condensate on the inside
accuracy and precision of your secondary of the container. What is your procedure to
method? obtain a result?
(b) Your results with the secondary method are 3. You have the following gravimetric results:
consistently high (about 1 %), based on the weight of dried pan and glass disk is 1.0376 g,
secondary method you chose. What are weight of pan and liquid sample is 4.6274 g, and
some potential problems and how would weight of the pan and dried sample is 1.7321 g.
your correct them? What was the moisture content of the sample
and what is the percent solids?
11. During a 12 h period, 1000 blocks (40 lbs each)
from ten different vats (100 blocks per vat) of Answers
Cheddar cheese were produced. It was later
realized that the cooking temperature was too 1. The weight of the soup initially is superfluous
low during cheesemaking. You are concerned information. By condensing the soup to 26.54 %
that this might increase the moisture content of solids from 8.67 % solids, the volume is reduced to
the cheese above the legal requirement. Describe 326.7 gal [(8.67 %/26.54 %) × 1000 gal]. You need
the sampling plan and method of analysis you to reduce the volume further to obtain 28.63 %
would use to ­determine the moisture content of solids [(8.67 %/28.63 %) × 1000 gal], or 302.8 gal.
the cheese. You want the results within 48 h so The difference in the gallons obtained is 23.9 gal
you can determine what to do with the cheese. (326.7−302.8 gal), or the volume of water that
12. Compare and contrast moisture content and must be removed from the partially condensed
water activity in terms of measurement soup to comply with company standards.
approaches, effects on food texture and mois- 2. This problem focuses on a real issue in the food
ture migration with the environment, influence processing industry – when do you analyze a
on microbial growth, and roles in vitamin stabil- sample and when don’t you? It would appear
ity, lipid oxidation, and the Maillard reaction. that the peas have lost moisture that should be
13. Identify three factors that would lead to under- within the vegetable for correct results. You will
estimation of the water activity of a food sam- need to grind the peas in a food mill or blender.
ple. Then identify three factors that would lead If the peas are in a Mason jar or one that fits a
to overestimation of the water activity of a food blender head, no transfer is needed. Blend the
sample. peas to a creamy texture. If a container transfer
14. You are put in charge of creating moisture sorp- was made, then put the blended peas back into
tion isotherms of several food products. the original container. Mix with the residual
Describe what method you will use and what moisture to a uniform blend. Collect a sample
approach you will take for assuring you have for moisture analysis. You should note on the
created moisture sorption isotherms and not report form containing the results of the analy-
moisture sorption profiles of: sis that the pea samples had free moisture on
container walls when they arrived.
(a) Potato chips 3. Note Eqs. 15.2, 15.3, and 15.4 in Sect. 15.2.2.1.7.
(b) An intermediate moisture nutrition bar con- To use any of the equations, you must subtract
taining oats, chocolate pieces, and a binder the weight of the dried pan and glass disk.
syrup Then you obtain 3.5898 g of original sample
(c) A gummy candy and 0.6945 g when dried. By subtracting these
(d) Cheddar cheese results, you have removed water (2.8953 g).
286 L.J. Mauer and R.L. Bradley Jr

Then (0.6945 g/3.5898 g) × 100 = 19.35 % solids 19. Bouraoui M, Richard P, Fichtali J (1993) A review of mois-
and (2.8953 g/3.5898 g) × 100 = 80.65 % water. ture content determination in foods using microwave
oven drying. Food Res Inst Int. 26:49–57
20. Mitchell J Jr, Smith DM (1948) Aquametry. Wiley, New York
21. Giese J (1993) In-line sensors for food processing. Food
Technol 47(5):87–95
REFERENCES 22. Wilson RH, Kemsley EK. On-line process monitoring
using infrared techniques. . Food processing automation
1. Pomeranz Y, Meloan C (1994) Food analysis: theory and II proceedings of the American society of agricultural
practice, 3rd edn. Chapman & Hall, New York engineers, ASAE Publication 02–92. St. Joseph, MI:
2. Josyln MA (1970) Methods in food analysis, 2nd edn. American Society of Agricultural Engineers; 1992.
Academic, New York 23. Pande A. Handbook of moisture determination and con-
3. US Department of Agriculture, Agricultural Research trol. New York: Marcel Dekker; 1974.
Service (2016). USDA National Nutrient Database for 24. Anonymous. Kett Microwave Moisture Meters Santiago,
Standard References, Release 28. Nutrient Data CA: Kett; 2009 [cited 2016]. Available from: www.kett.
Laboratory Home Page. http://ndb.nal.usda.gov com.
4. Fontana Jr AJ, Campbell CS (2004) Water activity (Chapter 25. Reid DS. Water activity: Fundamentals and relationships.
