Part 2 Interpretation

Room temperature susceptibility

How do we interpret a value of magnetic susceptibility measured at room temperature?
Magnetic susceptibility gives us information about the mineralogy and geochemistry of
environmental materials. From mineralogy we can often deduce additional information about
the material, such as its origin or the chemistry of its environment. Information about the origin
may give us further information about the environmental conditions which gave rise to the
minerals. An infinite number of environmental conditions gives rise to a very wide range of
mineralogies and magnetic susceptibility values, as shown in Figure 2.1. It shows that virtually
all types of environmental materials have been measured with values ranging from <0.001 to
>30 x 10-6 m3 kg-1 encompassing five to six orders of magnitude.

Ferrimagnetic minerals

Burned soils
Basic / ultrabasic rocks

Topsoils

Intermediate igneous rocks

Canted antiferromagnetic minerals

Acid igneous rocks

Coarse metamorphic rocks

Paramagnetic minerals

Medium / fine metamorphic rocks

Sedimentary rocks
Diamagnetic minerals

-0.01 -0.001 0.001 0.01 0.1 1 10 100 1000

Figure 2.1 Typical ranges of room temperature magnetic susceptibility values measured for
environmental materials and minerals

One kind of interpretation procedure is shown in Figure 2.2. Not all measurements can or need
to be interpreted strictly in this way. Whilst representing a useful approach, the procedure
followed in conjunction with the sections below acts as a tutorial in the subject for users who
are unacquainted with magnetic and environmental theories. More experienced users may wish
to move directly to certain other sections or to tables of data in order to make comparisons
with their own results. It begins by making the assumption that the magnetic susceptibility of
most environmental materials is controlled by the mixture of minerals present. Minerals are
naturally occurring chemical compounds which are usually in the form of crystals. Their
internal structure is determined by the arrangement of atoms of chemical elements into a three-
dimensional pattern termed a lattice. The magnetic behaviour of a mineral is controlled both by
the particular atoms making up the lattice and the way in which the lattice is structured. In
some minerals the same atoms give rise to different magnetic states because they can form
alternative lattice structures. The minerals which show strong magnetic responses tend to
contain atoms of iron (Fe) in their lattice.

36
 Magnetic susceptibility
value of a sample

Mixtures of minerals and materials

Ferrimagnetic Non-Ferrimagnetic
minerals minerals

Mineral properties
concentration
composition
shape size

Primary minerals Secondary minerals

Rock forming Burning / combustion

minerals Soil formation
Bacteria
Authigenic iron sulphides

Figure 2.2 A procedure for interpreting magnetic susceptibility

37
Table 2.1 Magnetic behaviour and magnetic susceptibility

Ferromagnetic Strong positive susceptibility

e.g. pure iron, nickel, chromium

Ferrimagnetic Strong positive susceptibility

Some iron oxides and sulphides, e.g. magnetite, maghemite,
pyrrhotite, greigite

Canted antiferromagnetic Moderate positive susceptibility

Some iron oxides, e.g. hematite, goethite

Paramagnetic Weak positive susceptibility

Many Fe-containing minerals and salts, e.g. biotite, olivine,
ferrous sulphate

Diamagnetic Weak negative susceptibility

e.g. water, organic matter, plastics, quartz, feldspars, calcium
carbonate

The idea of minerals and materials having different magnetic status was introduced in Part 1.
There are five categories of magnetic behaviour, shown in Table 2.1 ranked in decreasing
order of typical magnetic susceptibility. From this information and the previous discussion of
magnetic susceptibility theory (Figure 1.1) it can be presumed that a sample of rock or soil
containing predominantly ferrimagnetic minerals will have a higher susceptibility than one
containing, say, all paramagnetic minerals. And this is certainly a good guide to explaining the
relative magnitude of susceptibilities in samples of pure minerals. However, it is uncommon
for a natural sample to contain only one category of magnetic minerals.

It is therefore necessary to consider virtually all samples as a mixture of minerals often falling
into two or three categories of magnetic behaviour and each having a different magnetic
susceptibility value. Therefore, we need to know the magnetic susceptibility values of
individual minerals in order to interpret the magnetic susceptibility values of bulk samples. A
list of individual magnetic susceptibilities of minerals and materials is given in Table 2.2.

There are exceptions to a mineral basis for environmental magnetism. The weakest
diamagnetic group (Table 2.1) of materials contains non-crystalline forms like wood, water
and plastic. The strongly magnetic ferromagnetic group (Table 2.1) comprises ‘pure’ particles
of the elements iron, nickel and chromium, members of the transition element range or Group
8 of the Periodic Table.

