The Magnetic View on the Marine Paleoenvironment
577
silica, water) or paramagnetic (e.g., Fe-rich silicates including clays). Dia- and paramagnetic minerals acquire a weak induced magnetization in magnetic fields, but they cannot carry a remanent magnetization' i.e., a permanent magnetization in zero
field. They are discerned by their susceptibility
which is negative for diamagnetic and positive for
paramagnetic minerals.
Natural magnetic minerals are classified as
ferrimagnetic «titano-)magnetite, maghemite,
pyrrhotite, greigite) or antiforromagnetic (hematite,
goethite). F errimagnets exhibit much higher but less
stable magnetizations as compared to antiferromagnets.
The basic physical principles offerromagnets
(e.g., iron) also apply to ferrimagnets and antiferromagnets (e.g., Chikazumi 1997). To emphasize
their fundamentally different origin, diamagnetic
and paramagnetic contributions to the magnetic
properties of sediments are referred to as nonferromagnetic. (Note that there are no natural
ferromagnetic minerals).
The phenomenon of permanent magnetism is
associated to the existence of magnetic domains
"'lith uniformly oriented atomic magnetic moments.
As a function of grain-size, the number of domains
within a magnetic particle increases from one to
several hundreds. Single-domain (SD) and multidomain (MD) particles differ considerably in their
magnetic characteristics. MD grains with a few
domains « 1 0) have similar properties as SD
particles and are categorized as pseudo-singledomain (PSD) particles. The magnetization of
ultra-fine particles below a certain threshold size
is unstable due to thermal activation. Because of
their high susceptibility and, on the other hand, their
incapacity to carry a remanence, they are called
superparamagnelic (SP) particles.
The grain-size thresholds between the four
domain states (SP, SD, PSD and MD) primarily
depend on mineralogy. For magnetite the SP/SD
transition is at 0.03 flm, the SD/PSD boundary at
0.1 flmandparticleslargerthan 10-20 flmshow
MD behavior (Parry 1965; Dunlop 1973; Bailey
and Dunlop 1983). The properties of natural magnetic particles are discussed in detail by Stacey and
Bane~iec (1974), Dunlop (1981; 1986; 1990; 1995)
and Dunlop and Ozdemir (1997). The textbook of
Chikazumi (1997) gives a detailed treatment of the
physics of ferromagnetism.
Experimental Methods
Common rock magnetic parameters are derived
from
- magnetic hysteresis loops,
- remanence acquisition and demagnetization
experiments,
- magnetic susceptibility measurements,
- thermomagnetic analyses,
- evaluation of magnetic anisotropies.
The last two topics are covered by the following
references: Lowrie (1990), Mullender et al. (1993),
Hrouda (1994), Hrouda et al. (1997), Hamilton and
Rees (1970), Rochette et al. (1992), Borradaile and
Stupavsky (1995) and Lehmannet al. (1996). They
will not be further discussed here.
Magnetic Hysteresis Loops
The term magnetic hysteresis refers to the effect
that a sample's magnetization depends not only on
the actual magnetizing field, but also on the preceding magnetization states (Fig. 2).
The shape of a hysteresis loop, for which the
field cycles between positive and negative extremes,
delineates re-magnetization processes. A magnetically hard (high-coercive) mineral (e.g., hematite)
yields a flat and broad hysteresis loop, while a
magnetically soft (low-coercive) mineral (e.g.,
magnetite) displays a steep and narrow loop. In
Figure 2a both hysteresis branches merge at
medium field strengths, but still increase towards
higher fields. From where the outer branch is strictly
linear, the increase in magnetization can exclusively
be attributed to para- or diamagnetic minerals of the
sedimentary matrix. Subtraction of the linear asymptote, i.e., of the non-ferromagnetic susceptibility
Knf' results in the slope corrected hysteresis loop
(Fig. 2b). In the presence of SP and/or antiferromagnetic particles linearity can generally not be
reached as maximum fields are limited to about 1 T
in most laboratories. Sophisticated mathematical
algorithms allow to estimate the effect of partly
unsaturated hard magnetic constituents by the
approach to saturation analysis (von Dobeneck 1996).
