FORC analysis offrozen ferrofluids

Spring 2001, Vol. 11, No. 1Institute for Rock Magnetism

Inside...Visiting Fellows’ Reports 2Current Abstracts 3Rock Magnetism Conference 7

Ferrofluids are a good example of aninteracting single-domain (SD) particlesystem. A ferrofluid consists of ~10-nmsized magnetite particles immersed in anoil carrier fluid. Surface tension keepsthe particles immersed in the fluid.Ferrofluids have several commercial, andeven pharmaceutical applications. Atroom temperature, ferrofluid particles aremobile and superparamagnetic. Belowliquid nitrogen temperature, the carrierfluid is frozen and the particles becomestable SD. For basic science, ferrofluidshave the advantage over other SDsystems that their concentration, andhence interaction strength, can becontrolled.

In recent years, we have beenstudying the hysteresis properties of anumber of magnetic systems, such asmagnetic recording media, permanentmagnetic materials (e.g. Alnico), andnatural samples. To investigate hysteresisproperties, we have employed a class ofhysteresis curves known as First OrderReversal Curves (FORCs, see below).To better visualize the information in aset of FORCs, we transform them into acontour plot known as a FORC diagram(described below). A FORC diagramcontains much more detailed informationabout hysteresis properties than themajor hysteresis loop or the remanencecurves. It can be used to probe subtlevariations in magnetic properties thatmight not be discernable with theconventional hysteresis measurements.

With regard to natural samples, ourgoal is to use FORC diagrams to obtaininformation on the distribution ofdomain states and particle sizes andmineral phases in natural samples, and togain insight into effects such as thermalrelaxation and interparticle interaction.

ChristopherPikeUniversity ofCalifornia-Davis

Jim MarvinIRM

Ferrofliuds have been used for magnetic domain imaging since the technique was pioneered byFrancis Bitter in the 1930’s. High-resolution images such as this (aluminum- and magnesium-substituted titanomagnetite; scale bar = 2.5 µm) are obtained by scanning electron microscopyon dried Bitter patterns.

Mo

sko

witz

, B

. M

., H

alg

ed

ah

l, S

. L

., a

nd

La

wso

n,

C.

A.,

19

88

, M

ag

ne

tic d

om

ain

son u

npolis

hed a

nd p

olis

hed s

urf

ace

s of

titaniu

m-r

ich t

itanom

agnetit

e:

Journ

al

of

Geophys

ical

Rese

arc

h B

: S

olid

Eart

h,

v. 9

3,

p.

3372–3386.

ferrofluids continued on page 10...

However, natural samples are notoriouslycomplicated systems, and the interpreta-tion of the FORC diagrams of naturalsamples is commensurately difficult.Therefore, we have tried to develop aninterpretational framework starting withmore simple and well-behaved samples,such as ferrofluids. We recentlyperformed low-temperature FORCdiagram measurements on a ferrofluid atthe IRM. Before describing our results,we will review the construction of aFORC diagram.

Let us begin with some definitions.The measurement of a First OrderReversal Curve (FORC) begins bysaturating a sample in a large positiveapplied field (see Figure 1). The field isdecreased to a reversal field Hr, and theFORC is the magnetization curve thatresults when the applied field isincreased back to saturation [1]. Byrepeating this measurement for differentvalues of Hr, one obtains a set of curvessuch as those shown in Figure 2(a). Themagnetization at the applied field Ha on

the FORC with reversal field Hr isdenoted by M(Hr, Ha), where Ha > Hr(see Figure 1). Each individual FORCreflects a combination of reversible (e.g.,small displacements of domain walls inan energy well) and irreversible magneti-zation changes (e.g., displacements ofwalls through an energy barrier into adifferent well). The difference betweensuccessive FORCs is a consequence ofirreversible changes that occur betweenthe successive reversal fields. TheFORC distribution, which characterizesin detail the distribution of critical fieldsfor irreversible magnetization changes, isdefined as the mixed second derivative:

Edito

r’s N

ote

: T

his

is t

he s

eco

nd in

a s

eries

of

art

icle

s on d

ata

analy

sis

and i

nte

rpre

tatio

n.

( ) ( )ar

arar HH

HHMHH

∂∂∂

ρ ,,

2

−≡

where ρ(Hr, Ha) is well defined for Ha >Hr. (The notation here is slightlydifferent from that used in our previouswork on FORC diagrams [2,3]).

(1)

Investigating the role ofmicrobial versus chemicalprocesses in magnetic min-eral formation in soils

Authigenesis of magnetic mineralscan occur via abiotic/inorganic as well asvarious biological processes in soils andsediments. For instance, inorganically-formed ultrafine-grained magnetite wasreported from soils in the UK (Maher &Taylor, 1988). The chemical conditionsprevailing in soils due to hydrocarbonexposure are also thought to be capableof causing the formation of magneticphases such as magnetite (Machel, 1996,Machel & Burton, 1991). Alternatively,several types of bacteria can causemagnetic mineral formation eitherinternally in magnetotactic microorgan-isms (e.g., Bazylinski et al., 1993) orexternally by iron-reducing bacteria (e.g.,Lovley, 1990, Stolz et al., 1990). Theformer have been studied in greater detailand the magnetic minerals they producecan be uniquely magnetically character-ized because of their specific grainshapes and size range (Bazylinski &Moskowitz, 1997, Moskowitz et al.,1993). In the case of iron-reducingbacteria, the magnetic mineral depositsare dominantly superparamagnetic (SP)magnetite (Moskowitz et al., 1989,Sparks et al., 1990) and are similar tothose formed inorganically in soils. Theresemblance between inorganically-formed magnetite and magnetic grainsformed by iron-reducing bacteria makesit difficult to determine the origin ofsome magnetic minerals in soils andsediments. Developing a better under-standing of microbially-formed magneticminerals and in particular of the role ofiron-reducing bacteria and the conditionsunder which they produce magneticminerals is important in order to evaluatetheir potential contribution to themagnetic signal in soils and geologicdeposits. This work investigatesbiologically-produced magnetic mineralswith focus on iron-reducing bacteria as apotentially important factor controllingmagnetic mineral formation in soils ascompared to chemical processes.

Two soil types were collected inOklahoma to investigate the formation ofmagnetic minerals under a variety oflaboratory-controlled conditions. Thefirst soil type was collected in twohorizontal cores retrieved at depths of 5-12 cm and 15-22 cm. The shallow coreconsists of a dark grayish-brown siltyclayloam that contains decaying plantmaterial. This decaying organic matter

slightly after prolonged periods of timebut remains above the initial χ

b. Several

biological samples showing increases inχ

b and their sterile controls were chosen

for further analyses at the IRM alongwith an original/untreated soil sample.The main tool for magnetic mineralcharacterization used at the IRM isthermal demagnetization of a saturationmagnetization acquired at low-tempera-ture. In addition, other methodsincluding hysteresis and Curie-balanceanalysis were used on some samples.

Samples from both cores and allgroups indicate a somewhat steep initialdrop in remanence followed by a moregradual drop upon heating to roomtemperature and a weakly developedVerwey transition. This is interpreted toindicate the presence of some ultrafine-grained (SP) and coarse (multi-domain)

Monika CogoiniUniversity of Oklahoma

Cogoini@ou.edu

Visiting Fellows’ Reports

2

Figure 1. Normalized χb for

samples of the shallow core: a) treatedwith methane; b) treated with hydrogen.Biological samples generally showelevated χ

b with time, whereas the

magnetic sterilized samples showrelatively constant χ

b or occasionally

slightly lower values. Standard devia-tions are commonly less than 5%.

appears to be the cause of a colorationchange to yellowish along thin layers(usually less than 1 mm) of the soilsurrounding the plant material. The soilcontained in the deep core is also a siltyclayloam, dark grayish- to greenish-brown, and contains less plant materialand therefore, there are fewer of theyellowish lenses in the soil. The coreswere collected above an aging oil fieldand the soil is known to have beenexposed to hydrocarbons. The secondsoil is an orange-to-brown clayey loamwith abundant ferric iron oxides, whichare dominantly hematite.