3). In: Nollet L (ed.) Handbook of Food Analysis, 2nd In: Barbosa-Canovas BV, Fontana AJ, Schmidt SJ, Labuza
edn. Marcel Dekker, Inc., New York p. 39–54. TP (eds). Water activity in foods. Ames, IA: Blackwell
5. Fontana Jr AJ (2007) Measurement of water activity, mois- Publishing; 2007. p. 15–28.
ture sorption isotherms, and moisture content of foods 26. Anonymous. Code of Federal Regulations. Title 21 Parts
(Chapter 7). In: Barbosa-Canovas BV, Fontana AJ, 108, 113, 114. Washington DC: US Government Printing
Schmidt SJ, Labuza TP (eds). Water activity in foods: Office; 2015.
Fundamentals and applications. Blackwell Publishing, 27. Scott WJ. Water relations of food spoilage microorgan-
Ames, IA, p. 155–71. isms. Advances in Food Research, 1957; 7:83–127.
6. Reid DS, Fennema OR (2007) Water and ice (Chapter 2). 28. International N. Non-potentially hazardous foods. . Ann
In: Damodaran S, Parkin K, Fennema OR (eds). Fennema’s Arbor, MI: 2000 Nov. 10. Report No.: ANSI/NSF 75–2000.
food chemistry, 4th edn. CRC, Boca Raton, FL 29. Anonymous. International Standard ISO 21807:
7. Arunan EDG, Klein RA, Sadlej J, Scheiner S, Alkorta I, Microbiology of food and animal feeding stuffs –
Clary DC, Crabtree RH, Dannenbery JJ, Hobza P, Determination of water activity. Switzerland:
Kjaergaard HG, Legon AC, Mennucci B, Nesbitt DJ (2011) International Organization for Standardization; 2004.
Definition of the hydrogen bond (IUPAC recommenda- 30. Labuza TP, Tannenbaum, SR, Karel, M. Water content
tions 2011). Pure Applied Chemistry 83(8):1637–41 and stability of low moisture and intermediate moisture
8. Martiniano H, Galamba, N (2013) Insights on hydrogen- foods. Journal of Food Technology. 1970; 24:543–50.
bond lifetimes in liquid and supercooled water. J Physical 31. Scott VN, Clabero RS, Troller JA. Chapter 64. Measurement
Chem B. 117:16188–95 of water activity (aw), acidity, and brix. In: Downes FP,
9. Schmidt S (2004)Water and solids mobility in foods. Adv Ito K, (eds). Compendium of Methods for the
Food and Nutr Res. 48:1–101 Microbiological Examination of Foods, 4th Edn: American
10. Zografi G (1988) States of water associated with solids. Drug Public Health Association; 2001. P. 649–657.
Development And Industrial Pharmacy. 14(14):1905–26 32. Decagon Devices. AquaLab water activity meter opera-
11. Emmons DB, Bradley RL Jr, Sauvé J P, Campbell C, Lacroix tor’s manual, Version 4. Pullman, WA: Decagon Devices;
C, Jimenez - Marquez SA (2001) Variations of moisture 2009.
measurements in cheese. J AOAC Int. 84:593–604 33. Rodel W. Water activity and its measurement in food
12. Emmons DB, Bradley RL Jr, Campbell C, Sauve JP (2001) (Chapter 16). In: Kress-Rogers E, Brimelow CJB (eds).
Movement of moisture in refrigeration cheese samples Instrumentation and Sensors for the Food Industry.
transferred to room temperature. J AOAC. 84:620–622 Cambridge: Woodhead; 2001. p. 453–83.
13. Bradley RL Jr, Vanderwarn MA (2001) Determination of 34. Airaksinen S, Karjalainen M, Shevchenko A, Westermarck
moisture in cheese and cheese products. J AOAC Int. S, Leppanen E, Rantanen J, et al. Role of water in the
84:570–592 physical stability of solid dosage formulations. J Pharm
14. AOAC International (2016) Official methods of analysis, Sci. 2005:2147–65.
20th edn. (online). AOAC International, Rockville, MD 35. Sing KSW, Everett DH, Haul RAW, Moscou L, Pierotti
15. Wehr HM, Frank JF (eds) (2004) Standard methods for the RA, Rouquerol J, et al. Reporting physisorption data for
examination of dairy products, 17th edn. American gas/solid systems with special reference to the determi-
Public Health Association, Washington, DC nation of surface area and porosity. Pure and Applied
16. AACC International (2010) Approved methods of analy- Chemistry. 1985; 57(4):603–19.
sis, 11th edn. AACC International, St. Paul, MN 36. Trussel F, Diehl H. Efficiency of chemical desiccants.
17. Nelson OA, Hulett GA (1920) The moisture content of Analytical Chemistry. 1963; 35:674–7.
cereals. J Ind Eng Chem. 12:40–45 37. Bell LN, Labuza TP. Moisture Sorption: Practical Aspects
18. International Dairy Federation (1982) Provisional of Isotherm Measurement and Use. 2nd Edn. St. Paul,
Standard 4A. Cheese and processed cheese: determina- MN: American Association of Cereal Chemists; 2000.
tion of the total solids content. International Dairy 38. CRC Handbook of Chemistry and Physics, 96th ed. CRC
Federation, Brussels, Belgium Press: Boca Raton, FL, 2015.