Rock magnetism of remagnetized carbonate rocks: another lookgilder/Re... · 1976), carbonate rocks represented a new fron-tier for palaeomagnetic research. These and other weakly

Rock magnetism of remagnetized carbonate rocks: another look

MIKE JACKSON* & NICHOLAS L. SWANSON-HYSELL

Institute for Rock Magnetism, Winchell School of Earth Sciences,

University of Minnesota, Minnesota, US

*Corresponding author (e-mail: [email protected])

Abstract: Authigenic formation of fine-grained magnetite is responsible for widespread chemicalremagnetization of many carbonate rocks. Authigenic magnetite grains, dominantly in the super-paramagnetic and stable single-domain size range, also give rise to distinctive rock-magnetic prop-erties, now commonly used as a ‘fingerprint’ of remagnetization. We re-examine the basis of thisassociation in terms of magnetic mineralogy and particle-size distribution in remagnetized carbon-ates having these characteristic rock-magnetic properties, including ‘wasp-waisted’ hysteresisloops, high ratios of anhysteretic remanence to saturation remanence and frequency-dependentsusceptibility. New measurements on samples from the Helderberg Group allow us to quantifythe proportions of superparamagnetic, stable single-domain and larger grains, and to evaluatethe mineralogical composition of the remanence carriers. The dominant magnetic phase ismagnetite-like, with sufficient impurity to completely suppress the Verwey transition. Particlesizes are extremely fine: approximately 75% of the total magnetite content is superparamagneticat room temperature and almost all of the rest is stable single-domain. Although it has been pro-posed that the single-domain magnetite in these remagnetized carbonates lacks shape anisotropy(and is therefore controlled by cubic magnetocrystalline anisotropy), we have found strong exper-imental evidence that cubic anisotropy is not an important underlying factor in the rock-magneticsignature of chemical remagnetization.

The usefulness of a palaeomagnetic remanence isdirectly related to the accuracy with which its age isknown (e.g. Van der Voo 1990). Primary remanence,acquiredduringorverysoonafter rock formation,pro-videsdirect informationonpalaeofieldorientationandstrength at that time. The possibility of partial or com-plete remagnetization at a later time complicatespalaeomagnetic interpretation, ashas longbeenrecog-nized (e.g. Graham 1949). Every robust palaeomag-netic study must therefore include some effort toconstrain the ages of identified components of thenatural remanent magnetization (NRM). Relativedating with respect to sedimentary processes, struc-tural tilting or cross-cutting relationships is the basisof the classical geometric palaeomagnetic tests (foldtest and conglomerate test, Graham 1949; bakedcontact test, Everitt & Clegg 1962; unconformitytest, Kirschvink 1978). The special circumstancesrequired for application of these tests are often notavailable however, or positive test results mayprovide relatively loose constraints allowing magneti-zation ages that significantly post-date the age ofthe rock formation. As a result, other evidence(although generally more indirect) becomes essentialin evaluating the age, origin and significance ofNRM components. Such evidence commonlyincludes petrographic observations, isotopic and geo-chemical data and rock magnetic characterization.

By their very existence, magnetic overprintsprovide evidence of some process or event of

potential geological significance (e.g. McCabe &Elmore 1989; Elmore & McCabe 1991). The ‘usualsuspects’ in remagnetization are a set of mechanismsinvolving elevated temperature, stress and chemicalactivity, acting alone or in concert, and producingchanges in the magnetic minerals themselves and/or in the remanence that they carry. When theNRM and the population of magnetic carriers areboth modified by some process or event, carefulrock magnetic analysis may shed light on the culprit.

Our focus here is chemically remagnetizedcarbonate rocks where any primary remanence iscompletely obscured by ancient yet secondarymagnetizations. Many such units have been docu-mented in recent decades (e.g. McCabe & Elmore1989). With the advent of superconducting magnet-ometers in the 1970s and 1980s (Goree & Fuller1976), carbonate rocks represented a new fron-tier for palaeomagnetic research. These and otherweakly magnetic sedimentary rock strata had beenunmeasurable during progressive demagnetizationwith the limited-sensitivity spinner magnetometersthen available. Superconducting magnetometersenabled a rapid expansion into these untappedpalaeomagnetic archives, but by the mid 1980s itwas already evident that many Palaeozoic carbonaterocks had been completely remagnetized in latePalaeozoic time and it became important to under-stand the processes responsible (e.g. McCabe &Elmore 1989).

From: Elmore, R. D., Muxworthy, A. R., Aldana, M. M. & Mena, M. (eds) 2012. Remagnetization and ChemicalAlteration of Sedimentary Rocks. Geological Society, London, Special Publications, 371,http://dx.doi.org/10.1144/SP371.3 # The Geological Society of London 2012. Publishing disclaimer:www.geolsoc.org.uk/pub_ethics

at University of California Berkeley on July 30, 2013http://sp.lyellcollection.org/Downloaded from

http://sp.lyellcollection.org/

Review of previous work

Origins of the secondary remanence and its

carriers

Widespread remagnetization of Palaeozoic carbon-ate strata in the Appalachian Basin was initiallyrecognized and documented through combinedpalaeomagnetic, structural and petrographic obser-vations. These observations included the recogni-tion that the characteristic remanent magnetization(ChRM) directions sometimes clustered signifi-cantly more tightly prior to correction for structuraltilting, and sometimes clustered most tightly atintermediate unfolding. This led to interpretationsthat the magnetizations were acquired after and/orduring folding (McCabe & Elmore 1989). Petro-graphic work led to the observation of spheroidal,botryoidal and framboidal magnetite grains of pureend-member Fe3O4 composition that were inter-preted to be of low-temperature diagenetic origin(McCabe et al. 1983). Early rock-magnetic charac-terization focused on acquisition and demagnetiza-tion of isothermal remanent magnetization (IRM)and anhysteretic remanent magnetization (ARM),leading in many cases to the conclusion that theremanence carriers were dominantly pseudo-single-domain (PSD) or multi-domain (MD) magnetite(e.g. McElhinny & Opdyke 1973; Kent 1979).Hysteresis measurements eventually led to therecognition that the dominant magnetic carriers inmany remagnetized carbonates are in fact veryfine grained, mainly in the superparamagnetic (SP)and stable single-domain (SSD) size range(Jackson 1990; McCabe & Channell 1994; Suk &Halgedahl 1996; Dunlop 2002b). Moreover, thehigh proportions of SP and SSD particles in the fer-rimagnetic inventory of these rocks result in unusualcombinations of the summary parameters conven-tionally used to represent hysteresis behaviour,namely coercivity Bc, remanent coercivity Bcr, sat-uration magnetization Ms, saturation remanenceMrs and the ratios Bcr/Bc and Mrs/Ms. On the ‘Dayplot’ (Mrs/Ms v. Bcr/Bc; Day et al. 1977; Dunlop2002a, b), remagnetized carbonates are largelyisolated in a region that is distinct from theregions occupied by most other rocks, sedimentsand synthetic materials (Fig. 1), leading to the sug-gestion that hysteresis properties could be used asa fingerprint for recognition of remagnetizedstrata (Jackson 1990; Channell & McCabe 1994;McCabe & Channell 1994). We will explore thebasis and the limitations of this idea in some depthin the present paper.

The origin of the ancient yet secondary char-acteristic remanence was attributed by McCabeet al. (1983) to chemical processes (growth ofnew minerals and/or replacement of pre-existing

minerals). The spatial/temporal association ofwidespread remagnetization with Appalachian–Alleghenian–Variscan–Hercynian orogenesis (forsummaries see e.g. Weil & Van der Voo 2002;Tohver et al. 2008) soon led to various hypothesesinvolving continental-scale flow of orogenic fluids(Oliver 1986; Bethke & Marshak 1990; Garven1995) to account for not only remagnetization, butalso ore mineralization and hydrocarbon distri-bution in a number of different basins. Isotopicand geochemical evidence in some cases ruled outa significant role for exotic brines (e.g. Elmoreet al. 1993; Ripperdan et al. 1998), and pointedinstead to more closed-system mechanisms includ-ing clay diagenesis (Katz et al. 2000; Woods et al.2002), organic maturation (Blumstein et al. 2004)and pressure solution (Zegers et al. 2003; Elmoreet al. 2006; Oliva-Urcia et al. 2008). Kent (1985)questioned whether the remagnetization necessarilyinvolved chemical mechanisms at all, suggestingthat viscous acquisition during the Kiaman super-chron was a viable alternative. Noting that therange of observed unblocking temperatures couldbe accounted for by the Viscous Remanent Magneti-zation (VRM) model of Walton (1980), Kent (1985)demonstrated moreover that the Brunhes-agedviscous overprint in these rocks had unblockingtemperatures similar to those predicted by the samemodel. However, this model was later shown to beinappropriate for predicting unblocking temperatures(Dunlop et al. 1997a, b), undermining the theoreticalbasis for a viscous origin but not the empirical obser-vations. Similarly, Kligfield & Channell (1981)invoked thermoviscous mechanisms to explain wide-spread Brunhes-age remagnetization of limestonesfrom the Swiss Alps, in which the ChRM wascarried chiefly by pyrrhotite with fine grain sizesspanning the SP–SSD range. They noted that the pyr-rhotite may have formed by alteration of early-diagenetic pyrite, but did not address the possibilityof a chemical origin for the remanence.

Subsequent work on the remagnetized Appala-chian Basin carbonates and remagnetized carbon-ates in other basins has largely supported theidea that the remanence carriers are predominantlyof late diagenetic origin, and that one or morechemical mechanisms were involved in producingthe secondary characteristic components of naturalremanence (Reynolds 1990; Suk et al. 1990a, b,1991, 1993; Weil & Van der Voo 2002; Zwinget al. 2005). Our focus here is on the importantrole that rock magnetic studies have played inthe understanding of carbonate remagnetization,but of course we recognize that a complete pic-ture also includes careful petrographic, geo-chemical, isotopic, microstructural and analyticalmicroscopic work (Elmore & McCabe 1991;Elmore 2001).