For the first soil, an array ofexperiments was set up to test formicrobial versus chemical causes for theobserved changes in the magnetic signal.Each experiment was conducted bydividing the soil into three:groups:natural samples, positive controlsamples, and sterile control samples. Thenatural samples contain the soil ascollected in the field. To obtainpositive controls, a culture of iron-reducing bacteria known to becapable of producing magnetitewas introduced into selectedsamples of the natural soil. Inorder to test for a chemicalprocess, samples of the natural soilwere sterilized in an autoclave. Atotal of five sets, each constitutingof triplicate (or 4) samples of eachof the three groups (natural,positive and sterile controls), wereprepared for treatment under fivedifferent environmental condi-tions. Each set was subjected toone of the following treatments:addition of methane, hydrogen,nitrogen, a mixture of methaneand oxygen, or oxygen. During thecourse of approximately 180 days,bulk magnetic susceptibility (χ

b)

measurements were conductedperiodically and elevation of themagnetic signal generally occurredin biological samples, i.e. naturaland positive control samples(Figure 1). The sterile sets showedno consistent increases in the χ

b

signal through time, and somedrop slightly below the initial χ

b

value (Figure 1a and b). Theaugmentation of χ

b occurs under

anaerobic environmental conditions inmost of the natural and positive controlsamples of the shallow core and in manybiological samples of the deeper core.The χ

b in natural samples from the

shallow core may increase by up to 35 %whereas the χ

b of the deep core samples

reaches up to a maximum of 24 %. Onlya few of the aerobically treated biologicalsamples (mostly positive control samplesof the shallow core) show slightlyelevated χ

b. In some samples that showed

initial increases in χb the signal decreases

VF Reportscontinued on page 6...

reduction. Nevertheless we were able todefine a stable characteristic remanence formost of the core.

Yamazaki, T., Solheid, P. A., and Frost, G. M.,2000, Rock magnetism of sediments in theAngola-Namibia upwelling system withspecial reference to loss of magnetizationafter core recovery: Earth, Planets andSpace, v. 52, no. 5, p. 329-36.These Leg 175 sediments have a high total-organic-carbon content and very lowconcentrations of ferrimagnetic minerals.Severe and rapid loss of remanent magnetiza-tion occurred during storage, with less than10% of the initial intensity remaining a after afew months. Decreases in S-ratio suggestpreferential loss of magnetite, and hysteresisparameters and ARM/SIRM ratios suggestpreferential dissolution of finer grains. Low-temperature magnetometry revealed thepresence of magnetite in the sediments evenafter the completion of sulfate reduction.

Anisotropy

Yaouancq, G., and MacLeod, C. J., 2000,Petrofabric investigation of gabbros fromthe Oman ophiolite: comparison betweenAMS and rock fabric: Marine GeophysicalResearches, v. 21, p. 3-4.From the paleo-Moho to the top of thefoliated gabbros level, 73% of the rocksdisplay a good correspondence in orientation,between the magnetic and rock fabricorientation. AMS is controlled by secondarymagnetites located in the fracture network ofthe olivines, and probably, but to a lesserextent, by secondary magnetites located in theexsolution lamellae of the clinopyroxenes.The magnetic foliation is constrained by theorientation of the olivine fracture planes,which is in turn constrained by the orientationof the overall magmatic rock fabric. Nocorrelation between the shape and magnitudeof the AMS and plagioclase fabrics can beestablished

Carriers and Origins of NRM

Turner, G. M., 2001, Toward an under-standing of the multicomponent magnetiza-tion of uplifted Neogene marine sedimentsin New Zealand: Journal of GeophysicalResearch, v. 106, no. B4, p. 6385-97.The NRM consists of three components: (1) alow (<150° C) T

b component interpreted to be

a recent TVRM; (2) an intermediate (150-250° C) T

b component, thought to be all that

remains of the primary DRM; and (3) a high(>250° C) T

b component that is considered to

be diagenetic in origin. Rock magnetic datawhich show that specimens carrying stronghigh T

b components have a distinctly different

spectrum of ferrimagnetic grains, specificallya population with much higher coercivitiesthan is seen in specimens which lack a highT

b component. It is argued that these high-

coercivity, high Tb grains are secondary in

origin, leading to the conclusion that theintermediate T

b component is the primary

detrital magnetization.

Alteration, Diagenesis andRemagnetization

Doh, S.-J., Suk, D.-W., and Kim, B.-Y., 1999,Paleomagnetic and rock magnetic studiesof Cretaceous rocks in the Eumsung Basin,Korea: Earth, Planets and Space, v. 51, no.5, p. 337-49.Early Cretaceous red beds and greenishmudstones carry a ChRM that was acquiredduring or after tilting during late Cretaceousto early Tertiary. Electron microscopeobservations and rock magnetic experimentsshow secondary authigenic hematite andmagnetite grains of SD-PSD size.

Elmore, R. D., Kelley, J., Evans, M., andLewchuk, M. T., 2001, Remagnetization andorogenic fluids: testing the hypothesis inthe central Appalachians: GeophysicalJournal International, v. 144, no. 3, p. 568-76.The Helderberg Group, an aquitard, and theOriskany Formation, a palaeoaquifer, bothcontain similar, synfolding Late PalaeozoicCRMs residing in magnetite. The OriskanyFormation may have been remagnetized byorogenic fluids, because there is geochemicalevidence for the migration of such fluids. Incontrast, there is no geochemical evidence forthe widespread migration of fluids into theHelderberg Group and thus anotherremagnetization mechanism, which remainselusive, must be responsible for theHelderberg CRM.

Lewchuk, M. T., Al-Aasm, I. S., Symons, D.T. A., and Gillen, K. P., 2000, Late Laramidedolomite recrystallization of the HuskyRainbow “A” hydrocarbon Devonianreservoir, northwestern Alberta, Canada:paleomagnetic and geochemical evidence:Canadian Journal of Earth Sciences, v. 37,no. 1, p. 17-29.Middle Devonian carbonates here contain awell-defined ChRM, residing in SD to PSDmagnetite, of Middle Eocene to MiddleMiocene age. Four generations of dolomiteare recognized, and all have similar isotopiccompositions (δ18O of -10.7 to -16.5% (PDB),δ13C of +0.7 to +3.2% (PDB), and Sr ratios of0.70826 to 0.70846) that do not matchDevonian carbonate or seawater values. Thesedata indicate that mixed pre-Laramide basinalfluids, heated by burial during the LaramideOrogeny, were present when the dolomiteswere recrystallized and (or) precipitated priorto petroleum migration and accumulation inthe reservoir.

Richter, C., Hayashida, A., Guyodo, Y., Valet,J. P., and Verosub, K. L., 1999, Magneticintensity loss and core diagenesis in long-core samples from the East Cortez Basinand the San Nicolas Basin (CaliforniaBorderland): Earth, Planets and Space, v.51, no. 5, p. 329-36.High-resolution (1-2 cm interval) magneticmeasurements made on about 180 m ofcontinuous U-channel samples from two Leg167 cores showed that by ca. 4-6 months afterthe cores were drilled , the magnetic intensityhad decreased to 7%-10% of the initialmagnetization. Both cores were stronglyaffected by the decay of a metastablemagnetic mineral or the dissolution of fine-grained magnetite by organic matter

Current AbstractsA list of current research articlesdealing with various topics in thephysics and chemistry of magnetism isa regular feature of the IRM Quar-terly. Articles published in familiargeology and geophysics journals areincluded; special emphasis is given tocurrent articles from physics, chemis-try, and materials-science journals.Most abstracts are culled fromINSPEC (© Institution of ElectricalEngineers), Geophysical Abstracts inPress (© American GeophysicalUnion), and The Earth and PlanetaryExpress (© Elsevier Science Publish-ers, B.V.), after which they aresubjected to Procrustean editing andcondensation for this newsletter. Anextensive reference list of articles(primarily about rock magnetism, thephysics and chemistry of magnetism,and some paleomagnetism) is continu-ally updated at the IRM. This list,with more than 5200 references, isavailable free of charge. Yourcontributions both to the list and tothe Abstracts section of the IRMQuarterly are always welcome.