38
Mixtures of minerals
In theory, we can explain or predict the magnetic susceptibility of a sample in terms of the sum
of the magnetic susceptibility values of the individual minerals and materials shown in Table
2.2. The idea of interpreting each measurement in terms of many different minerals sounds a
fairly daunting task. But in practice we can simplify matters by making some assumptions
about which minerals are significant in a sample.

Samples which are not contaminated by ferrous metal do not usually contain ferromagnetic
materials. In their absence, the susceptibility of a sample is most likely to be controlled by the
ferrimagnetic component, and less likely by the other categories of minerals present, shown in
Table 2.1. Magnetite, for instance, is about 1000 times more magnetic than the strongest
canted antiferromagnetic or paramagnetic mineral, and about 10000 times stronger than the
weakest clay mineral. Ferrimagnetic iron oxides, like magnetite, are found in virtually all
environments.

In igneous rocks, magnetite may represent about 1-2% of the minerals. But even in these
relatively small proportions, its high susceptibility will often mean that it contributes more to
the susceptibility of the whole sample than does the combined effect of all the other minerals.
The dominating effect of magnetite can be seen by considering a hypothetical soil. Imagine a
soil composed of organic matter, water, quartz sand, clay and iron oxides (Table 2.3). With the
information about the concentration or proportion of the different materials and minerals
present (column 1), and their individual susceptibilities taken from Table 2.2, we can see how
the total susceptibility value is produced. By multiplying the fraction or concentration (column
2) by the specific susceptibility (column 3) we obtain the susceptibility (column 4) of each
mineral/material component in this soil. The total susceptibility of the soil, 0.5855 x 10-6 m3 kg-
1
is obtained by summing the susceptibility values of the components. Column 5 shows
calculations of the fraction of the total susceptibility held by each type of mineral/material.
Note that a very small concentration (0.1%) of magnetite provides 85.4% of the sample’s
susceptibility, and the largest fractions, the paramagnetic minerals, organic matter, sand and
water (90% of the soil) provide only 6% or so of the total sample susceptibility.

39
Table 2.2 Minerals and magnetic susceptibility

Mineral/Material Formula Iron Mass specific magnetic

(%) susceptibility (10-6 m3 Kg-1)
Ferromagnetic metals
Iron αFe 100 276000
Cobalt Co 204000
Nickel Ni 68850

Ferrimagnetic
Magnetite Fe3O4 72
(0.012-0.069 μm) 513-1116
500-1000
(0.09-2000 μm) 596 ± 77
(1-250μm) 440-716
390-580
Maghemite γFe2O3 70 410, 440
286-371
Titanomagnetite Fe3O4-Fe2TiO4 169-290
Titanohaematite Fe2O3-FeTiO3 281-315
Pyrrhotite Fe7S8 50, 53
Greigite Fe3S4

(Canted) antiferromagnetic
Hematite αFe2O3 70 1.19-1.69
0.58-0.78
0.49-0.65
0.27, 0.31, 0.6, <0.63
Goethite αFeOOH 63 0.35, 0.38, 0.7, <1.26

Paramagnetic (20 °C)

Ilmenite FeTiO3 37 1.7, 2
Ulvospinel Fe2TiO4
Olivine 4[(Mg,Fe)2SiO4] <55 0.01-1.3
Siderite FeCO3 48 1.0
Biotite Mg,Fe,Al silicate 31 0.05-0.95
Pyroxene (Mg,Fe)2Si2O6 <12 0.04-0.94
Chamosite oxidised chlorite 0.9
Nontronite Fe-rich clay 0.863
Amphibole Mg,Fe,Al silicate 0.16-0.69
Epidote Ca,Fe,Al silicate 31 0.25-0.31
Pyrite FeS2 47 0.3
Lepidocrocite γFeOOH 63 0.5-0.75, 0.69
Prochlorite mica-like mineral 0.157

40
Vermiculite complex silicate 0.152
Illite K1Al4(Si,Al)8O2O(OH)4 0.15
Bentonite complex silicate 0.058
Smectite complex silicate 0.05, 0.027
Chalcopyrite CuFeS2 30 0.03
Attapulgite complex silicate 0.02
Dolomite CaMg(CO3)2 0.011

Diamagnetic
Calcite CaCO3 -0.0048
Alkali-feldspar Ca,Na,K,Al silicate -0.005
Plastic -0.005
Quartz SiO2 -0.0058
Organic matter -0.009
Water H2O -0.009
Halite NaCl -0.009
Kaolinite Al4Si4O10(OH)8 -0.019

Data from published and unpublished sources, showing ranges and individual measurements of
susceptibility values and iron content.