M. JACKSON & N. L. SWANSON-HYSELL



In the interest of clarity, we will take a specificand limited set of previously studied remagnetizedcarbonate units from the Appalachian Basin as pro-totypes of the distinctive set of rock-magnetic prop-erties and NRM characteristics that we wish tounderstand. These include the Devonian OnondagaLimestone (Kent 1979; Jackson et al. 1988; Suket al. 1990a; Lu et al. 1991) and HelderbergGroup (Scotese et al. 1982) and the OrdovicianTrenton Limestone (McElhinny & Opdyke 1973;McCabe et al. 1984) and Knox Dolomite (Bachtadseet al. 1987; Suk et al. 1990b). All of these carry atwo-component NRM with a well-documentedreverse-polarity Permian-aged ChRM and a normal-polarity recent viscous overprint, and all exhibitrock-magnetic properties dominated by SP andSSD ferrimagnets (Jackson 1990; Jackson et al.1993; Jackson & Worm 2001; Dunlop 2002b). Byunderstanding as well as possible the relationshipbetween the ChRM and the magnetic phases andtheir origins in these prototypical units, our goal isto improve our ability to interpret similarities anddifferences in properties observed in other carbon-ate strata for which the magnetization history isless well known.

Unresolved questions

Nature of grain-scale anisotropy. The hysteresisratios of several of the remagnetized Palaeozoic car-bonate units from the Appalachian Basin follow a

power-law trend, visible as a linear array of pointson a bilogarithmic Day plot as in Figure 1. Thistrend is best explained as resulting from a mixtureof SP and SSD ferrimagnets (Dunlop 2002a, b).Further, it has been suggested (Jackson 1990) thatthe SSD fraction (at least) in remagnetized Appala-chian carbonates has properties dominated by cubicmagnetocrystalline (multiaxial) anisotropy ratherthan by shape anisotropy. In a population of ran-domly oriented single-domain ferrimagnetic par-ticles dominated by uniaxial shape anisotropy, theratio Mr/Ms should approach 0.5 as the coercivitiyratio approaches 1 (Stoner & Wohlfarth 1948). Incontrast, the Mr/Ms intercept for the remagnetizedAppalachian carbonates is c. 0.89, as expected forrandomly oriented single-domain (SD) grains withcubic anisotropy (Wohlfarth & Tonge 1957). Itshould be noted that the error bars on this interceptvalue are relatively large, and moreover that linearextrapolation is probably not valid (Dunlop 2002a,b). Nevertheless, the suggestion of dominantlycubic anisotropy is interesting and has importantimplications. Populations of ferrimagnetic grainsthat are dominated by multiaxial anisotropy areunusual in nature, because even a slight elongation(e.g. length to width ratio of c. 1.1 and greater) issufficient to make magnetostatic (shape) anisotropythe dominant factor controlling the hysteresis andremanence behaviour of magnetite or maghemite.The strong implication of a nearly complete lackof particle elongation is that the remanence-carrying

(a) (b)

g–1

Ms

Mr

Bc

Fig. 1. (a) Summary Day plot of published hysteresis parameters for carbonate rocks, in which the ratio ofremanent coercivity (Bcr) to coercivity (Bc) is plotted against the ratio of saturation remanence (Mr) to saturationmagnetization (Ms). Data from carbonates interpreted to carry primary remanence are marked with circles, anddata from carbonates interpreted to be remagnetized are marked with diamonds. (b) An example hysteresis loopfrom sample H2-5.35 of the Manlius Formation of the Helderberg Group is shown with Bc, Mr and Ms marked. Notethat the loop had not reached saturation by 100 mT (the maximum of the x-axis), such that Ms is not reached until higherapplied fields (off the plot).

ROCK MAGNETISM: REMAGNETIZED CARBONATES



particles have authigenic/diagenetic morphologiessimilar to those observed for the much larger par-ticles retrieved by magnetic extraction of insolubleresidues (e.g. McCabe et al. 1983) and that theunobserved, much finer, SSD particles originatedin the same way.

On the basis of hysteresis parameters, a similarargument for multiaxial anisotropy was later madeby Gee & Kent (1995) for mid-ocean-ridge basalts(MORB), whose remanence resides chiefly in tita-nomagnetites of composition Fe3-xTixO4 with xc. 0.6 (TM60). This argument has been criticizedby Fabian (2006) on two grounds. First, theoreticalconsiderations, combined with known intrinsicmaterial properties, suggest that cubic magneto-crystalline (multiaxial) anisotropy should almostalways be less important than magnetostrictiveand shape (uniaxial) anisotropies for titanomagne-tites. Second, careful measurements showed thatmagnetization of MORB samples is often not fullysaturated in the field range used for hysteresis loopmeasurements (typically not exceeding 1 Tesla(T), and incomplete saturation in combination withconventional linear high-field slope calculationsyield erroneously low values for Ms (and thus erro-neously high Mr/Ms ratios). Fabian (2006) showedfor several MORB samples that even althoughloops measured on a vibrating-sample magnet-ometer (with peak fields of 1 T) yielded apparentMr/Ms ratios well above 0.5, measurement of thesame samples in fields up to 7 T in a super-conducting magnet resulted in significantly higherMs estimates and Mr/Ms ratios below the uniaxiallimit. Fabian (2006) concluded that uniaxial stressanisotropy exerts the dominant control on thehysteresis behaviour of MORBs. Although he didnot discuss remagnetized carbonates, his findingsraised questions about the identification of multi-axial anisotropy in any material through standardhysteresis experimental protocols.

Fabian’s (2006) arguments and experimentalresults appeared conclusive, but the potential impor-tance of magnetocrystalline anisotropy in MORBshas not been put to rest. A strong counterargumentin favour of such multiaxial anisotropy in MORBtitanomagnetites was recently made by Lanci(2010), by means of a clever technique based onmeasurements of field-impressed anisotropy of sus-ceptibility (illustrated in Fig. 2a). The essence ofLanci’s approach involves the ‘inverse’ anisotropyof magnetic susceptibility (AMS) of SD particles:an individual SD grain has minimum (essent-ially zero) susceptibility parallel to its remanentmoment (because it is already magnetized to satur-ation in that direction) and maximum susceptibilityperpendicular to that easy axis. For a populationof uniaxial SD grains this results in the well-known ‘inverse-fabric’ effect (Potter & Stephenson

1988); a preferential alignment of long axes givesminimum susceptibility parallel to lineation. Whenuniaxial particles are randomly oriented, the acqui-sition of a strong-field remanence has only veryweak effects on AMS, because particle momentsare spread out all over the directional hemispherearound the net remanence (i.e. around the appliedfield orientation). Lanci’s (2010) idea is that multi-axial particle anisotropy makes remanence effectsmuch more extreme: even if easy axes are ran-domly oriented, acquisition of a strong-field rema-nence causes the particle moments to be confinedto a much smaller range of orientations close tothe applied field direction because there are moreeasy axes per particle. This therefore becomesthe minimum susceptibility direction (Fig. 2a).Very large anisotropies of 200–300% can be pro-duced by acquisition of a saturation isothermalremanent magnetization (SIRM) in populationsof randomly oriented SD grains with multiaxialanisotropy. Lanci’s experimental results showedthat imprinting a 1 T IRM on several MORBsamples with very weak initial AMS resulted ina strong anisotropy of susceptibility (10–25%)with kmin parallel to the IRM, and showed thatthe effect could be erased by tumbling alternating-field (AF) demagnetization. He calculated that20–30% of the susceptibility in those MORBsamples can be attributed to titanomagnetite par-ticles with multiaxial anisotropy.

Further evidence for multiaxial anisotropy inMORB samples has been presented by Mitra et al.(2011) using another novel experimental approach,involving asymmetry of forward and reverse IRMacquisition (illustrated in Fig. 2b). An individualuniaxial SD grain has the same switching thresholdin fields of either polarity. A population of uniaxialSD grains, initially in a weakly magnetized state(e.g. NRM state, or after tumbling AF), willacquire an IRM of intensity Mr

+ in a non-saturatingapplied field Bapp. Application of an equal-strengthfield in the opposite direction will produce an IRMof equal intensity (Mr

2 ! Mr+) antiparallel to the

positive-field IRM, as each uniaxial particle thatreversed its moment into alignment with the positivefield reverses again into alignment with the negativefield. This equivalence between the Mr

2 and Mr+

does not hold for multiaxial SD grains (Fig. 2b).In contrast to uniaxial grains where the moment isconfined to the long axis of the grain and only twopossible remanent orientations exist, multiaxialgrain moments have a number of possible remanentorientations only one of which is antiparallel to theoriginal moment orientation. Application of a non-saturating field may cause the moment of a particu-lar grain to jump from one easy axis to another,closer to the applied field direction and thus contri-buting to Mr

+. Subsequent application of a reverse




field of equal strength to the population of multiaxialgrains will not necessarily cause the moment of thesame grain to reorient, because the angle betweenthe moment and the field is now more acute thanit was in the initial state. Mr

2 is consequently oflower intensity than Mr

+ for equal and oppositeapplied non-saturating fields (Fig. 2b). Mitra et al.(2011) demonstrated such positive/negative asym-metry in IRM acquisition experiments for MORBsamples and for hematite samples (with multi-axial basal-plane grain-scale anisotropy) with inten-sity differences reaching 100% for intermediateapplied fields, decreasing to zero for saturatingfields. Similar measurements on other controlsamples showed essentially identical forward and

reverse IRM acquisition, as expected for sampleswith uniaxial remanence carriers.

Although these arguments over multiaxial aniso-tropy have focused almost entirely on MORBsamples, the same questions apply to remagnetizedcarbonates. Suk & Halgedahl (1996) pointed outthat observed coercivities of remagnetized Appla-chian Basin carbonates were two to three timeshigher than can be theoretically produced by magne-tocrystalline anisotropy in magnetite, but theysuggested that a mixture of uniaxial and multiaxialparticles could account for all observed properties.Evidence for multiaxial grain-scale anisotropycomes almost entirely from the intercept value ofthe trend line on the bilogarithmic Day plot; few

(a)

(b)

Fig. 2. Schematic illustration of tests for multiaxial particle anisotropy in single-domain populations. The testdepicted in (a) involves the effects of strong-field remanence acquisition on anisotropy of low-field susceptibility (Lanci2010); with multiaxial particle anisotropy, a strong alignment of particle moments results in a strong anisotropy ofmagnetic susceptibility (AMS). The test depicted in (b) involves asymmetric stepwise acquisition of isothermalremanent magnetization (IRM) in positive and negative fields of equal strength (Mitra et al. 2011). See text for detailedexplanation of both tests.




individual loops have yielded remanence ratiosabove 0.5 and, for those, non-saturation may be aproblem. Moreover, no independent evidence formultiaxial anisotropy has been found for carbonaterocks. One goal of the present study is to apply thenew techniques of Lanci (2010) and Mitra et al.(2011) to some of the well-studied Palaeozoic car-bonates of the Appalachian Basin, to test thehypothesis that multiaxial anisotropy is a character-istic feature of their magnetic mineralogy.