Guests in the home of Arthur Ewing were invited to sketch a pig whileblindfolded. Among the “Pig-Book” sketches are the works of: (A)Arthur Eddington, astronomer and physicist known for his work on thestructure and evolution of stars and relativity; (B) Alexander GrahamBell, inventor of the telephone; (C) J.J Thomson, discoverer of theelectron and winner of the 1906 Nobel prize for his work on electricalconductivity of gases; (D) William Bragg, who shared the 1916 Nobelprize with his son Lawrence, for their work on x-ray diffraction andcrystal structures; (E) David Lloyd George, British Prime Minister from1916-1922; (F) Winston Churchill, Prime Minister from 1940-45 and1951-55.

abstractscontinued on page 4...

3

sourc

e: T

he M

an o

f R

oom

40:

The

Life

of

Sir

Alfr

ed E

win

g,

by

A.

W.

Ew

ing

,1939,

Hutc

hin

son &

Co.

Ltd

., L

ondon.

Yoshida, M., Khan, I. H., and Ahmad, M. N.,1998, Remanent magnetization of ooliticironstone beds, Hazara area, LesserHimalayan thrust zone, northern Pakistan:its acquisition, timing, andpaleoenvironmental implications: Earth,Planets and Space, v. 50, no. 9, p. 733-44.Chamosite-hematite type oolitic ironstone isdistributed in the Cretaceous-Tertiary (K-T)boundary zone. Ferromagnetic properties aredominated by fine-grained hematite thatformed from iron hydroxides during earlydiagenesis. The ChRM direction indicatesthat the magnetization was acquired duringEarly Paleocene time. The initial Fe-richsediments were deposited under anoxicconditions, probably in a non-marine orbrackish environment. During earlydiagenesis, after development of oolitictextures, the paleoenvironment became arid,and the CRM was acquired through theconversion of amorphous hydroxides orgoethite to fine-grained hematite.

Data Processing and Analysis

Oda, H., and Shibuya, H., 1998, Animprovement in ABIC-minimizingdeconvolution for continuously measuredmagnetic remanence data: Earth, Planetsand Space, v. 50, no. 1, p. 15-22.Deconvolution of magnetic remanence whichhas been continuously measured along a long-core sample is improved by eliminating thedisturbance of severe intensity variationconfined to a narrow layer. The deconvolutionis constrained by the smoothness of themagnetization measured by the L

2-norm of

second order difference. The previous modeluses a single smoothness parameter for theentire sample and the optimum smoothness isobtained by minimizing ABIC (Akaike’sBayesian information criterion), while anadditional parameter is introduced in themodified scheme for a layer where intensityfluctuation is strong. The modifieddeconvolution scheme was applied to 5-mminterval magnetic remanence data for twosections of ODP Hole 767B. We assumedanother smoothness for the uppermost 1.5 cmof Section 767B-6H1 where the coringdisturbance is severe. For Section 767B-6H2,the Australasian microtektite layer (108-113cm), where the magnetization intensity showsa strong peak, is expressed by a differentsmoothness parameter from other part. In bothcases the minimum ABIC was lowered andthe calculated error was reduced, demonstrat-ing the improvement of the model byincorporating additional information otherthan magnetic remanence data in the priordistribution of the data.

Tsunakawa, H., Okada, M., and Niitsuma, N.,1999, Further application of thedeconvolution method of post-depositionalDRM to the precise record of theMatuyama-Brunhes reversal in thesediments from the Boso Peninsula, Japan:Earth, Planets and Space, v. 51, no. 3, p. 169-73.The Matuyama-Brunhes geomagnetic reversalwas continuously recorded in massivesiltstones with a time resolution better than 9years. However the Boso sediments probablyhave post-depositional detrital remanentmagnetization (pDRM) depressing short-4

period variations due to the convolution of thegeomagnetic field. The previous applicationof the deconvolution method of pDRM to the900 year records gave a half fixing depth of21 cm. Further application of this fixing depthto the 4100 year record clearly shows nearly100 year variations in both of the inclinationand declination.

Utada, H., Neki, M., and Kagiyama, T., 2000,A study of annual variations in thegeomagnetic total intensity with specialattention to detecting volcanomagneticsignals: Earth, Planets and Space, v. 52, no.2, p. 91-103.A model is proposed in which changes inmagnetization of near-surface rocks due toseasonal changes in atmospheric temperatureproduce annual variations in geomagnetictotal intensity. First, amplitude and phasedifference of annual variations in the totalintensity and ground temperature data weredetermined by time series analyses.Considering thermal diffusion from thesurface into the ground, the phase differencewas converted to a characteristic depth, andthen the amplitude of annual temperaturevariation at the depth was estimated. Finally,the observed total intensity variations werecompared with the expected change on thebasis of the temperature dependence of therock magnetization measured in thelaboratory, and good agreement was found.Correction of annual variations by usingtemperature data will enable more accuratedetection of volcanomagnetic signals.

Environmental Magnetism

Abdul-Razzaq, W., and Gautam, M., 2001,Discovery of magnetite in the exhaustedmaterial from a diesel engine: AppliedPhysics Letters, v. 78, no. 14, p. 2018-19.Magnetite was detected in the particulatematter collected from diesel engine exhaustusing a total exhaust dilution tunnel. Thisdiscovery is important in terms of the healtheffects of exposure to magnetite or itsinteraction with static magnetic fields or low-frequency electromagnetic fields. Magnetite isthe best absorber of microwave radiation ofany biological material in the 0.5-10 GHzfrequency range through the process ofelectromagnetic resonance. This includes thefrequencies that are normally used in thecellular telephone industry.

Evans, M. E., 2001, Magnetoclimatology ofaeolian sediments: Geophysical JournalInternational, v. 144, no. 2, p. 495-7.In the classic sites of the Chinese LoessPlateau, soil formation during interglacialsleads to magnetic enhancement in thepalaeosols. In other regions (Alaska, Siberia),magnetic susceptibility peaks during glacialintervals. A simple qualitative analysis basedon density sorting determines the regimesunder which one or the other dominates.Published data permit a certain degree ofquantification, and lead to the conclusion thatproximal sites are likely to manifest Alaskan/Siberian-type magnetoclimatological patterns.Beyond a few hundred kilometres, however,distal sites will be dominated by Chinese-typepatterns.

Fukuma, K., and Torii, M., 1998, Variable

shape of magnetic hysteresis loops in theChinese loess-paleosol sequence: Earth,Planets and Space, v. 50, no. 1, p. 9-14.The shape of magnetic hysteresis loops varieswith susceptibility χ. When the ratio of theferrimagnetic (FM) to antiferromagnetic(AFM) contribution is relatively low (low χ),the broad loop is controlled by lithogenic FMand AFM minerals. For samples withintermediate χ, constricted loops originatefrom the addition of a broad loop from thelithogenic fraction and a narrow loop from apedogenic fraction with high superparamag-netic content. Then with further χ increase,the constricted shape almost disappears andthe loop is dominated by the pedogenicfraction.

Heslop, D., Langereis, C. G., and Dekkers, M.J., 2000, A new astronomical timescale forthe loess deposits of Northern China: Earthand Planetary Science Letters, v. 184, no. 1,p. 125-39.A refined timescale for the entire sequence ofQuaternary Chinese loess relies upon thecorrelation of detailed monsoon records to theastronomical solution of Laskar (1990) andthe oceanic ODP677 δ18O record ofShackleton et al. (1990). The chronologicalscheme considers in detail the relativestructures of the palaeoclimatic andpalaeomagnetic records to produce anaccurate timescale that is consistent with thecurrent understanding of loess depositionaland post-depositional processes. Analysis ofthis chronological framework demonstratesdownward displacement of thepalaeomagnetic horizons with respect to theclimatic record.

Extraterrestrial Magnetism

Gunnlaugsson, H. P., 2000, Analysis of themagnetic properties experiment data onMars: results from Mars Pathfinder :Planetary and Space Science, v. 48, no. 15, p.1491-504.Spectral data for the material adhering to themagnets, as well as the time dependence ofmagnetic dust accumulation, are the basis of anew model of the magnetic properties of thedust. Optical reflection spectra of themagnetic dust differ slightly from thereflection of ground soil/dust in the vicinity ofthe lander. Magnetic dust has been removedfrom the magnets at least twice during the 80sol mission by wind gusts, in a manner similarto sand blasting. A lower limit for thesaturation magnetisation of the most magneticparticles is found to be 5 Am2/kg.