577
silica, water) or paramagnetic (e.g., Fe-rich silicates including clays). Dia- and paramagnetic minerals acquire a weak induced magnetization in magnetic fields, but they cannot carry a remanent magnetization' i.e., a permanent magnetization in zero
field. They are discerned by their susceptibility
which is negative for diamagnetic and positive for
paramagnetic minerals.
Natural magnetic minerals are classified as
ferrimagnetic «titano-)magnetite, maghemite,
pyrrhotite, greigite) or antiforromagnetic (hematite,
goethite). F errimagnets exhibit much higher but less
stable magnetizations as compared to antiferromagnets.
The basic physical principles offerromagnets
(e.g., iron) also apply to ferrimagnets and antiferromagnets (e.g., Chikazumi 1997). To emphasize
their fundamentally different origin, diamagnetic
and paramagnetic contributions to the magnetic
properties of sediments are referred to as nonferromagnetic. (Note that there are no natural
ferromagnetic minerals).
The phenomenon of permanent magnetism is
associated to the existence of magnetic domains
"'lith uniformly oriented atomic magnetic moments.
As a function of grain-size, the number of domains
within a magnetic particle increases from one to
several hundreds. Single-domain (SD) and multidomain (MD) particles differ considerably in their
magnetic characteristics. MD grains with a few
domains « 1 0) have similar properties as SD
particles and are categorized as pseudo-singledomain (PSD) particles. The magnetization of
ultra-fine particles below a certain threshold size
is unstable due to thermal activation. Because of
their high susceptibility and, on the other hand, their
incapacity to carry a remanence, they are called
superparamagnelic (SP) particles.
The grain-size thresholds between the four
domain states (SP, SD, PSD and MD) primarily
depend on mineralogy. For magnetite the SP/SD
transition is at 0.03 flm, the SD/PSD boundary at
0.1 flmandparticleslargerthan 10-20 flmshow
MD behavior (Parry 1965; Dunlop 1973; Bailey
and Dunlop 1983). The properties of natural magnetic particles are discussed in detail by Stacey and
Bane~iec (1974), Dunlop (1981; 1986; 1990; 1995)
and Dunlop and Ozdemir (1997). The textbook of
Chikazumi (1997) gives a detailed treatment of the
physics of ferromagnetism.
Experimental Methods
Common rock magnetic parameters are derived
from
- magnetic hysteresis loops,
- remanence acquisition and demagnetization
experiments,
- magnetic susceptibility measurements,
- thermomagnetic analyses,
- evaluation of magnetic anisotropies.
The last two topics are covered by the following
references: Lowrie (1990), Mullender et al. (1993),
Hrouda (1994), Hrouda et al. (1997), Hamilton and
Rees (1970), Rochette et al. (1992), Borradaile and
Stupavsky (1995) and Lehmannet al. (1996). They
will not be further discussed here.
Magnetic Hysteresis Loops
The term magnetic hysteresis refers to the effect
that a sample's magnetization depends not only on
the actual magnetizing field, but also on the preceding magnetization states (Fig. 2).
The shape of a hysteresis loop, for which the
field cycles between positive and negative extremes,
delineates re-magnetization processes. A magnetically hard (high-coercive) mineral (e.g., hematite)
yields a flat and broad hysteresis loop, while a
magnetically soft (low-coercive) mineral (e.g.,
magnetite) displays a steep and narrow loop. In
Figure 2a both hysteresis branches merge at
medium field strengths, but still increase towards
higher fields. From where the outer branch is strictly
linear, the increase in magnetization can exclusively
be attributed to para- or diamagnetic minerals of the
sedimentary matrix. Subtraction of the linear asymptote, i.e., of the non-ferromagnetic susceptibility
Knf' results in the slope corrected hysteresis loop
(Fig. 2b). In the presence of SP and/or antiferromagnetic particles linearity can generally not be
reached as maximum fields are limited to about 1 T
in most laboratories. Sophisticated mathematical
algorithms allow to estimate the effect of partly
unsaturated hard magnetic constituents by the
approach to saturation analysis (von Dobeneck 1996).