Mineral mixtures and the importance of pyrrhotite.Wasp-waisted hysteresis loops and Day plot trendssimilar to those of the remagnetized carbonates inFigure 1 require the presence of two or more dis-tinct coercivity fractions, which may correspond toeither different remanence-carrying minerals withstrongly contrasting coercivities (Jackson et al.1990; Muttoni 1995; Roberts et al. 1995) or differ-ent size fractions of a single mineral (especiallyif one is superparamagnetic; Tauxe et al. 1996;Dunlop 2002a, b; Lanci & Kent 2003). These twoscenarios have different implications for theorigins of the carriers and, correspondingly, of thenatural remanence. A single process or remagneti-zation event could account for the co-occurrenceof SP and SSD magnetites in chemically overprintedrocks, with a unimodal size distribution of mono-genetic carriers naturally producing the strong coer-civity contrast (Lanci & Kent 2003). Polymineraliccarrier populations imply a more complicatedhistory where either a primary phase was preservedwith a secondary authigenic phase, or where thereare multiple preserved generations of secondaryauthigenic phases.

The magnetic mineralogy of many previouslystudied remagnetized carbonates is thought to com-prise primarily magnetite, largely on the basis ofunblocking temperatures, but remains incompletelyknown. In some cases pyrrhotite (Xu et al. 1998;Zegers et al. 2003) or hematite (see list in McCabe& Elmore 1989) have been documented as overprintcarriers. Multiple studies have argued for the preser-vation of primary magnetite remanence in carbon-ates that have a secondary remanence held bypyrrhotite (Muttoni & Kent 1994; Symons et al.2010). Strong-field thermomagnetic data are gener-ally rare, due to the difficulty of obtaining measur-able signals from small quantities of this weaklymagnetic material. While a thermomagnetic curveon a sample chip of the Manlius Formation of theHelderberg Group revealed a smooth reversiblecurve with a Curie temperature at c. 565 8C (Sco-tese et al. 1982), this result has not been confirmedfor other remagnetized carbonates or for otherHelderberg Group samples. Laboratory heating alsotypically results in some degree of alteration of themagnetic mineralogy, especially for pyrrhotite and

other sulphides, adding uncertainty to the interpret-ation. More readily measurable unblocking tempera-ture spectra of NRM and of artificial remanenceshave generally shown low-to-moderate thermal stab-ility, with typical median demagnetization tempera-tures (i.e. temperatures required to erase half theinitial remanence) of 200–300 8C and maximumunblocking temperatures c. 450–500 8C (e.g. Jacksonet al. 1993; Zegers et al. 2003; Elmore et al. 2006).These data allow multiple possibilities, including(but not requiring) significant contributions frompyrrhotite or detrital titanomagnetites as well asfine-grained pure magnetite (with Tub , Tc) and/or magnetite with dominantly magnetocrystallineanisotropy.

Low-temperature remanence measurements(Jackson et al. 1993) on samples of the DevonianOnondaga Limestone and Ordovician TrentonLimeston of New York and of the Ordovician KnoxDolomite have shown a general absence of cleartransitions, but subtle indications have been foundin some samples of the magnetite Verwey transition(Tv c. 120 K) (Verwey & Haayman 1941; Walz2002) and/or of the pyrrhotite Besnus transition(TBs c. 32 K) (Besnus 1966; Dekkers et al. 1989;Rochette et al. 1990). As has long been known(Verwey & Haayman 1941; Ozdemir et al. 1993;Walz 2002; Ozdemir & Dunlop 2010), the Verweytransition is very sensitive to deviations fromperfect stoichiometry. Small degrees of cation sub-stitution or moderate cation deficiency (due tolow-temperature oxidation) depress the transitiontemperature and diminish the changes in measurablephysical properties, and the transition is entirelysuppressed by just a few percent content of substi-tuted metal cations or vacancies due to high-temperature oxidation (Verwey & Haayman 1941).Low-temperature oxidation produces a maghemiteshell on a stoichiometric magnetite core, and someexpression of the Verwey transition persists to highoverall degrees of oxidation (Ozdemir & Dunlop2010). The variables controlling the Besnus tran-sition are much less well known, and even its funda-mental mechanism remained unclear until quiterecently (Wolfers et al. 2011). Nevertheless, thereis a clear grain-size dependence of the magnetic be-haviour across the transition. The drop in magnetiza-tion (or susceptibility) on cooling through TBs

increases strongly with increasing particle size, andthe fractional recovery of remanence on rewarmingthrough the transition is greater for the finer sizes(Dekkers et al. 1989). The smallest fraction studiedby Dekkers et al. (1989) was 0–5 mm and for thisfraction the drop in remanence on cooling acrossTBs was about 15% of the initial room-temperatureintensity, with nearly 100% recovery on rewarming.Finer-grained pyrrhotites (but of uncontrolled andimprecisely known submicron sizes) in claystones




exhibit what has been termed the P-transition(Aubourg & Pozzi 2010), a ramp-like and perfectlyreversible change in magnetization over the temp-erature range 50–10 K. This is in contrast to the step-like change in larger pyrrhotite particles over avery narrow temperature range around TBs. TheP-transition is quite sensitive to even very weakambient fields, and may be manifested as an increaseor a decrease in magnetization while cooling inthe imperfectly controlled zero field (+500 nTafter careful adjustment) of the Quantum DesignsMagnetic Property Measurement System (MPMS)instrument commonly used for low-temperaturemeasurements. Depending on the relative abun-dances of paramagnetic iron-bearing minerals, sub-micron pyrrhotite and other remanence-bearingphases, the P-transition may be difficult or imposs-ible to discern.

Given the possibility of poorly expressed orentirely suppressed low-temperature transitions inmagnetite and pyrrhotite, any definitive quantifi-cation of their respective contributions to naturaland artificial remanences remains challenging. Thelarge spheroidal particles extracted from a num-ber of remagnetized carbonates (McCabe et al.1983; Suk et al. 1990a, b, 1993; Suk & Halgedahl1996) have been definitively shown – by x-ray dif-fraction and scanning electron microscopy (SEM)energy-dispersive analysis – to consist in mostcases of nearly pure magnetite. No evidence of mag-netic sulphides was found in any of these studies onAppalachian Palaeozoic carbonates, although somegrain morphologies such as framboids almost cer-tainly indicate formation of magnetite by replace-ment of early-diagenetic pyrite (e.g. Reynolds1990; Suk et al. 1990a) or perhaps greigite (e.g.Roberts et al. 2011). Pyrrhotite has been observedpetrographically in carbonates from the lower Car-boniferous Leadville Formation of central Coloradothat also contain authigenic magnetite (Xu et al.1998). By measuring hysteresis loops of individualextracted spheroidal particles using an extremelysensitive alternating gradient magnetometer, Suk &Halgedahl (1996) made two key additional obser-vations. First, the saturation magnetic moment ofmany of the spherules was much less than it wouldbe for the same mass of pure magnetite; these mustcontain large portions of void space, non-magneticminerals (e.g. silicates or pyrite) and/or weaklymagnetic phases such as hematite or goethite.Second, although some relatively large (up to100 mm) spherules yielded Mr/Ms ratios of 0.5 ormore (generally those with quite low Ms), thetrend for the spherule samples on the Day plotdiffered strongly from that of the bulk rocks; thesecan be interpreted as SSD–MD mixtures andSP–SSD mixtures, respectively (Suk & Halgedahl1996; Dunlop 2002b). The bulk rocks contain a

large population of ultrafine SP and SSD particlesthat evade or are destroyed by the extraction process(e.g. Sun & Jackson 1994; Weil & Van der Voo2002) and whose mineralogy is not constrained byany direct evidence.

New results

We will now attempt to address some of the out-standing questions described above through theresults of experiments conducted on a new sampleset collected from Devonian carbonates of the Hel-derberg Group in New York State (Fig. 3a). Thesenew samples were collected in the Hudson Valleyfold-thrust belt just to the north of Kingston, NYat sites along roadcuts on the north side of Route199 that correspond to two of the localitiessampled by Scotese et al. (1982) (Fig. 3b, c).Oriented cores were collected from: limestone mud-stone with fine-scale laminations of the ManliusFormation (our stratigraphic section H2; site 24 ofScotese et al. 1982); limestone packstone comprisedpredominantly of brachiopod shells with a fine-grained mud matrix of the Becraft Limestone(our stratigraphic section H1 metre levels 0 to 1.4;site 23 of Scotese et al. 1982); and from limestonemudstone and wackestone of the Alsen Formation(our stratigraphic section H1 metre levels above1.4; also site 23 of Scotese et al. 1982). Thethermal demagnetization of Scotese et al. (1982)isolated a characteristic remanence direction fromthese localities interpreted to be held by magnetitewhose direction corresponds with that of otherremagnetized Appalachian basin carbonates.

In addition to analyses on these new samples, wealso reanalyse some of the earlier datasets (Jackson1990; Jackson et al. 1993) in light of more recentdevelopments in hysteresis loop processing (Fabian2006; Jackson & Sølheid 2010), with particular atten-tion to the question of undersaturation.

Reanalysis of previous hysteresis loops

The measurements of Jackson (1990) on samplesof the Trenton, Onondaga and Knox carbonatesused maximum fields of 1 T or less; in some cases500 mT or even 300 mT was the maximum appliedfield. With fields of this magnitude it is legitimateto question whether the magnetization was saturatedover the region used for high-field slope fitting.Although a few of those samples were archived,most were not, so we cannot remeasure them inhigher fields. However, the original loop measure-ment data are still available and can be reprocessedusing statistical tests for non-saturation and non-linear approach-to-saturation fitting (Fabian 2006;Jackson & Sølheid 2010) in the high-field interval.