Kletetschka, G., Wasilewski, P. J., and Taylor,P. T., 2000, Mineralogy of the sources formagnetic anomalies on Mars: Meteoritics &Planetary Science, v. 35, no. 5, p. 895-9.The intense magnetic anomalies on Mars arelikely due to TRM and possible carriersinclude SD and MD magnetite, hematite, andpyrrhotite. The intensity of TRM (in 0.05 mT)is in descending order: SD magnetite, SDpyrrhotite, MD hematite, MD pyrrhotite, MDmagnetite, SD hematite. Each of theseminerals is used in a thin sheet model toestimate the concentrations required togenerate the observed magnetic anomaly(1500 nT at 100 km altitude) assuming TRMacquisition in a 0.05 mT magnetic field.

separated from basalts of the Rajmahal Trapsof India that formed during the CretaceousNormal Polarity Superchron. These datasuggest a time-averaged paleomagnetic dipolemoment of 12.5±1.4*1022 Am2, three timesgreater than mean Cenozoic and EarlyCretaceous-Late Jurassic dipole momentswhen geomagnetic reversals were frequent.This result supports a correlation betweenintervals of low reversal frequency and highgeomagnetic field strength.

Magnetization Processes

Katari, K., and Bloxham, J., 2001, Effects ofsediment aggregate size on DRM intensity:a new theory: Earth and Planetary ScienceLetters, v. 186, no. 1, p. 113-22.Inter-particle attractions due to electrostatic orvan der Waals forces, or biologically mediatedflocculation, lead to the formation ofaggregates, which prevent the settling ofindividual magnetic grains. Viscous dragbecomes important for large clay-magnetiteaggregates, which can be tens of micrometersin diameter, and larger magnetic fields arenecessary to bring their moments intoalignment. A new model of DRM acquisitionassumes a log-normal size distribution offlocs, all of which have the same magneticmoment. The undersaturation of magneticintensity in laboratory DRM experiments isexplained by the larger viscous dragencountered by the larger aggregates whichprevents perfect alignment with the appliedmagnetic field.

Kosterov, A., 2001, Magnetic hysteresis ofpseudo-single-domain and multidomainmagnetite below the Verwey transition:Earth and Planetary Science Letters, v. 186,no. 2, p. 245-54.Hysteresis loops have been measured as afunction of temperature between 10 K androom temperature for two size fractions ofcrushed magnetite: 1-5 µm (pseudo-single-domain, PSD) and 100-150 µm (multidomain,MD). The material under study appears to befairly close to stoichiometric magnetite, witha Verwey transition temperature of 115 K forthe MD sample and 118 K for the PSDsample. For both samples, hysteresisproperties below the Verwey transition dependcritically on whether or not the samples werecooled through the transition in the magneticfield. For the PSD sample, an even morecomplex behavior has been observed:temperature dependences of coercive forceobtained after zero-field-cooling differsignificantly for the demagnetized versusmagnetized starting magnetic state.

Maksimochkin, V. I., 2000, The pressureeffect on the magnetization of basalts:Izvestiya, Physics of the Solid Earth, v. 36,no. 7, p. 565-70.The effect of pressure application time on theformation of the pressure-induced viscousremanent magnetization (PVRM) in basalts isstudied. The PVRM at P=160 MPa is shownto be 1.8 to 7.3 times higher than the viscousremanent magnetization. The pressure effecton the PVRM process in basalts with Z=0.4-0.75 single-phase oxidation oftitanomagnetites is 2 to 4 times higher thanthe similar effect in titanomagnetites thatwere not oxidized or experienced high- 5

Taylor, A. P., Barry, J. C., and Webb, R. I.,2001, Structural and morphologicalanomalies in magnetosomes: possiblebiogenic origin for magnetite inALH84001: Journal of Microscopy, v. 2720,no. 2001, p. 0022-2720.We report biogenic magnetite whiskersresembling the magnetite whiskers detected inthe Martian meteorite ALH84001 by Bradleyet al., and interpreted by those authors asevidence of vapour-phase (abiogenic) growth.The structures of the magnetite particles inALH84001, their spatial arrangement andcoprecipitation with carbonates and proximityto silicates are all consistent with beingbiogenic. Electron-beam-induced flash-melting of magnetosomes produced numerousscrew dislocations in the {111}, {100} and{110} lattice planes and induced fusion ofplatelets. From this, the lack of screwdislocations reported in the magnetiteparticles in ALH84001 (McKay et al., andBradley et al.) indicates that they have a low-temperature origin.

Instruments and Measurements

Leslie, K. E., Binks, R. A., Lewis, C. J., Scott,M. D., Tilbrook, D. L., and Du, J., 2001,Three component spinner magnetometerfeaturing rapid measurement times: IEEETransactions on Applied Superconductivity, v.11, no. 1, p. 252-5.The remanent magnetisation of rockspecimens is measured using a hightemperature superconductor rf SQUID.Samples are rotated around two orthogonalaxes to facilitate the calculation of the three-component remanence vector with aminimum of operator intervention. Measure-ments of remanence ranging from 101 to 10-5

A/m, with errors of 1*10-5 A/m, have beenachieved in measurement times of 20 s, in atypical geophysical laboratory environment.

Magnetic Field Records andPaleointensity Methods

Harrison, C. G. A., 2001, Resource Letter:G-1: Geomagnetism: American Journal ofPhysics, v. 69, no. 5, p. 534-42.This Resource Letter provides a guide to theliterature on the internally generatedgeomagnetic field. The topics covered includethe origin of the geomagnetic field and theenergy requirements and core propertiesnecessary for its generation; a description ofthe geomagnetic field, including its sphericalharmonic representation, and its rates ofchange, as well as models of the secularvariation; observations of the field relevant topaleomagnetism and continental drift;information about magnetic anomalies in theoceans, sea floor spreading and platetectonics; reversals of the geomagnetic fieldand magnetochronology.

Tarduno, J. A., Cottrell, R. D., and Smirnov,A. V., 2001, High geomagnetic intensityduring the mid-Cretaceous from Thellieranalyses of single plagioclase crystals:Science, v. 291, no. 5509, p. 1779-83.The authors report data from 56 Thellier-Thellier experiments on plagioclase crystals

temperature multi-phase oxidation.

Spinu, L., Srikanth, H., Wiemann, J. A., Li,S., Tang, J., and O’Connor, C. J., 2000,Superparamagnetism and transversesusceptibility in magnetic nanoparticlesystems: IEEE Transactions on Magnetics, v.36, no. 5, p. 3032-4.The temperature and field dependence of thetransverse susceptibility (χ

T) of a magnetic

nanoparticle system has been investigated.The T- dependence is found to exhibitcharacteristic features that can be related torelaxation behavior in nanoscale magneticsystems. The experimental data are inter-preted using a modified coherent magnetiza-tion rotation model based on a two levelapproach.

Stinnett, S. M., and Doyle, W. D., 2000,Relationship between the intrinsic thermalswitching field at 10-9 s and the anisotropyfield in magnetic recording media: IEEETransactions on Magnetics, v. 36, no. 5, p.2456-8.New measurements of the anisotropy fieldfrom rotational hysteresis are compared withprevious measurements of the intrinsicswitching field in a variety of particulate andthin film media. All of the particulate mediahave intrinsic switching fields of H

0~H

K/3

independent of particle size or orientation. Incontrast to the particulate samples, the thinfilm samples have a larger ratio of H

0/H

K from

0.5 to 0.8 possibly indicating a change inreversal mode from incoherent to coherentrotation. Further, results on γ-Fe

2O

3

particulate samples confirm that below a fewnanoseconds, thermal switching ceases todrive the increase in H

CR(t).

Modeling and Theory

Muxworthy, A. R., 2001, Effect of graininteractions on the frequency dependenceof magnetic susceptibility: GeophysicalJournal International, v. 144, no. 2, p. 441-7.Magnetic susceptibility and κ

fd have been

modelled for weakly interacting assemblagesof SD magnetite grains, near the SP - stableSD threshold known as the blocking volume,v

b. Weak interactions between SP grains

effectively increase the anisotropy, whichreduces v

b, causing a decrease in the peak

values of κfd, by over 50 per cent for certain

grain distributions. However, it is also foundthat the reduction of v

b as a result of

interactions can actually increase κfd for

certain grain distributions.