Table 2.3 Magnetic susceptibility of soil components

1 2 3 4 5
Mineral/material Fraction χlf Component χlf % Fraction
10-6 m3 kg-1 10-6 m3 kg-1 total χlf
Magnetite 0.001 500 0.5 85.4
Goethite 0.099 0.5 0.0495 8.5
Paramagnetic minerals 0.75 0.05 0.0375 6.4
Organics, quartz, sand and water 0.15 -0.01 -0.0015 -0.3
Total 1.0 500.54 0.5855 100.0

Samples of rocks and soils showing purely paramagnetic behaviour rarely show χlf values
exceeding 0.1 x 10-6 m3 kg-1. Therefore, as a rule-of-thumb, the χlf of any sample with a value
less than this is probably controlled by the concentration of paramagnetic minerals and for
values greater than this by ferrimagnetic minerals. There are exceptions to this rule, especially
in some weak samples where the susceptibility may be controlled by minute concentrations of
ferrimagnetic minerals (see Weak samples).

41
Ferrimagnetic minerals
Most environmental materials have room temperature magnetic susceptibility values controlled
by ferrimagnetic minerals. The iron oxides, magnetite, maghemite, titanomagnetite,
titanomaghemite are the dominant ferrimagnetic minerals in many soils, rocks, sediments and
dusts. Ferrimagnetic iron sulphides are much less common, and are usually found in a narrow
range of environments; greigite in some sediments and waterlogged soils, and pyrrhotite in
some metamorphic and basic igneous rocks. The next sections focus on ferrimagnetic iron
oxides.

Table 2.4 Magnetic susceptibility of magnetite

Titanomagnetite Single domain Ultrafine Fraction

-6 3 -1 -6 3 -1 -6 3 -1
10 m kg 10 m kg 10 m kg % ppm
20 60 100 10 100000
2 6 10 1 10000
0.2 0.6 1 0.1 1000
0.02 0.06 0.1 0.01 100
0.002 0.006 0.01 0.001 10
0.0002 0.0006 0.001 0.0001 1
0.00002 0.00006 0.0001 0.00001 0.1

Magnetic susceptibility of samples with different fractions of pure titanomagnetite (200 x 10-6
m3 kg-1 ), stable single domain magnetite (600 x 10-6 m3 kg-1) and ultrafine superparamagnetic
magnetite (1000 x 10-6 m3 kg-1) in a matrix of zero susceptibility. Fractions defined in mass
percentages, parts per million (ppm). Find the nearest χlf of a sample in columns 1, 2 and 3 (or
recalculate for intermediate values of χlf) and read across for equivalent fractions of minerals
in sample. Thus, a sample with χlf of 20 x 10-6 m3 kg-1 could represent either ~ 10% coarse
titanomagnetite (reading across top line from column 1), ~3.3% stable single domain magnetite
or ~2% ultrafine magnetite (based on recalculations of top values in columns 2 and 3).

Mineral concentration
The range of susceptibilities where it is thought that ferrimagnetic minerals dominate, that is
greater than 0.1 x 10-6 m3 kg-1 covers three to four orders of magnitude (Figure 2.2). Natural
samples lying at either end of this susceptibility range vary in their concentration of
ferrimagnetic minerals by a factor of about 300. Therefore the major factor controlling the
susceptibility of samples in this range will be the concentration of ferrimagnetic minerals; in
other words the total number or volume of ferrimagnetic crystals. The concentration of
magnetite in a sample is estimated by dividing the bulk susceptibility value of the sample by
the susceptibility of the assumed or known magnetite type or size (Table 2.2 and Figure 2.3).
Table 2.4 can be used to estimate concentrations of ultrafine superparamagnetic magnetite,
stable single domain magnetite and coarse titanomagnetite minerals (see Crystal size and
domains below). As Figure 2.2 shows, concentration is only one of four factors controlling
magnetic susceptibility. Magnetic susceptibility also depends on the mineral composition,
crystal size and crystal shape.

42
Mineral composition
The composition of ferrimagnetic iron oxides varies from the ‘pure’ oxides of magnetite and
maghemite, to ‘impure’ oxides such as titanomagnetite and titanomaghemite, in which the Fe
atoms are partially substituted by atoms of titanium (Ti). There are continuous sequences or
solid solutions of minerals between these two sets which have varying titanium contents.
Titanium substitution reduces the Fe content and magnetic moment of the mineral, and hence
lowers the magnetic susceptibility. This is confirmed in Table 2.2, which shows
titanomagnetite with susceptibility values as low as 15% of the highest values for magnetite.
Where titanium substitution has progressed beyond a certain point the minerals lose
ferrimagnetic status and become transformed into the paramagnetic titanium oxides ilmenite
and ulvospinel.