The highest Mr/Ms ratios in the original study (insome cases exceeding 0.5) were obtained for theKnox Dolomite and, in a number of cases, thesesamples were measured in maximum fields of300 mT under the assumption that this would be suffi-cient to saturate magnetite. Reprocessing shows thatthis was far from a safe assumption: F tests (Jackson& Sølheid 2010) comparing total-misfit variancewith pure-error variance (quantified by departuresfrom inversion symmetry) indicate that we mustreject the hypothesis of linearity over the intervalfrom 70% to 100% of the maximum applied field,

with a high degree of certainty (F ranges from 300to over 1000; compared to a critical value ofabout 1.5). Recalculation using a non-linear fittingmethod yields significantly higher estimates of Ms

and accordingly diminished estimates of Mr/Ms

(Fig. 4). The calculated coercivity is also affectedby the slope correction, and both coordinates on theDay plot therefore differ from those originallypublished by Jackson (1990). This revised analysisundermines the original argument for multiaxial ani-sotropy that was based on the apparent occur-rence of Mr/Ms ratios above 0.5, the theoretical

(a)

(c)

(b)

Fig. 3. (a) Map of New York State with the Helderberg Group mapped out in blue (bedrock geology from Fisher et al.1970). (b) Simplified geological map in the vicinity of Kingston and Albany, NY (modified from Fisher et al. 1970). Thesites sampled in this study to the north of Kingston, New York comprised a stratigraphic section across the contactbetween the Becraft and Alsen Formations (H1; 41.976048N, 073.975818W) and a stratigraphic section within theManlius Formation (H2; 41.976048N, 073.975818W). (c) Stratigraphic relationships between the formations thatcomprise the Helderberg Group between Coxsackie and Kingston, NY above its basal unconformity with the underlyingOrdovician flysch of the Austin Glen Formation (modified from Fisher 1987).




upper limit for uniaxial anisotropy. However, theselowered recalculated Mr/Ms ratios alone do not ruleout the possibility of a significant contribution fromsuch particles.

Tests for multiaxial anisotropy

The experimental approach of Lanci’s (2010) test formultiaxial anisotropy (Fig. 2) is straightforward, butfor instrumental reasons we have modified it slightly.Lanci’s technique began with an initial AMS determi-nation followed by pulse magnetizing in 1 T toproduce SIRM, and then involved repeated AMSmeasurements following tumbling AF demagnetiza-tion steps to document the gradual removal of thestrong field-impressed AMS in multiaxial SDgrains. Lacking a tumbling AF instrument, we haveslightly modified the experiment for the HelderbergGroup samples and have worked in the opposite direc-tion. We begin in the NRM state where moments arevery weakly aligned and, after measuring initial

AMS, we imprinted an IRM in one or more fieldsteps, remeasuring AMS after each step.

AsshowninTable1, theeffectofexposuretostrongfields on the AMS of the Helderberg Group sampleswas negligible. Diagonal tensor elements change bymuch less than 1%. Because most of the susceptibilitycomes from paramagnetic and superparamagneticsources, there is a relatively large correction thatmust be applied. Consequently there are rather largeuncertainties in the corrected degree of anisotropy,but it is nevertheless clear that the AMS of the potentialremanence carriers (i.e. the non-SP ferrimagneticpopulation) is strongly controlled by uniaxial grain-scale anisotropy.

The dominance of uniaxial anisotropy in the ana-lysed Helderberg Group samples is also confirmedin experiments using the approach of Mitra et al.(2011) (Fig. 2). These experiments began in theNRM state with no prior exposure to DC or ACfields, and continued with stepwise application offields of increasing strength ultimately to 1 T. For

Table 1. Field-impressed changes in AMS on Helderberg Group limestones

Specimen_ID kxx kyy kzz Treatment

H1-2.3C (Alsen Formation) 54.25 56.09 59.73 None (NRM state)H1-2.3C 54.37 56.28 59.81 1 T field along zPercent change 0.22% 0.34% 0.13%

H1-2.6B (Alsen Formation) 147.63 154.00 156.18 None (NRM state)H1-2.6B 147.42 154.77 156.37 100 mT along xPercent change 0.14% 0.50% 0.12%

H1-2.6B 147.68 154.93 156.63 200 mT along xPercent change 0.04% 0.60% 0.29%

H2-5.35B (Manlius Formation) 216.98 226.28 222.78 None (NRM state)H2-5.35B 216.59 225.94 221.99 1 T field along zPercent change 0.18% 0.15% 0.35%

kxx, kyy, kzz: diagonal elements of susceptibility tensor (mSI) in the specimen coordinate system

Fig. 4. (a) Day plot for Onondaga and Knox samples of Jackson (1990), together with new data for mid-continentsamples (Jackson et al. 1992) and for new Helderberg Group samples. (b) Recalculated parameters for the KnoxDolomite and Onondaga Limestone loops, using an approach-to-saturation fit to estimate Ms (see text).




each field step, the field was first applied in thepositive direction; the resulting remanence (Mr

+)was measured. The same field was then applied inthe negative direction and again the resulting rema-nence (Mr

2) was measured. For several samples theexperiment was carried out using a pulse magnetizerand a 2G superconducting magnetometer. For a fewsamples, an equivalent experiment was conducted,using a vibrating-sample magnetometer to measurea series of minor hysteresis loops with increasingpeak fields (of equal strength in the positive andnegative directions). In all cases, the differencebetween Mr

+ and Mr2 was negligible, thereby dem-

onstrating that the anisotropy of the particles inthe analysed Helderberg Group carbonates is over-whelmingly uniaxial (Fig. 5).

Superparamagnetic fraction

Although the SP population plays no role in carryingthe NRM, it is of considerable interest in understand-ing the origins and significance of the (presumably)cogenetic SSD particles that do carry the palaeomag-netic information in these rocks. Further, any quantitat-ive analysis and interpretation of in-field (hysteresisand susceptibility) measurements must properlyaccount for the SP fraction in order to isolate the con-tribution of the stable remanence carriers.

Magnetic properties change dramatically withincreasing size within the SP range and acrossthe SP–SSD threshold (or blocking volume), andare strongly temperature dependent (Neel 1949;Dunlop & Ozdemir 1997). SP particles that aremuch smaller than the blocking volume at a given

temperature remain in equilibrium with ambientfields, even those that alternate at high frequency; theytherefore have frequency-independent susceptibilitythat is fully in-phase, and not especially high(e.g. Worm 1998). During hysteresis experiments,these SP particles approach saturation slowly in rela-tively high fields (Tauxe et al. 1996). Larger particlesnear the SP–SSD threshold show more strongly time-or frequency-dependent magnetization on laboratorymeasurement timescales; susceptibilities are veryhigh with in-phase (k!) and quadrature (k!!) com-ponents that vary with frequency. These ‘viscoussuperparamagnetic’ (VSP) particles saturate in rela-tively low fields, thereby ‘shearing’ the loop of accom-panying stably magnetic phases during hysteresisexperiments to generate ‘wasp-waisted’ loops (Fig. 1b;Tauxe et al. 1996; Fabian 2003).

At room temperature, a relatively strong (10–20% per decade) frequency dependence of suscep-tibility and high ratios of xf/Ms (c. 50 mm A21)in previously studied remagnetized carbonates(Jackson et al. 1993) and in our new HelderbergGroup samples clearly indicates the presence of asignificant population of grains with Tb c. 295 K(for magnetite, this Tb implies grains with diametersof c. 15–25 nm; Worm 1998). During cooling, asthe size window for VSP behaviour shifts to finersizes, a strong frequency dependence of k! and a sig-nificant k! ! persist down to 20 K for the HelderbergGroup samples (Fig. 6), indicating a broad distri-bution of nanometric particle sizes.

Hysteresis loops measured on a QuantumDesigns MPMS, using maximum fields of 2.5 Tand temperatures from 300 K down to 10 K, showstrong, progressive increases in Mr and Bc oncooling (Fig. 7) as the SP fraction becomes ther-mally stable. The loop constriction that is so promi-nent at room temperature gradually gives way tomore ‘rectangular’ or ‘pot-bellied’ shapes (Tauxeet al. 1996; Fabian 2003). Quantitative loop analysisrequires initial separation of the ferrimagnetic signalfrom the dia/paramagnetic background signal.This becomes somewhat problematic at the lowesttemperatures however, where paramagnetic magne-tization becomes a distinctly non-linear function ofapplied field (especially in fields over 1 T). Further,antiferromagnetic nanoparticles of materials such asferrihydrite may contribute to low-temperaturehigh-field non-linearity. Here we use a consistentfitting method, in which the high-field interval istested for statistically significant deviations fromlinearity (Jackson & Sølheid 2010). An exponentialapproach-to-saturation law (Fabian 2006) is appliedif the linearity test fails. We recognize that thismethod overestimates Ms for the lowest tempera-tures, because it incorporates some of the non-linearparamagnetic and/or antiferromagnetic magnetiza-tion into the modelled ferrimagnetic loop; however

Mr (+)

Mr (-)

Fig. 5. Results of a test for multiaxial anisotropy basedon forward and reverse IRM acquisition (Fig. 2; Mitraet al. 2011). At each field step a positive field is applied,the resulting remanence (Mr

+) is measured, a negativefield of equal strength is applied and the remanence (Mr

2)is measured. Intensity differences for this sample of theManlius Formation of the Helderberg Group arenegligible, showing that the particle anisotropy isoverwhelmingly uniaxial.




this will not be critical to our analysis. For tempera-tures down to about 40 K, Ms changes by only a veryslight amount; Mr however increases to 3 times itsroom-temperature value and Bc increases by afactor of 6 (Figs 8a, b), both as a result of SP–SSDtransitions in progressively finer particles oncooling. The shape parameter shys defined byFabian (2003) quantifies the dramatic change fromstrongly wasp-waisted loop shapes to more rec-tangular or pot-bellied shapes at lower temperatures(Fig. 8c). The sharp increase in calculated Ms valuesbelow 40 K is no doubt partly due to the aforemen-tioned non-linear paramagnetic signal affecting theapproach-to-saturation calculations, but may alsobe partly due to ordering of some iron-bearingphase. Jackson & Worm (2001) saw evidence of

this ordering in low-temperature susceptibility datafor specimens of the Trenton Limestone.