Shcherbakov, V. P., Lamash, B. E., andSycheva, N. K., 2000, The magneticsusceptibility and viscous magnetization ofinteracting single-domain grains: Izvestiya,Physics of the Solid Earth, v. 36, no. 5, p.430-6.Numerical calculations based on Boltzmannstatistics for an ensemble of single-domain(SD) particles, with volume concentrationranging from c = 0 to 5%, show thatmagnetostatic interaction has virtually has noeffect on the equilibrium value of thesusceptibility. The calculations also show thatinteraction significantly affects viscousbehavior at c≥0.5%, due to the drasticbroadening of the relaxation time spectrum as

abstracts continued on page 8...

magnetite (Figure 2). A comparison ofthe low-temperature analysis of a naturalsample that was exposed to methane andwhich was later subjected to additionaltreatment in the laboratory to enhance the

growth of iron-reducing bacteria isshown in Figure 2 (inset). The thermaldemagnetization curves reveal a generalshift towards higher remanence at roomtemperature, an increased drop ofremanence upon heating through theVerwey transition, and a slight shift ofremanence loss towards higher tempera-tures after the additional treatment. Theseresults could indicate growth and/orfurther production of magnetite in thesoil after enhancing for iron-reducingbacteria. If growth of the magnetiteoccurred, this could also explain theobserved decrease in the χ

b measure-

ments after prolonged times in some ofthe biological samples since coarsermagnetite has lower χ

b than SP grains.

Hysteresis measurements on a hydrogenexposed natural and a sterile controlsample indicate that the ferrimagnetic χis responsible for most of the χ

b signal

(99.9% and 99.2% of χb for the natural

and sterile sample, respectively). AllCurie-balance measurements had verylow signal to noise ratios due to the smallsample sizes and relatively low concen-trations of magnetic material. Noadditional magnetic phases wereidentified within the soil by any of themethods used. The rock-magnetic resultsindicate that small amounts of magnetiteof various grain sizes were originallypresent in the soil and magnetitecontinues to grow in the biological, i.e.natural and positive control samples. Theresults also show that chemical processesalone are not capable of furthering theproduction of magnetic minerals underthe conditions provided in the experi-ments.

One of the remaining unresolvedissues is: why are the differences noticedmainly in the coarse magnetite ratherthan the SP fraction despite the increasesin the χ

b? Iron-reducing bacteria are

known to produce SP-sized magnetite. Isit possible that the bacterial magnetiterapidly aggregates to coarser magnetiteunder favorable conditions for growth ofiron-reducing bacteria – or are theconditions prevailing in the microcosmsfavorable for inorganic growth of themagnetite once a nucleus has formed viabiologic pathways?

The second soil type that wasinvestigated is an orange-brown clayeyloam with abundant ferric iron oxides,which are dominantly hematite. This soilwas initially treated in the laboratory toenhance the amount of iron-reducingbacteria present within the soil and theexperiment was monitored by χ

b

measurements. The results showed rapidinitial increases in the χ

b signal, which

tripled within less than 80 days, and aconcurrent color change from orange-brown to black suggesting the presenceof magnetite. However, further analysesof the magnetic mineralogy at the IRM

using thermal demagnetization of asaturation remanence acquired at low-temperature, hysteresis measurements,Curie-balance analysis, pARM, andMössbauer spectroscopy showedinconclusive results. The low-tempera-ture thermal decay pattern of a treatednatural sample was compared to that of aferric oxy-hydroxide sample, the materialadded to the soil in order to enhance thegrowth of iron-reducing bacteria (Figure3). The low-temperature curve of theferric oxy-hydroxide sample shows avery steep initial drop in remanence andit retains no remanence at temperaturesabove approximately 15°K. The naturalsoil sample initially loses its remanencerapidly, but it carries a small amount ofremanence up to higher temperatures.Similar to the ferric oxy-hydroxidesample, the natural sample has almost nomagnetic remanence at room tempera-ture. These results may indicate thepresence of extremely ultrafine-grainedmagnetite (few nm) in the natural soil,the formation of which may have beencaused by iron-reducing bacteria.However, it is also possible that some ofthe ferric oxy-hydroxide that is used toenhance the growth of iron-reducingbacteria changed to a ferric ironcomponent of different crystallinityduring the course of the experimentwhich may show somewhat differentdemagnetization patterns. All othermethods used in an attempt to decipherthe magnetic mineralogy within the soilwere unconvincing. Additional experi-ments on the clayey loam to test for amicrobial versus chemical process ofmagnetic mineral formation were set upin the laboratory prior to my visit at theIRM. These experiments were inconclu-sive with regard to the χ

b behavior of

most samples, which is probably due tothe low initial values inherent in this soiltype. Further investigations on this soiltype were not conducted because of thedifficulties encountered in determiningthe magnetic mineralogy and the weakinitial χ

b signal. The observed enhance-

ment in the χb in the initial experiment is

probably due to the fact that ferric ironwas introduced to the soil in a readilybioavailable form, i.e. the ferric iron inthe compound is easily utilized bymicroorganisms such as iron-reducingbacteria. Since the natural soil containsmainly hematite, a poorly bioavailableferric iron compound of high crystallin-ity, magnetic mineral formation could beeither very slow or not occurring evenunder conditions that otherwise couldsustain the growth of iron-reducingbacteria. Based on the results presented,it appears that the first soil type de-scribed above is favorable for biologicalproduction of magnetic minerals,probably via iron-reducing bacteria. Thepresence of abundant organic material (in

Figure 2. Thermal demagnetization of low-temperature SIRMs.Inset: comparison of a hydrogen-treated natural and a sterilecontrol sample shows minor differences, and both suggest thepresence of some SP grains and small amounts of MD magnetitein the original soil. The sterilized sample shows only a veryweakly developed Verwey transition, which can be recognizedmore easily in the first derivative (not shown here). The other twocurves represent a methane-treated natural sample and amethane-treated sample after enhancing the amount of iron-reducing bacteria present. The additional treatment of the samplepossibly resulted in a coarsening of the magnetite grains present.

Figure 3. Thermal demagnetization of low-temperature SIRMsfor an iron-gel sample (ferric oxy-hydroxide) and a naturalsample treated with the iron-gel to enhance growth of iron-reducing bacteria present in the soil. Although the SIRMs wereacquired at slightly different temperatures, the demagnetizationpatterns are similar in having steep initial drops. The iron-gelsample loses its remanence within a temperature range of 10 -15°. The natural sample carries some remanence up to at least200K. The additional remanence-carrying material present inthe soil could be ultrafine-grained magnetite.

...VF Reportscontinued from page 2

6

this case plant material and hydrocar-bons) and bioavailable ferric iron areprobably the key factors for enablingmicrobial magnetic mineral formation.The additional growth of magnetitegrains could be caused either by bacteriaor an inorganic process, which probablyrequired the presence of previouslyformed magnetite.

My visit at the IRM was of greatvalue from a scientific as well as from apersonal viewpoint. I am very gratefulfor the opportunity of visiting such anexceptional place. I wish to express mysincere appreciation to all members ofthe IRM for their support not onlyconcerning the technical aspects but alsofor the many interesting discussions andhelpful suggestions. And at last, but notthe least, I would like to acknowledgeAnn Hirt for her friendship and for themany valuable and inspiring discussionswe had during the long working nightsspent at the institute.

References

Bazylinski, D.A., and Moskowitz, B.M.,1997, Microbial biomineralization of

magnetic iron minerals: Reviews inMineralogy, v. 35, p. 181-223.

Bazylinski, D.A., Heywood, B.R., Mann,S., and Frankel, R.B., 1993, Fe3O4and Fe3S4 in a bacterium: Nature, v.366, p. 218.

Lovley, D.R., 1990, Magnetite formationduring microbial dissimilatory ironreduction, in Frankel, R.B. andBlakemore, R.P., eds., IronBiominerals: New York, PlenumPress, p. 151-166.

Machel, H.G., 1996, Magnetic contrastsas a result of hydrocarbon seepageand migration, in Schumacher, D.and Abrams, M.A., eds., Hydrocar-bon migration and its near-surfaceexpression: AAPG Memoir, v. 66, p.99-109.