The type of iron oxides which occur in specimens of igneous rock depends upon the chemical
composition of the liquid magma and the mode of crystallisation. The main oxide is
titanomagnetite and the proportion of titanium tends to be higher in basic rather than acid
rocks. Oxidation at high temperatures (>600 °C) during cooling will result in conversion to
magnetite, whilst at low temperatures (<400 °C) oxidation produces titanomaghemite.
Titanomagnetites tend to form during rapid rather than slow cooling of the magma. The iron
oxides will have a range of crystal dimensions from <1 μm to several millimetres, and the
crystal shape can vary from spheres to rods. Weathering of igneous rocks can alter crystal
structure and chemical composition, and has the crucial effect of releasing minerals into soils
and sediments. The iron oxides in sedimentary rocks will in part be inherited from igneous
rocks.

Crystal size and domains

Ferrimagnetic grains are divided up into different regions or cells of magnetisation, known as
domains. Above diameters of ~110 μm, magnetite grains are referred to as multidomain
(MD) because energetically it is favourable to have more than one domain. In small grains
<0.2 μm, the restricted volume allows only one domain to form, and these are termed single
domain (SD) grains. Grains in the interval 0.2 - 110 μm are large enough to favour more than
one domain but show the magnetic properties of single domain grains; these are termed
pseudo-single domain (PSD). Ultrafine grains <0.03 μm are SD but display unique properties.
The magnetisation is strong but unstable due to thermal energies counteracting induced
magnetisation very quickly after a magnetic field is removed. This behaviour is similar to
paramagnetism, but with a much greater susceptibility. Hence it is termed superparamagnetic
(SP) behaviour. Measurements on natural and synthetically produced magnetites of known
sizes (Table 2.2) have helped to define the major changes in susceptibility with crystal size and
domain, and these are shown in Figure 2.3. Recent measurements suggest that the variations of
χlf with crystal size are smaller than previously thought and may be considered to have a
constant mean of 3.1 SI (± 0.4 SI) equivalent to a χlf value of ~596 x 10-6 m3 kg-1 (± 77 x 10-6
m3 kg-1 ) over a very wide range (0.09 - 6000 μm) of crystal sizes. This range includes all the
MD, PSD and stable SD domain states (SSD). In crystal diameters less than 0.03 μm, the
domain state is essentially SP and values of χlf are higher and may exceed 1000 x 10-6 m3 kg-1.
The SP crystals are also characterised by their response to susceptibilities measured at different
frequencies and are detected by frequency dependent measurements (see below).

43
Crystal shape
The crystal shape of the ferrimagnetic mineral assemblage influences magnetic susceptibility
values. Longer crystals have the effect of shifting the boundaries between the domain states
towards coarser crystal sizes. For example, rod-shaped crystals might exhibit SP behaviour in
crystals as long as 0.05 μm. However, without other magnetic measurements or direct
observations by transmission electron microscope, evaluation of the effects of crystal shape on
susceptibility is difficult. In summary, where ferrimagnetic minerals dominate the magnetic
mineral assemblage (i.e. κlf >0.1 x 10-6 m3 kg-1), magnetic susceptibility is controlled by the
following factors in order of decreasing importance:

mineral concentration (χlf varies by factor of 200-300)

mineral composition (χlf varies by factor of 3-4)
crystal size (χlf varies by factor of <2)
crystal shape (χlf varies by factor of <2)

Primary and secondary minerals

Ferrimagnetic minerals which were formed within igneous rocks and which retain all or
virtually all of their magnetic properties are referred to in this text as primary minerals. Other
ferrimagnetic iron oxides and sulphides are referred to as secondary minerals; that is, they
have formed by other processes that include burning, fossil fuel combustion, bacteria, soil
formation, diagenesis and authigenesis. The crystal size or domain state provides a clue as to
the processes of formation of magnetite. Table 2.5 shows that primary rock minerals and
products of fossil fuel combustion tend to fall into MD, PSD and SSD ranges, while burning,
pedogenic processes and bacterial action tend to produce fine secondary crystals of SSD or SP
behaviour. Therefore, in some situations the ability to distinguish between domain and grain
sizes will help to identify the process of crystal formation. However, room temperature
measurements of magnetic susceptibility cannot alone distinguish between secondary or
ferrimagnetic minerals and domains. The next sections therefore deal with how further
measurements of frequency-dependent susceptibility, and low and high temperature
susceptibility may help to identify minerals and domains.

44