Thermal demagnetization of a low-temperaturestrong-field isothermal remanence can be used,together with some assumptions, to quantify theSP content in a sample (Banerjee et al. 1993) andto calculate the size distribution of particles thatunblock while warming to room temperature (e.g.Worm & Jackson 1999). The key assumption isthat the loss of remanence is related to the SSD-SPtransition, rather than to multi-domain phenomenasuch as wall unpinning or domain reorganizationdue to the changes in temperature-dependent aniso-tropy constants (e.g. Moskowitz et al. 1998). Thisassumption is well justified for the remagne-tized carbonates that we are discussing, as detailed

(a)[m

3 kg

–1]

[m3

kg–1

]

(b)

Fig. 6. (a) In-phase and quadrature susceptibility at five frequencies for sample H2-4.4 (Manlius Formation, HelderbergGroup), measured from 20 to 300 K. Strong frequency dependence over the whole temperature range indicates abroad size distribution of nanoparticles (e.g. Worm 1998). (b) Expanded view of quadrature susceptibilities,compared to the derivative of in-phase susceptibility with respect to ln( f ).

(a) (b)

Fig. 7. (a) Low-temperature high-field hysteresis loops for sample H2-5.35 (Manlius Formation, Helderberg Group),uncorrected for dia- and paramagnetic background. (b) Expanded view of low-field portion of the loops, showingthe characteristic wasp-waisted shape at room temperature (light grey). This wasp-waisted shape dissipatessignificantly on cooling as the superparamagnetic particles become progressively blocked in.




below. For the analysed Helderberg Group samples,roughly three-quarters of the low-temperature rema-nence imposed isothermally by a 2.5 T field at 20 Kis erased by zero-field warming to room temper-ature. This result implies that approximately 75% ofthe ferrimagnetic material present is in the SP state at300 K and does not contribute to the ancient NRM(Fig. 9a). This result agrees well with the estimateof Dunlop (2002b) for the Onondaga and TrentonLimestones that was based on modelling of room-temperature hysteresis data. The uninflected demag-netization curve suggests a broad, unimodal sizedistribution of the SP particles, and simple calcu-lations using the method of Worm & Jackson (1999)make this explicit (Fig. 9b). The distribution peaksat or below 10224 m3 (approximately 10 nm diam-eter), in very good accord with the modelledSSD-SP mixing curves of Dunlop (2002b).

Stable SD fraction

Whereas in-field measurements such as hysteresisand susceptibility are strongly influenced by SP

particles, we can isolate the properties of the SSDand larger fraction by measuring remanent magneti-zations. IRM acquisition and AF demagnetizationcurves for the new Helderberg Group samples areessentially identical to those obtained previouslyfor the Onondaga and Trenton limestones (Jackson1990), intersecting at fields of about 50 mT, withcrossover parameters (Cisowski 1981) of R c. 0.5(Fig. 10). Roughly 10% of the 1 T SIRM survivesAF demagnetization at 200 mT. The acquisitioncurves strongly resemble those found by Elmoreet al. (2006) in Helderberg Group samples fromWest Virginia. They also exhibit some similarity tothose defined by unmixing analysis (Gong et al.2009) as the dominant component in remagnetizedCretaceous carbonates from the Organya Basin ofnorthern Spain, showing significant lack of satur-ation in 300 mT. Gong et al. (2009) interpreted thiscomponent as due to very fine magnetite, near theSP–SSD transition. Low-temperature cycling of aroom-temperature SIRM (Fig. 11) suggests the pres-ence of minor amounts of remanence-carryinghematite and pyrrhotite, with subtle inflections

(a) (b) (c)

Ms[

A m

2 kg

–1]

Mr[A

m2

kg–1

]Fig. 8. (a) Saturation remanence, (b) coercivity and (c) shape parameter as a function of temperature as calculatedfor the low-temperature hysteresis loops of sample H2-5.35 (Fig. 7). (a) Saturation remanence increases significantly oncooling, as SP particles become thermally stable. Saturation magnetization changes little, at least down to 40 K,which appears to be the ordering temperature of an unidentified iron-bearing phase. Loop shapes are constricted(positive value of s) down to 40 K, below which most of the ferrimagnetic signal comes from thermally stable carriers.

(a) (b)

M [A

m2

kg–1

]

Fig. 9. (a) Thermal demagnetization of low-temperature remanence for sample H2-4.4 shows progressive unblockingof the nanophase ferrimagnets, with no indication of the Verwey transition. FC remanence was imprinted by coolingin a 2.5 T field from 300 K to 20 K; ZFC remanence was imparted isothermally at 20 K by application and removalof a 2.5 T field (after zero-field cooling from 300 K). Both remanences were measured while warming in zerofield. (b) Particle-size distribution calculated for the ZFC curve in (a).




near the Morin transition (c. 250 K; e.g. Ozdemir &Dunlop 2006) and near the Besnus transition (32 K;Besnus 1966; Dekkers et al. 1989; Rochette et al.1990). However, the magnetization changes verylittle overall (less than 8%) during cooling andrewarming, with some small irreversible loss ofremanence. The major feature of the rewarmingcurve is a broad maximum between 150 and200 K, resembling the hallmark features noted foroxidized magnetite by Ozdemir & Dunlop (2010)and for pyrrhotite by Dekkers (1989).

Anhysteretic remanent magnetization (ARM)acquisition characteristics provide additional evi-dence that the remanence carriers in many remagne-tized carbonate units overwhelmingly comprisenon-interacting SSD particles. ARM intensityincreases rapidly as a function of DC bias field(Bdc), with ratios of ARM/SIRM reaching nearly30% for Bdc ! 0.2 mT (Fig. 12; see also Jacksonet al. 1992, 1993). The initial slope (calculated for0 " Bdc " 0.01 mT) is the normalized anhystereticsusceptibility xa/SIRM with a value of 2.1 mm A21,comparable to theoretical values for non-interactingSD magnetite (Egli & Lowrie 2002; Egli 2006),to values observed for cultured magnetotacticbacteria (Moskowitz et al. 1993) and for igneousmaterials such as the Tiva Canyon Tuff (Till et al.2010) and the Lambertville plagioclase (Dunlop &Ozdemir 1997, fig. 11.6) that are known to containdominantly non-interacting SSD magnetic particles.

To summarize, ARM and IRM characteristicstogether indicate that the thermally stable (non-SP) room-temperature remanence carriers in thenew sample set from the Helderberg Group, andfrom similar previously analysed remagnetizedcarbonates in the Appalachian Basin and the

mid-continental USA, are almost entirely in theSD state with very minimal contributions fromPSD and MD grains. These SSD particles aredominated by uniaxial anisotropy and constituteroughly a quarter of the ferrimagnetic material

Fig. 10. Acquisition and AF demagnetization of IRMfor sample H2-2.8 (Manlius Formation, HelderbergGroup) shows near-ideal non-interacting SD behaviour(Cisowski 1981). The crossover point between theacquisition and demagnetization curves is 54 mT,R ! 0.46.

M [A

m2

kg–1

]

Fig. 11. Low-temperature demagnetization (LTD) cyclefor H2-4.4. The sample was magnetized isothermally at300 K in a 2.5 T field, then measured in zero field whilecooling to 20 K and rewarming to room temperature. Thebroad maximum around 150–200 K resembles thatobserved by Ozdemir & Dunlop (2010) in oxidizedsubmicron magnetite; however, here we see no distinctindication of the Verwey transition. The Morin transition(250 K) and the Besnus transition (32 K) indicate smallremanence contributions from hematite andpyrrhotite, respectively.

0.00

0.05

0.10

0.15

0.20

0.25

0.30

0 0.02 0.04 0.06 0.08 0.10 0.12 0.14 0.16 0.18 0.2

ARM

/SIR

M

H1-0.2a

Fig. 12. Non-linear anhysteretic remanentmagnetization (ARM) acquisition for sample H1-0.2(Becraft Formation, Helderberg Group) showsnon-interacting SD behaviour (Dunlop & Ozdemir 1997,fig. 11.8; Egli & Lowrie 2002; Egli 2006). Initial slopegives xa/SIRM ! 2.1 # 1023 m A21, comparable totheoretical values for non-interacting SD magnetite (Egli& Lowrie 2002; Egli 2006), to values observed forcultured magnetotactic bacteria (Moskowitz et al. 1993)and for igneous materials such as the Tiva Canyon Tuff(Till et al. 2010) and the Lambertville plagioclase(Dunlop & Ozdemir 1997, fig. 11.6)




in these samples. Almost all of the rest of themagnetic mineralogy is superparamagnetic atroom temperature, as shown by low-temperaturehysteresis, low-temperature remanence, frequency-dependent susceptibility and mixing models ofroom-temperature hysteresis. From these exper-iments the distribution of domain states is welldefined, but specifics of the mineralogy remainelusive.

Ferrimagnetic mineral composition

Given the lack of pronounced diagnostic low-temperature transitions, the best remaining hopefor definitive identification of the magnetic min-eralogy is through measurements above room

temperature. Low-field high-temperature AC sus-ceptibility of the new Helderberg Group samples,measured on a KLY-2 Kappabridge in an argonatmosphere in a series of heating–cooling cycleswith successively higher peak temperatures, showsa weak and noisy signal but essentially reversiblebehaviour up to 400 8C followed by sharp increasesin susceptibility for higher-temperature cycles (Fig.13). This behaviour is virtually identical to thatobserved by Zegers et al. (2003), which they attrib-uted to formation of new magnetite by oxidationof pyrite in the temperature range 420–500 8C.It differs significantly from the results obtained forthe Trenton Limestone by Jackson & Worm (2001),in which significant mineral alteration also beganfor heating runs to maximum temperatures above400 8C, but the new mineral was most likely pyrrho-tite with a Curie temperature near 320 8C.