Machel, H.G., and Burton, E.A., 1991,Cause and spatial distribution ofanomalous magnetization inhydrocarbon seepage environments:American Association of PetroleumGeologists Bulletin, v. 75, p. 1864-1876.

Maher, B.A., and Taylor, R.M., 1988,Formation of ultrafine-grainedmagnetite in soils: Nature, v. 336, p.

368-370.Moskowitz, B.M., Frankel, R.B., and

Bazylinski, D.A., 1993, Rockmagnetic criteria for the detection ofbiogenic magnetite: Earth andPlanetary Science Letters, v. 120, p.283-300.

Moskowitz, B.M., Frankel, R.B.,Bazylinski, D.A., Jannasch, H.W.,and Lovley, D.R., 1989, A compari-son of magnetite particles producedanaerobically by magnetotactic anddissimilatory iron-reducing bacteria:Geophysical Research Letters, v. 16,p. 665-668.

Sparks, N.H.C., Mann, S., Bazylinski,D.A., Lovley, D.R., Jannasch, H.W.,and Frankel, R.B., 1990, Structureand morphology of magnetiteanaerobically-produced by a marinemagnetotactic bacterium and adissimilatory iron-reducing bacte-rium: Earth and Planetary ScienceLetters, v. 98, p. 14-22.

Stolz, J.F., Lovley, D.R., and Haggerty,S.E., 1990, Biogenic magnetite andthe magnetization of sediments:Journal of Geophysical Research, v.95, p. 4355-4361.

Focus: This conference is a jointinitiative of the "Mag-Net" EuropeanNetwork for Mineral Magnetic Studies ofEnvironmental Problems and theInstitute for Rock Magnetism. Updatedreviews, new observations on nano-phaseto submicron magnetic particles andexamples of successful inter-disciplinaryproblem-solving will be given bykeynote speakers, together with posterpresentations of current research byparticipating scientists. Significant timewill be kept for open discussion on datainterpretation, supported by short oralinput from participants.

Location: The Conference will be held atthe "Ettore Majorana" Foundation andCentre for Scientific Culture (EMFCSC,http://www.ccsem.infn.it/) in the historicand artistic town of Erice, Italy, locatedat the western end of Sicily. Participantsshould arrive on Wednesday, June 26,and depart in the afternoon of Monday,July 1. The cost of the conference isUS$500. This includes all food andlodging for 5 days and nights, meetingcosts (e.g. registration, abstract book,ice-breaker, coffee breaks, conferencedinner), transportation to/from Palermoairport and one field excursion.

Participation: There is space for amaximum of 100 participants and 20accompanying persons, with priority tothose who register early and who plan tostay for all conference sessions. Weespecially encourage application of PhDstudents and young (< 36 years)researchers. The objective is to have 50%of PhD students and young researchersand 25% of invited senior speakers. Thecost of the conference may be partlycovered with external funding (i.e. EC,NSFS.) with priority to young research-ers.

Applications: Obtain forms from PierreRochette (rochette@cerege.fr) or SilviaNardi (nardi_s@ingv.it).

Deadlines: - response to first circular bySeptember 30, 2001 - selection ofconference applicants if required (byNovember 1, 2001) - submission ofextended abstracts by December 31,2001 (maximum 2 pages and 1 figure) tobe published in a book of abstracts editedby the Istituto Nazionale di Geofisica eVulcanologia, Rome, Italy.

FUNDAMENTAL ROCK MAGNETISMAND ENVIRONMENTAL APPLICATIONSCONFERENCEErice, Italy - 26 June to 1 July 2002

http://www.sicily.infcom.it/uk/localita/TP/erice/fotografie/castelli/erice3.html 7

compared with the case of noninteractingparticles.

Oceanic Crust Magnetism

Nazarova, K. A., Wasilewski, P. J., and Dick,H. J. B., 2000, Magnetic study ofserpentinized harzburgites from the IslasOrcadas Fracture Zone: Marine Geophysi-cal Researches, v. 21, no. 5, p. 475-88.Magnetite in these serpentinites is variablymaghemitized, shown by optical microscopy,thermomagnetic analyses and low temperaturecycling of SIRM. Hysteresis ratios fall in arestricted range regardless of coercivity. Theapparent grain size configured in a threedimensional vein network plus maghemitiza-tion might be responsible for this observation.Maghemitization does not affect thermalmagnetic stability and enhances thegeophysical importance of remanence inserpentinites, which may be very importantfor generation of long wavelength aeromag-netic and possibly even satellite magneticanomalies.

Zhou, W., Van der Voo, R., Peacor, D. R.,Wang, D., and Zhang, Y., 2001, Low-temperature oxidation in MORB oftitanomagnetite to titanomaghemite: Agradual process with implications formarine magnetic anomaly amplitudes:Journal of Geophysical Research, v. 106, no.B4, p. 6409-21.Electron microscopic and rock magneticstudies indicate that the degree of oxidation oftitanomagnetite in MORB increases onlygradually with sample age. The envelope ofdata in a z versus age plot can be representedby z=p+q log t, where p ≈0.38, q ≈0.38, and tis in millions of years (for t>100 ka). The rateof maghemitization is controlled by manyfactors on both macroscopic and microscopicscales, including regional oceanic crustalstructures, lithological features, grain size,and surrounding matrices. The NRM intensityof MORB varies as a function of age,concentration and grain size of magneticmaterials, and the degree of alteration.

Paleomagnetism & Tectonics

Tarduno, J. A., and Smirnov, A. V., 2001,Stability of the Earth with respect to thespin axis for the last 130 million years:Earth and Planetary Science Letters, v. 184,no. 2, p. 549-53.Recently continental paleomagnetic datarotated into a hotspot reference frame havebeen used to suggest the Earth rotated 18°with respect to the spin axis at 110 Ma. Theauthors test this true polar wander (TPW)hypothesis using paleomagnetic data fromwidespread granitic rocks of cratonal NorthAmerica. These data conflict with TPWpredictions leading to rejection of thehypothesis. The calculated polar displacementis instead consistent with motion of theAtlantic hotspots during the mid-Cretaceous.This analysis suggests that the time-averagedposition of the spin axis has deviated by nomore than 5° over the last 130 million yr,indicating that mantle mass heterogeneities

have not changed rapidly enough to driveTPW.

Synthesis and Properties ofMagnetic Minerals

Ayub, I., Berry, F. J., Bilsborrow, R. L.,Helgason, O., Mercader, R. C., Moore, E. A.,Stewart, S. J., and Wynn, P. G., 2001,Influence of zinc doping on the structuraland magnetic properties of ααααα-Fe2O3:Journal of Solid State Chemistry, v. 156, no.2, p. 408-14.Zinc-doped α-Fe

2O

3 has been prepared by

hydrothermal methods. The zinc K-edgeEXAFS and interatomic potential calculationsindicate that Zn2+ ions substitute for Fe3+ ionsin the lattice and are accompanied by oxygenvacancies. The 57Fe Mossbauer spectrarecorded between 296 and 18 K demonstratethat the Morin transition in α-Fe

2O

3, is

completely suppressed within this temperaturerange by the presence of zinc. The 57FeMossbauer spectra recorded in situ attemperatures between 300 and 915 K showthe material to undergo conversion to thespinel-related ZnFe

2O

4 phase in this

temperature regime.

de Boer, C. B., and Dekkers, M. J., 2001,Unusual thermomagnetic behaviour ofhaematites: neoformation of a highlymagnetic spinel phase on heating in air:Geophysical Journal International, v. 144,no. 2, p. 481-94.A magnetic phase with T

C~475° C is formed

during routine thermomagnetic analysis ofvarious haematites. The formation tempera-ture can be as low as 400° C for syntheticsamples, whereas temperatures of 700-800° Care required for natural samples. The newphase has a cubic spinel structure with unitcell a

0=0.8350±0.0005 nm, similar to pure

maghemite. The low Tc of this particular

maghemite variety suggests that the vacancy(and/or cation) ordering is different fromtypical maghemite; a cation-deficientstructure with part of the vacancies ontetrahedral sites is suggested. Thermallyactivated release of incorporated hydroxylgroups would trigger the formation ofmaghemite traces on the surface of well-crystalline haematite planes. The formation oftraces of this highly magnetic mineral duringroutine stepwise thermal demagnetization orduring annealing haematite at high tempera-tures may seriously affect NRM measure-ments or may be erroneously taken ashaematite’s defect moment.