High-temperature hysteresis measurements(Fig. 14) also had rather marginal signal/noiseratios, due to the combination of weak intensitiesand small allowable sample sizes. Loops and back-field remanence curves were measured betweenroom temperature and 400 8C, in order to focus onthe naturally occurring ferrimagnetic populationand avoid the formation of new magnetic materialat higher temperatures. Saturation magnetizationdecreases by about 50% between 20 and 400 8C,significantly more than the drop of c. 38% expectedfor pure magnetite over this temperature interval.This may be due to a small contribution from ferri-magnetic pyrrhotite as a slight dip in the Ms(T)curve near 300 8C suggests; however, the data arenot of sufficient quality to allow decisive determi-nation. Nevertheless, the trend suggests that mono-clinic pyrrhotite is not a major phase in these

0

5.0E-08

1.0E-07

1.5E-07

2.0E-07

2.5E-07

3.0E-07

3.5E-07

4.0E-07

0 100 200 300 400 500 600 700

susc

eptib

ility

[m3 k

g–1]

Temperature [ºC]

heating

cooling

heating

cooling

heating

cooling

heating

cooling

Tmax = 300

Tmax = 400

Tmax = 500

Tmax = 600

H2-5.35c2

Fig. 13. Multicycle high-temperature susceptibilitymeasurements for sample H2-5.35 (Manlius Formation,Helderberg Group) show nearly reversible behaviourbelow 400 8C (with some decrease due to the 400 degreestep), followed by a major increase on heating to 500 8C.

H2-4.4b2

B [T]10.50-0.5-1

4.0E-3

3.0E-3

2.0E-3

1.0E-3

0

-1.0E-3

-2.0E-3

-3.0E-3

-4.0E-3

20ºC100ºC200ºC300ºC400ºC

Temperature [ºC]35030025020015010050

1.5E-3

1.0E-3

5.0E-4

Temperature [K]650600550500450400350300

2.5E-4

2.0E-4

1.5E-4

1.0E-4

5.0E-5

0

0

(a)(b)

(c)

Fig. 14. (a) High-temperature hysteresis loops for sample H2-4.4, measured on a Princeton Measurements MicroMagVSM (temperatures from 20 to 400 8C). (b) Calculated saturation magnetization and (c) saturation remanence asfunctions of temperature.




samples. Saturation remanence drops by nearly 80%over this temperature range, and the ferrimagneticpopulation is almost entirely in the SP state by400 8C. The backfield remanence curves (Fig. 15)show a corresponding progressive shift of the coer-civity spectrum to lower fields, with a decrease inBcr by roughly two-thirds over this temperaturerange.

The method of Lowrie (1990) for identifyingremanence carriers according to both coercivityand unblocking-temperature ranges yields similarresults (Fig. 16). Orthogonal IRMs were imprintedwith successively decreasing pulsed fields: 1 Talong the specimen z axis, 300 mT along x and100 mT along y. The two lower-field treatmentseach reoriented most of the remanence acquired inprevious steps (Fig. 16a), confirming that hard anti-ferromagnets such as hematite could only be makinga minor contribution to the IRM. Stepwise thermaldemagnetization shows that the three coercivity

fractions all unblock at approximately the samerate, with more than 80% of the remanence erasedby 400 8C and nearly 100% demagnetized by5008 (Fig. 16b). A slight change in slope in theintermediate-coercivity x component from 300 to350 8C may be due to pyrrhotite, but overall theremanence appears to reside in a magnetite-likephase with relatively low unblocking temperatures.

The low-temperature data clearly show that stoi-chiometric magnetite is nearly absent from thesesamples. Cation-deficient magnetite is a good can-didate to account for the low-T properties, but isless successful in accounting for Ms(T) aboveroom temperature since oxidation increases theCurie temperature (Ozdemir & Banerjee 1984).The observed Ms(T) trend suggests a Curie temp-erature below that of pure magnetite. Cation substi-tution could account for this, but would be quitesurprising as it would seem to be inconsistentwith a low-temperature chemical origin for the

H2-4.4b2

Field [T]0.0-0.1-0.2-0.3

Field [mT](a) (b)0-50-100-150-200-250-300

Rem

anen

t Mag

netiz

atio

n

3.0e-4

2.0e-4

1.0e-4

0.0e0

-1.0e-4

-2.0e-4

-3.0e-4

20 ºC100 ºC200 ºC300 ºC400 ºC

Temperature [K]650600550500450400350300

Temperature [C]35030025020015010050

B cr [m

T]

40.00

30.00

20.00

10.00

0.00

Bcr

Fig. 15. (a) High-temperature back-field remanence curves for sample H2-4.4, measured on a Princeton MeasurementsMicroMag VSM (temperatures from 20 to 400 8C). (b) Remanent coercivity determined from the back-fieldremanence curves as a function of temperature.

-1.0E-06

0.0E+00

1.0E-06

2.0E-06

3.0E-06

4.0E-06

5.0E-06

6.0E-06

7.0E-06

8.0E-06

02004006008001000

mom

ent [

Am2 ]

mom

ent [

Am2 ]

H2-5.75bZ (1000 mT)X (300 mT)Y (100 mT)

0.0E+00

1.0E-06

2.0E-06

3.0E-06

4.0E-06

5.0E-06

6.0E-06

0 100 200 300 400 500 600Temperature [ºC]

H2-5.75bZ (1000 mT)X (300 mT)Y (100 mT)

(a) (b)

Fig. 16. (a) Acquisition and (b) thermal demagnetization of a three-component IRM for sample H2-5.75 (ManliusFormation, Helderberg Group). Acquisition involved application of a 1 T pulsed field along the specimen’s z axis,followed by a 300 mT field along the x axis and a 100 mT field along the y axis. The two latter treatments each reorientmost of the remanence, thus shown to be carried predominantly by relatively soft (ferrimagnetic) phases. The threecomponents are removed at roughly the same rates by thermal demagnetization, and are almost completely removed by500 8C. A dip in the intermediate-coercivity x component at 300–350 8C may be due to pyrrhotite.




dominant magnetic phase. On the other hand, Xuet al. (1998) found magnetite grains of clearlyauthigenic origin that contained minor amounts ofsubstituted Ti and Mn (x , 0.15 in Fe3-xTixO4) inthe Leadville carbonates of central Colorado.Such compositions would suppress the Verweytransition and also depress the Curie temperatureand unblocking temperatures. However, authigeniccation-substituted magnetite has not (to our knowl-edge) been found in other studies of remagnetizedcarbonates. Probably the best way to explain boththe low- and high-temperature data is through amixture of maghematized magnetite and smalleramounts of monoclinic pyrrhotite nanoparticles,but cation-substituted authigenic magnetite mayalso be important.

Discussion

Do hysteresis properties provide reliable indicationsof the presence or absence of substantial chemicaloverprints in carbonate rocks? Certainly in situgrowth of a significant population of SP–SSD par-ticles should be expected to exert a strong influ-ence on both the NRM and on the rock-magneticproperties, and a number of studies have con-firmed this relationship. But how often might weencounter ‘false positives’ (hysteresis ratios alongthe trend defined by the remagnetized Appalachianbasin carbonates, in rocks where the characteristicremanence is primary) or ‘false negatives’ (hyster-esis ratios along the trend for SD–MD magnetites,in rocks that are substantially or completely remag-netized)? How might we recognize these situations?Given the variety of possible remagnetization mech-anisms and triggers (e.g. Elmore 2001), as well asthe variety of possible primary iron phases in car-bonates, there are a number of important pointsthat have to be considered.

Primary magnetic particles in carbonates:

sources, characteristics, preservation

Major proportions of SP material are relatively rarein geological materials, with notable exceptionsmostly including those in which the SP materialformed in situ: soils and palaeosols (e.g. Mullins1977; Maher & Taylor 1988; Dearing et al. 1997;Guyodo et al. 2006); rapidly cooled volcanics (e.g.Schlinger et al. 1991; Gee & Kent 1999; Till et al.2010); and marine or lacustrine sediments where SPmagnetite is produced in the sediment column bydissimilatory iron-reducing bacteria (e.g. Moskowitzet al. 1989; Maloof et al. 2007). The high surfacearea to volume ratios of SP magnetite likely contrib-utetotheirgeochemical instability indetritalenviron-ments (e.g. Li et al. 2009). Carbonates with primary

magnetization from which hysteresis data havebeen obtained reveal a trend on the compiled Dayplot of hysteresis parameters (Fig. 1) that indicatesmagnetite populations dominated by SD and MDparticles without significant SP contributions.These units are predominantly deepwater pelagiclimestones (Laytonville Limestone, Maiolica Lime-stone, ODP 728) and, in such lithologies, an SD–MD grain size distribution is consistent with detritalinput of magnetite and either a lack of SP input oralteration of any SP particles. However, in shallow-water carbonate depositional environments that areisolated from aqueous detrital input and wherethere is early cementation, is it possible that therecould be significant contributions of both ultrafineSP and SSD ferrimagnets? If so, they may yieldfalse positive indications of remagnetization. Inthe absence of aqueous detrital input there are twomain potential sources of magnetic minerals to car-bonate environments: biogenic precipitation andaeolian delivery (including extraterrestrial fluxes).

Freeman (1986) documented the common occur-rence of ‘cosmic spherules’ in magnetic extractsfrom pelagic limestones by detailed SEM work,but was unable to quantify the proportions of ultra-fine material in these extraterrestrial magnetic par-ticles. It is not uncommon for airborne dustcollections to show significant populations of SP–SSD ferrimagnets, especially when no major localdust sources are present (e.g. Oldfield et al. 1985).Magnetic material found in dust layers in the Green-land and Antarctic ice cores contained signifi-cant proportions of SP–SSD nanoparticles whoseorigins were ascribed to ablation of meteorites inEarth’s upper atmosphere (Lanci & Kent 2006;Lanci et al. 2007, 2008). Detrital inputs of ferrimag-netic nanoparticles could therefore potentiallymimic the hysteresis properities of the SP–SSDgrains that are produced through authigenic pro-cesses associated with remagnetization.

Extracellular formation of magnetite duringmicrobial iron reduction can lead to the in situ for-mation of prodigious amounts of SP magnetite dur-ing deposition (Frankel 1987; Maloof et al. 2007).Combined with SD grains formed by magnetotacticbacteria during or immediately following deposi-tion (e.g. Kopp & Kirschvink 2008; Moskowitzet al. 2008), these populations could also give falseindications of remagnetization based on hysteresisproperties. With the current published datasets pre-dominantly coming from deepwater carbonates, thepossibility that there can be preservation of a sig-nificant primary SP population in early cementedshallow-water carbonates is difficult to evaluate.This reality points to the need for more detailed hys-teresis data from shallow-water carbonates to seeif the preservation of primary SP grains can have asignificant influence on hysteresis parameters.




Primary detrital magnetic oxide grains in the PSDand MD size range can moreover be diminished toSP–SSD sizes by reductive dissolution (e.g.Karlin & Levi 1983; Tarduno 1995; Smirnov &Tarduno 2001). New iron sulphide minerals (predo-minantly greigite) commonly also form in the sedi-ment column during early diagenetic sulphatereduction (Maloof et al. 2007; Roberts et al.2011). The combination of such sulphides withresidual magnetite could conceivably mimic thehysteresis signature of the remagnetized carbonates,whether or not the new sulphides carry a significantportion of the remanence.

It has long been known that constricted (‘wasp-waisted’) hysteresis loops and high coercivityratios can arise from mixtures of hard (e.g. imperfectantiferromagnets such as hematite and goethite) andsoft (e.g. ferrimagnets such as magnetite andmaghemite) phases (e.g. Nagata & Carleton 1987).There are many situations imaginable (e.g. detritalmagnetite and late weathering products such asgoethite) where Day plot locations and constrictedloops might incorrectly suggest an ancient remagne-tization. In the case of a mix of soft ferrimagneticand hard antiferromagnetic carriers, values of Bc

and Bcr would be higher than in the case of a domi-nant SP–SSD population. Additionally, if the loopconstriction is due to multiple mineralogies theremay be discernible inflections in the gradient ofmagnetization acquisition and backfield demagneti-zation of an SIRM. Given the common occurrenceof goethite as a surface weathering product andthe precipitation of secondary iron sulphidesthrough fluid flow or early diagenesis, interpret-ations of hysteresis loop parameters should bemade cautiously and be paired with other rock mag-netic experiments that can demonstrate a largelymono-mineralic population of ferrimagnetic grains.

Sources of iron for formation of new magnetic

phases and mechanisms and triggers for

authigenic iron oxide growth

Oxidation of pyrite or other sulphides to magnetitein remagnetized carbonates has been documentedthrough detailed analytical microscopy (e.g. Fruitet al. 1995; Weil & Van der Voo 2002). Early diage-netic pyrite framboids, aggregates up to 100 mm indiameter comprised of large numbers of small(,1 mm) crystallites, have frequently been foundin North American Palaeozoic carbonates to becompletely altered by later diagenesis to magnetite(Suk et al. 1990a, b, 1991, 1993). Suk & Halgedahl(1996) found a variety of other iron-oxide spheroidmorphologies that may have formed in this way orby different processes (Freeman 1986), many ofwhich exhibited the MD magnetic behaviour

expected for particles of that size but some ofwhich had PSD or SSD hysteresis properties; afew even had wasp-waisted loops. They arguedthat despite the large particle dimensions, suchspheroids may be the major remanence carriers inthe remagnetized carbonates of the AppalachianBasin and elsewhere. This argument is difficult torefute, especially if we recognize that the spheroidpopulation recovered by magnetic extraction fromthe insoluble residue is likely to be strongly biasedin favour of the larger and more highly magneticparticles, and that therefore the spherules foundwith relatively low Ms (and high Mr/Ms) may bethe most representative of those in the bulk rock.However, it is not clear that the preponderance ofSP magnetite in bulk-rock remagnetized carbon-ate samples has any direct association with pre-existing sulphides or with the observed spheroids.Replacement of iron sulphides by magnetite is cer-tainly involved in many cases and may be a chiefmechanism of remagnetization when large-scalefluid flow is involved. It remains difficult to pre-dict with any certainty that a distinctive hyster-esis signature must necessarily accompany thisprocess, so the possibility of false negative testscannot be ruled out. Suk & Halgedahl (1996)found ‘normal’ unconstricted hysteresis loops inremagnetized Onondaga Limestone samples fromwestern New York, suggesting that SP magnetitewas not formed by the remagnetization process ornot preserved.

Many studies have convincingly demonstrated aconnection between remagnetization and clay dia-genesis, particularly the conversion of smectite toillite (Jackson et al. 1988; Hirt et al. 1993; Katzet al. 1998; Elliott et al. 2006a, b; Tohver et al.2008). Illitization of smectite occurs at relativelylow temperatures (between c. 60 8C and 120 8C)and liberates both water and cations such as iron,creating conditions suitable for the de novo for-mation of magnetite or other remanence-carryingminerals. Nucleation and growth processes can beexpected to produce nanoparticle populations withsize distributions that are ultimately governed byfactors including element availability, diffusionrates, reactivity, temperature and time. The growthof magnetite through this process is likely toproduce populations that span the SP–SSD sizerange. In such a scenario, the formation of bothabundant SP grains and the grains that reached thecritical size for thermal stability (and thus acqui-sition of a secondary remanence) can be attributedto a single chemical overprinting event. Thisprocess seems to us to be the situation in whichunique hysteresis properties may most reliablyindicate strong chemical remagnetizations analo-gous to those seen in North American Palaeozoiccarbonates.




The rock-magnetic signature of

remagnetization in carbonate rocks

Given the wide variety of possible remagnetizationmechanisms, parent phases and final carriers, andthe range of scenarios discussed above for potentialfalse positives and false negatives, it is some-what surprising that a rock-magnetic signature ofremagnetization would have any generality. This iseven more true now that we have eliminated one ofthe proposed rationales for a distinctive hysteresissignature, based on multiaxial individual-particleanisotropy in equidimensional authigenic magne-tites. However, many examples are now availablefor which demonstrably remagnetized units havesome or all of the putative rock-magnetic finger-prints (e.g. Channell & McCabe 1994; McCabe &Channell 1994; Molina Garza & Zijderveld 1996;Suk & Halgedahl 1996; Banerjee et al. 1997;Butler et al. 1997; Xu et al. 1998; d’Agrella Filhoet al. 2000; Dinares-Turell & Garcia-Senz 2000;Enkin et al. 2000; Zegers et al. 2003; Trindadeet al. 2004; Zwing et al. 2005; Elmore et al. 2006;Soto et al. 2008). We are not aware of any clear-cutfalse positives (hysteresis characteristics indicatingcomplete remagnetization in carbonate rocks bear-ing a primary NRM) and very few false negatives(hysteresis behaviour indicating primary NRMcarriers in fully chemically overprinted carbonaterocks). However, there is a number of rock unitswith ‘intermediate’ or ambiguous hysteresis charac-teristics, some of which clearly have multi-mineralicremanence-carrying populations (e.g. Weil & Vander Voo 2002; Zegers et al. 2003; Zwing et al.2005; Oliva-Urcia & Pueyo 2007).

Zwing et al. (2005) noted a strong association oflithology and hysteresis properties of remagnetizedrocks: reef carbonates, with low primary detritalcontent, exhibited hysteresis properties similar tothose of the archetypal remagnetized carbonates;platform carbonate samples had intermediate hys-teresis ratios; and remagnetized siliciclastics hadhysteresis properties suggestive of primary detritalmineralogy. They reached the reasonable con-clusion that a remagnetization test based on hyster-esis behaviour only works when authigenic phasesrepresent an overwhelming majority of the magneticmineralogy, which in most cases means that detritalphases cannot be present in any great abundance.Indeed, our compilation in Figure 1 also appearsto show a systematic dependence on depositionalenvironment, with samples that plot along the‘primary’ SD–MD mixing line coming largelyfrom pelagic limestones (Channell & McCabe1994; Tarduno & Myers 1994; Abrajevitch &Kodama 2009) while the Pleistocene Tahiti reef(Menabreaz et al. 2010) had significant inputof volcaniclastic sands. In each of these cases, it

appears magnetite from a detrital or bacterialsource is holding a primary remanence.

In our view, the most diagnostic indicator of mag-netite authigenesis in carbonate rocks is the occur-rence of a unimodal particle-size distribution thatpeaks well below the room-temperature SP–SSDthreshold. Such distributions are relatively rare ingeological materials and, when they do occur, arealmost always associated with in situ processes oforigin (nucleation and growth in volcanic glass,metabolic electron-transfer reactions mediatedextracellularly by bacteria, pedogenesis or variouscandidate processes in carbonate rocks). Particlesthat grow through the SP–SSD threshold acquire astable CRM, and the population characteristicsaccount for the distinctive hysteresis behaviour,high ARM/SIRM ratios and elevated frequency-dependence of susceptibility.Further work isneeded,however, to evaluate whether a similar particle-sizedistribution could arise in the primary magneticmineralogy of shallow-water carbonates that are iso-lated from detrital input.

Summary and conclusions

In the prototypical remagnetized carbonates thathave been the principal focus of this study, boththe ChRM and the rock-magnetic properties areintimately related to the SP–SSD nanoparticlepopulation that formed in situ during the Permian,many tens of millions of years after the stratawere deposited. Clay diagenesis and maturationof organic matter, driven by moderately elevatedburial temperatures as well as changes in geochem-ical conditions possibly due to introduction oftectonically driven brines, resulted in neoformationof these nanoparticles and alteration of pre-existingiron sulphides. Large proportions of the authi-genic magnetite never grew to sizes larger than20 nm; these SP particles are responsible for thefrequency-dependent susceptibility and, to a largeextent, for the wasp-waisted hysteresis loops andstrongly elevated Hcr/Hc ratios that constitute theproposed rock-magnetic fingerprint of remagneti-zation. Some fraction of the authigenic magnetitegrew through the SP–SSD threshold, acquiringthe ChRM and accounting for the elevated ARM/SIRM ratios that help to identify the carriers asauthigenic. In the Helderberg Group samples ofthis study, some 75% of the magnetite is superpara-magnetic at room temperature, and almost all ofthe rest is SSD. The exact mineralogical compo-sition remains unclear, but it is certainly not puremagnetite; some small degree of cation substitu-tion or a larger degree of cation deficiency isrequired for the almost complete lack of aVerwey transition.