Fan, R., Chen, X. H., Gui, Z., Liu, L., andChen, Z. Y., 2001, A new simple hydrother-mal preparation of nanocrystallinemagnetite Fe3O4: Materials ResearchBulletin, v. 36, no. 3, p. 497-502.Nanocrystalline magnetite particles have beenprepared by hydrothermal reaction of ferroussulfate (FeSO

4), sodium hydroxide and

sodium thiosulfate (Na2S

2O

3) at 140° C. The

X-ray powder diffraction pattern indicatesthat the product is single-phase magnetite inan inverse cubic spinel structure. Itscomposition is determined by traditionalchemical analysis combined with X-ray

photoelectron spectroscopy. Transmissionelectron microscopy images show that themagnetite particles are homogenous in theshape of quasi-sphere polyhedra with anaverage diameter of 50 nm.

Garcia, J., Subias, G., Proietti, M. G., Blasco,J., Renevier, H., Hodeau, J. L., and Joly, Y.,2001, Absence of charge ordering belowthe Verwey transition temperature inmagnetite: Physical Review B, v. 63, no. 5, p.054110.The electronic state of Fe

3O

4 below the

Verwey transition has been studied by X-rayresonant scattering at the iron K edge of theforbidden (0 0 2) and (0 0 6) reflections.These reflections are permitted by theanisotropy of the iron anomalous scatteringfactor. The energy and the azimuthal angledependence of the intensity of thesereflections do not change above and below thetransition temperature. Moreover, thespectrum at low temperature does not changeupon orientation of the crystal c axis. Theresults clearly show that no charge ordering ofFe3+ and Fe2+ occurs in the insulating phase ofmagnetite.

Zysler, R. D., Vasquez-Mansilla, M.,Arciprete, C., Dimitrijewits, M., Rodriguez-Sierra, D., and Saragovi, C., 2001, Structureand magnetic properties of thermallytreated nanohematite: Journal of Magnetismand Magnetic Materials, v. 224, no. 1, p. 39-48.The effect of thermal treatment on thestructure and magnetic properties of α-Fe

2O

3

rhombohedrical nanoparticles (30 nm)synthesized by chemical route has beenanalyzed by XRD, TEM, magnetizationmeasurements and Mossbauer spectroscopy.Annealing of these samples recrystallizes thenanoparticles maintaining their mean sizewhile changing the crystalline anisotropyenergy, thus leading to an increase of the spinreorientation Morin temperature and changesin the superparamagnetic-blocking behavior.

8

CHAPTER V.

MAGNETIC HYSTERESIS.

§ 77. Magnetic Hysteresis.-The curveswhich have been drawn to show the effectsof cyclic magnetising processes in iron,steel, nickel, and cobalt, have this impor-tant feature in common, that there is a ten-dency on the part of the metal to persist inany magnetic state which it may have ac-quired. This tendency is specially obvi-ous whenever an alteration begins to bemade in the character of the magnetisingprocess. Thus, when the magnetising forcehas been raised to its highest value, we find,on beginning to reduce the force, that themagnetism tends to remain. It does not allremain, but the rate at which it disappearsduring withdrawal of the magnetising forceis notably less than the rate at which mag-netism was being acquired during imposi-tion of the force, especially at the begin-ning of the withdrawal. The existence ofresidual magnetism when the force iswholly withdrawn is one result of this re-luctance on the part of the metal to changeits magnetic condition. But the results ofthis tendency go further. If, for example,

after withdrawing the magnetising force,we begin to re-apply it, we find in the earlystages of the process the same reluctanceto change; the metal begins to regain mag-netism, but not so fast as it was losingmagnetism during the last stages of the re-moval of the force. The rate, however,improves, and when the force has beencompletely restored we find that the piecehas recovered all, or nearly all (sometimeseven a little more than all), the magnetismit lost while the force was being withdrawn.The curve of magnetisation comes againto the same, or nearly the same, point asthat from which it started; but its path dur-ing the process of return differs entirelyfrom its path during removal of the force.The two curves form a loop, and any inter-mediate value of the magnetic force is asso-ciated with different values of themagnetisation during the two processes.

Moreover, this description appliesequally to the effects of any cyclic varia-tion of magnetic force, provided the rangethrough which the force is varied be notexceedingly small. Starting from any con-dition of magnetism and of magnetisingforce, if we remove and re-apply a part ofthe force, or if we apply and remove asupplementary force, and repeat the pro-cess until its effects become cyclic, we findthat the two stages of the process may berepresented by two curves, which do notcoincide, but differ in a way that may beconcisely described by saying that there isa tendency, at each change of process, forthe preceding magnetic condition to per-sist. The changes of magnetism lag be-hind the changes of force. This tendencyhas received the name of magnetic hyster-esis, from υστερω, to lag behind.*

* Proc. Roy. Soc., No. 216, 1881, p.22;Phil. Trans., 1885, p.524.

9

10

Preisach distribution, the FORCdistribution is always well-defined forany magnetic system. The FORCdistribution ρ(Hc, Hb) is defined usingonly a set of measured FORC data, asecond derivative, and a change ofcoordinates. Its definition is notdependent on the Preisach model, or anyother theoretical model.

The details of data acquisition andanalysis involved in obtaining a FORCdistribution have been described in detailelsewhere [2,3]. A certain amount ofnumerical smoothing is inherent in thecalculation of a FORC diagram fromexperimental data: this is quantified by asmoothing factor (SF), which can varybetween 2 for the highest quality dataand 5 for poor quality data. Figure 2shows both typical FORC data and theFORC diagram for a well-characterizedsample of Yucca Mountain ash flow tuff,from southern Nevada, which consists ofweakly interacting SD particles [4]. Thefact that the distribution in Figure 2(b) isnarrowly confined to the centre horizon-tal axis is characteristic of weaklyinteracting SD particles.

At the IRM we measured, for thefirst time, FORC diagrams on a VibratingSample Magnetometer at low tempera-tures. The ferrofluid, produced byFerrofluidics, Inc., comes with extracarrier fluid which can be added to dilutethe particle concentration. We studied itin the concentration out of the bottle, 0.2emu/mL, and at roughly a 10 to 1dilution. The results, shown in Figure 4,are satisfying in that they confirm manyaspects of FORC diagrams and SDparticle models. Note in both theundiluted and diluted samples thatdecreasing temperature pushes thedistribution peak to higher coercivity. Atlow enough temperature, this distributionwould be well separated from the verticalaxis, like the distribution in Figure 2; thefact that even at 5 K the ferrofluid

...FORCs & ferrofluidcontinued from page 1

equals 1 for large values of H, switchesto -1 when H falls below Hα, andremains at -1 until H rises above Hβ.The FORC distribution for γαβ is a pointdelta function which has an infinite peakat Hr = Hα and Ha = Hβ, and whichequals zero elsewhere. In Hc and Hbcoordinates, ρ(Hr, Ha) will consist of apoint delta function at Hc = (Hα - Hβ )/2,Hb = (Hα + Hβ )/2. However, (Hα - Hβ )/2 is just the half-width of γαβ, whichcorresponds to its coercivity, and (Hα +Hβ )/2 is the horizontal offset of γαβ,which we refer to as its bias (see Figure3). Hence, on a FORC diagram, the Hcand Hb coordinates of the point deltafunction give the coercivity and bias,respectively, of γαβ.

In the Preisach model, hysteresis isgenerated by the superposition of acollection of hysterons with a distribu-tion of coercivities and biases, where thedistribution of coercivities and biases isreferred to as a Preisach distribution,P(Hc, Hb). By a simple extension of theabove- described result with onehysteron, it can be shown that the FORCdistribution ρ(Hr, Ha) for such acollection of hysterons will be equivalentto P(Hc, Hb). Whenever a hysteresissystem can be rigorously described bythe Preisach model (i.e., as the superpo-sition of hysterons), the FORC andPreisach distributions will be equivalent.

The Preisach model was firstsuggested as a model of hysteresis ininteracting SD particle systems. In thismodel, each hysteron represents anindividual particle in the assemblage.The coercivity of a hysteron correspondsto the coercivity of a SD particle if itwere isolated from the system. The biasof a hysteron corresponds to a fixedinteraction field, which represents theinteraction of an individual particle withthe surrounding assemblage of particles.The Preisach distribution corresponds toa distribution of particle coercivities andinteraction fields. Therefore, themotivation for the mixed secondderivative and change of coordinates isthat, at least in this simple model, theFORC distribution ρ(Hr, Ha) will beequivalent to the Preisach distributionand to a distribution of particlecoercivities and interaction fields.

In real interacting single domainparticles systems, hysteresis is morecomplex and rich than can be describedby a simple superposition of hysterons,so the Preisach model must be consid-ered as a phenomenological constructionto describe hysteresis in an approximateway. Similarly, the interpretation of aPresiach distribution as a distribution ofcoercivities and interaction fields is onlyvalid in an approximate way. And thePreisach distribution associated with aspecific system is not uniquely specifiedor well-defined. In contrast to the

Figure 3. The hysteron γαβ equals 1 for largevalues of H, switches to -1 when H falls belowHα, and remains at -1 until H rises above Hβ.The half width of the hysteron is itscoercivity, and the horizontal offset of itscenter is referred to as its bias.

Figure 1. The measurement of a First Order Reversal Curve (FORC) beginswith saturation by a positive field. The field is ramped down to a reversalfield Hr. The FORC consists of a measurement of the magnetization as thefield is then increased from Hr back up to saturation. The magnetization atapplied field Ha on the FORC with reversal point Hr is denoted by M(Hr,Ha).The dotted line represents the major hysteresis loop. The FORC with areversal field at negative saturation is equivalent to the lower half of themajor hysteresis loop.

Figure 2. (a) FORC data for Yucca Mountain ash flow tuff sample, fromsouthern Nevada. (b) A FORC diagram for above data (SF = 2).

When a FORC distribution is plotted,it is convenient to change coordinatesfrom {Hr, Ha} to {Hc = (Ha - Hr)/2, Hb =(Ha + Hr)/2}. A FORC diagram is acontour plot of a FORC distribution withHb and Hc on the vertical and horizontalaxes, respectively. Since Ha > Hr, thenHc>0, and a FORC diagram is confinedto the right-hand half plane. In Figure2(b) we show the FORC diagramcalculated from the data in Figure 2(a).The Hc coordinate is referred to as themicrocoercivity.

To explain the motivation for thismixed second derivative and change ofcoordinates, it is necessary to give a briefintroduction to the Preisach model. Letus begin by defining a simple mathemati-cal construction referred to as a hysteron.As shown in Figure 3, the hysteron γαβ

This is primarily a result of noise. Wehave chosen a contour line close to azero value of the distribution. As aresult, when the distribution gets close tozero, noise can produce spuriouscrossings of the zero contour boundaryline, which gives rise to randomfluctuations of the dark shading.However, at 5 and 9 K, for the dilutedsample, there is a negative region in thelower left hand corner which is not justnoise. Weakly negative regions are, infact, commonly observed in FORCdiagrams, which is one aspect of thesediagrams that separates them fromPreisach diagrams.

Finally, an interesting result with theundiluted sample is a secondary peaklocated at the diagram origin, and to theleft of the main peak. At first sight, thispeak might seem to indicate some kind ofbimodal particle distribution. But wehave shown theoretically that thissecondary peak arises when the distribu-tion crosses the stable/superparamagneticthreshold. The interesting thing for us itthat this secondary peak is spread outvertically by the interactions in theundiluted sample.

The ferrofluid studied here is a well-behaved and well understood system, andnot necessarily of interest to geologists.However, as we gain experience withFORC diagrams in these simple systems,we will be better able to apply FORCdiagrams to more complicated naturalsystems.

References1. Mayergoyz, I.D., Mathematical

models of hysteresis, IEEE Trans.Magn., MAG-22, 603-608, 1986.

2. Pike, C.R., A.P. Roberts, and K.L.Verosub, Characterizing interactionsin fine magnetic particle systemsusing first order reversal curves, J.Appl. Phys., 85, 6660-6667, 1999.

3. Roberts, A.P., C.R. Pike, and K.L.Verosub, FORC diagrams: A newtool for characterizing the magneticproperties of natural samples, J.Geophys. Res., in press.

4. C.M. Schlinger, D.R. Veblen and J.G.Rosenbaum, Magnetism andmagnetic mineralogy of ash flowtuffs from Yucca Mountain, Nevada,J. Geophys. Res. B, 96, 6035–6052,1991

Ewing, (James) Alfredb. March 27, 1855, Dundee, Scotlandd. Jan. 7, 1935, Cambridge, England

Sir Alfred is credited with discovering andnaming the phenomenon of hysteresis. Astudent and protegé of Fleeming Jenkin atEdinburgh, Ewing went on to professorships inmechanical engineering in Tokyo, Ediburghand Cambridge. His wife, née Annie Washing-ton, was great-great grand niece of the first USpresident. Ewing wrote numerous papers onthe effects of stress on magnetization, andseveral books, including The Steam Engine andOther Heat Engines (1885) and MagneticInduction in Iron (1892). His research alsoincluded work on cryptography, seismology,and thermoelectric effects.

11

Figure 4. Ferrofluid results for diluted and undiluted ferrofluid. Thetemperature and averaging time of measurement are indicated. SF = 3. Thegreater vertical spread of the undiluted sample indicates stronger interactions.Increased temperature shifts the distribution to lower microcoercivities.

distribution meets the vertical axisindicates that a large fraction of thesystem is still unblocked at 5 K.

Notice that the degree of verticalspread is greater for the undiluted case.This confirms that, to a first order, thevertical spread is a measure of adistribution of interaction fields.Furthermore, the degree of verticalspread in the undiluted sample isrelatively constant from 5 to 9 to 12 K,while the distribution ofmicrocoercivities shifts substantially.This indicates that, to a first order, thestrength of interactions is independent ofthe particle coercivities and temperature,which is a satisfying result.

The darkest shaded regions of theseplots correspond to negative values ofthe FORC distribution. At the right handside of these diagrams, there arerandomly fluctuating negative regions.

press), or if you have any suggestionsto improve the newsletter, pleasenotify the editor:Mike JacksonInstitute for Rock MagnetismUniversity of Minnesota291 Shepherd Laboratories100 Union Street S. E.Minneapolis, MN 55455-0128phone: (612) 624-5274fax: (612) 625-7502e-mail: irm@geolab.geo.umn.eduwww.geo.umn.edu/orgs/irm/irm.html

I R MInstitute for Rock Magnetism

The U of M is committed to the policy thatall people shall have equal access to itsprograms, facilities, and employmentwithout regard to race, religion, color, sex,national origin, handicap, age, veteranstatus, or sexual orientation.

KE

LVIN

Co

llecto

r’s Se

ries #

21

source

: The M

an of Room

40: The Life of S

ir Alfred E

wing,

by A

.W

. Ew

ing, 1

939, H

utch

inso

n &

Co. L

td., L

ondon.

12

The Institute for Rock Magnetism isdedicated to providing state-of-the-artfacilities and technical expertise free ofcharge to any interested researcher whoapplies and is accepted as a VisitingFellow. Short proposals are acceptedsemi-annually in spring and fall for workto be done in a 10-day period during thefollowing half year. Shorter, less formalvisits are arranged on an individual basisthrough the Facilities Manager.

The IRM staff consists of SubirBanerjee, Professor/Director; BruceMoskowitz, Associate Professor/Associate Director; Jim Marvin , SeniorScientist; Mike Jackson, SeniorScientist and Facility Manager, and PeatSolheid, Scientist.

Funding for the IRM is provided bythe National Science Foundation, theW. M. Keck Foundation, and theUniversity of Minnesota.

The IRM Quarterly is published fourtimes a year by the staff of the IRM. Ifyou or someone you know would like tobe on our mailing list, if you havesomething you would like to contribute(e.g., titles plus abstracts of papers in

University of Minnesota291 Shepherd Laboratories100 Union Street S. E.Minneapolis, MN 55455-0128phone: (612) 624-5274fax: (612) 625-7502e-mail: irm@geolab.geo.umn.eduwww.geo.umn.edu/orgs/irm/irm.html

EWING

Nonprofit Org.U.S Postage

PAIDMpls., MN

Permit No. 155