One of the original explanations offered for acorrelation between hysteresis properties andstrong chemical remagnetization involved authi-genic SD particles having an almost complete lackof shape anisotropy (i.e. spheroidal morphologiesresembling those observed in large extracted mag-netite particles), and therefore having magneticproperties governed primarily by the cubic magne-tocrystalline (multiaxial) anisotropy of magnetite(Jackson 1990). We have now shown unequivocallythat this is not correct for the newly analysedsamples of the Helderberg Group. Two newapproaches (Lanci 2010; Mitra et al. 2011) appliedto these samples to test for the presence of multiax-ial anisotropy have revealed that the magnetiteparticles are instead controlled by uniaxial aniso-tropy. The fundamental basis of the distinctive hys-teresis behaviour of these remagnetized carbonatesis the dominant contribution from the SP and SSDsize fractions.

It should be borne in mind that although there isa strong link between rock-magnetic behaviour andsize distributions in these rocks, the link betweenparticle sizes and a late authigenic origin is moreindirect. It is conceivable that dominant SP–SSDsize fractions could also originate biogenically orby detrital input, and that rocks with a primary orvery early ChRM may therefore sometimes exhibitthe rock-magnetic characteristics associated withremagnetization. At this time, we are unaware ofany such ‘false positives’ in the published literaturewhere carbonates with primary magnetization yieldhysteresis parameters in the range typicallyrestricted to remagnetized carbonates. Nevertheless,there is a paucity of hysteresis data from shallow-water carbonates with primary ChRMs, and dataare needed from such units to further evaluatewhether the primary coexistence of SP and SSDgrains ever leads to false positives for remagnetiza-tion. ‘False negative’ results (in which remagnetizedunits have the rock-magnetic characteristic of unre-magnetized carbonates) appear to be more common,when detrital materials occur in sufficient abun-dance to shift the peak of the size distribution intothe PSD–MD range.

We thank A. Hirt and an anonymous reviewer for theirhelpful suggestions and S. Swanson-Hysell for assistancewith fieldwork. This is contribution 1106 of the Institutefor Rock Magnetism, which is supported by grants fromthe Instruments and Facilities Program, Division of EarthScience, National Science Foundation.

References

Abrajevitch, A. & Kodama, K. 2009. Biochemicalv. detrital mechanism of remanence acquisition inmarine carbonates: A lesson from the K-T boundary

interval. Earth and Planetary Science Letters, 286,269–277.

Aubourg, C. & Pozzi, J. P. 2010. Toward a new ,250 8Cpyrrhotite-magnetite geothermometer for claystones,250 8C pyrrhotite-magnetite geothermometer forclaystones. Earth and Planetary Science Letters, 294,47–57.

Bachtadse, V., Van der Voo, R., Hayes, F. M. &Kesler, S. E. 1987. Late Paleozoic remagnetizationof mineralized and unmineralized Ordovician carbon-ates from east Tennessee: evidence for a post-orechemical event. Journal of Geophysical Research B:Solid Earth, 92, 14 165–14 176.

Banerjee, S. K., Hunt, C. P. & Liu, X.-M. 1993.Separation of local signals from the regional paleo-monsoon record of the Chinese loess plateau: a rock-magnetic approach. Geophysical Research Letters,20, 843–846.

Banerjee, S., Elmore, R. D. & Engel, M. H. 1997.Chemical remagnetization and burial diagenesis:testing the hypothesis in the Pennsylvanian BeldenFormation, Colorado. Journal of GeophysicalResearch, 102, 24 825–24 842.

Belkaaloul, N. K. & Aissaoui, D. M. 1997. Natureand origin of magnetic minerals within the MiddleJurassic shallow-water carbonate rocks of the ParisBasin, France: implications for magnetostratigraphicdating. Geophysical Journal International, 130,411–421.

Besnus, M. J. 1966. Proprietes magnetiques de la pyrrho-tite naturelle. PhD thesis, University of Strasbourg.

Bethke, C. M. & Marshak, S. 1990. Brine migrationsacross North America: the plate tectonics of ground-water. Annual Reviews of Earth and PlanetarySciences, 18, 287–315.

Blumstein, A. M., Elmore, R. D., Engel, M. H., Elliot,C. & Basu, A. 2004. Paleomagnetic dating of burialdiagenesis in Mississippian carbonates, Utah. Journalof Geophysical Research, 109, 1–16.

Butler, R. F., Gehrels, G. E. & Bazard, D. R. 1997.Paleomagnetism of Paleozoic strata of the Alexanderterrane, southeastern Alaska. GSA Bulletin, 109,1372–1388.

Channell, J. E. T. & McCabe, C. 1994. Comparison ofmagnetic hysteresis parameters of unremagnetizedand remagnetized limestones. Journal of GeophysicalResearch B: Solid Earth, 99, 4613–4623.

Cisowski, S. 1981. Interacting v. non-interacting single-domain behavior in natural and synthetic samples.Physics of the Earth and Planetary Interiors, 26,77–83.

D’Agrella Filho, M. S., Babinski, M., Trindade,R. I. F., van Schmus, W. R. & Ernesto, M. 2000.Simultaneous remagnetization and U–Pb isotoperesetting in Neoproterozoic carbonates of the Sao Fran-cisco Craton, Brazil. Precambrian Research, 99,179–196.

Day, R., Fuller, M. & Schmidt, V. A. 1977. Hysteresisproperties of titanomagnetites: grain-size and compo-sitional dependence. Physics of the Earth and Plane-tary Interiors, 13, 260–266.

Dearing, J. A., Bird, P. M., Dann, R. J. L. & Benjamin,S. F. 1997. Secondary ferrimagnetic minerals in Welshsoils: a comparison of mineral magnetic detection




methods and implications for mineral formation. Geo-physical Journal International, 130, 727–736.

Dekkers, M. J. 1989. Magnetic properties of natural pyr-rhotite. II. High- and low-temperature behavior of Jrsand TRM as a function of grain size. Physics of theEarth and Planetary Interiors, 57, 266–283.

Dekkers, M. J., Mattei, J.-L., Fillion, G. & Rochette,P. 1989. Grain-size dependence of the magneticbehavior of pyrrhotite during its low-temperature tran-sition at 34 K. Geophysical Research Letters, 16,855–858.

Dinares-Turell, J. & Garcia-Senz, J. 2000. Remagne-tization of Lower Cretaceous limestones from thesouthern Pyrenees and relation to the Iberian plate geo-dynamic evolution. Journal of Geophysical Research,105, 19 405–19 418.

Dunlop, D. J. 2002a. Theory and application of the Dayplot (Mrs/Ms v. Hcr/Hc). 1. Theoretical curves andtests using titanomagnetite data. Journal of Geophysi-cal Research, 107, doi:10.1029/2001JB000487.

Dunlop, D. J. 2002b. Theory and application of the Dayplot (Mrs/Ms v. Hcr/Hc). 2. Application to data forrocks, sediments, and soils. Journal of GeophysicalResearch, 107, 1. doi:10.1029/2001JB000486.

Dunlop, D. J. & Ozdemir, O. 1997. Rock Magnetism.Fundamentals and Frontiers. Cambridge UniversityPress, Cambridge.

Dunlop, D. J., Ozdemir, O. & Schmidt, P. W. 1997a.Paleomagnetism and paleothermometry of theSydney Basin. 2. Origin of anomalously high unblock-ing temperatures. Journal of Geophysical Research,102, 27285–27295.

Dunlop, D. J., Schmidt, P. W., Ozdemir, O. & Clark,D. A. 1997b. Paleomagnetism and paleothermometryof the Sydney Basin. 1. Thermoviscous and chemicaloverprinting of the Milton Monzonite. Journal of Geo-physical Research, 102, 27271–27283.

Egli, R. 2006. Theoretical considerations on the anhys-teretic remanent magnetization of interacting particleswith uniaxial anisotropy. Journal of GeophysicalResearch B. Solid Earth, 111, doi:10.1029/2006JB004577.

Egli, R. & Lowrie, W. 2002. Anhysteretic remanent mag-netization of fine magnetic particles. Journal of Geo-physical Research-Solid Earth, 107, 2209, doi:10.1029/2001JB000671.

Elliott, W. C., Basu, A., Wampler, J. M., Elmore, R. D.& Grathoff, G. H. 2006a. Comparison of K-Ar agesof diagenetic illite-smectite to the age of a chemicalremanent magnetization (CRM): an example fromthe Isle of Skye, Scotland. Clays and Clay Minerals,54, 314–323.

Elliott, W. C., Osborn, S. G., O’Brien, V. J., Elmore,R. D., Engel, M. H. & Wampler, J. M. 2006b. On thetiming and causes of illite formation and remagnetiza-tion in the Cretaceous Marias River Shale, DisturbedBelt, Montana. Journal of Geochemical Exploration,89, 92–95.

Elmore, R. D. 2001. A review of palaeomagnetic data onthe timing and origin of multiple fluid-flow events inthe Arbuckle Mountains, southern Oklahoma. Pet-roleum Geoscience, 7, 223–229.

Elmore, R. D. & McCabe, C. 1991. The occurrence andorigin of remagnetization in the sedimentary rocks of

North America. Reviews of Geophysics, 29 (Suppl.),377–383.

Elmore, R. D., London, D., Bagley, D., Fruit, D. &Gao, G. 1993. Remagnetization by basinal fluids:testing the hypothesis in the Viola Limestone, southernOklahoma. Journal of Geophysical Research B, SolidEarth, 98, 6237–6254.

Elmore, R. D., Foucher, J. L. E., Evans, M., Lewchuk,M. & Cox, E. 2006. Remagnetization of the Tonolo-way Formation and the Helderberg Group in theCentral Appalachians: testing the origin of syntiltingmagnetizations. Geophysical Journal International,166, 1062–1076.

Enkin, R. J., Osadetz, K. G., Baker, J. & Kisilevsky, D.2000. Orogenic remagnetizations in the Front Rangesand Inner Foothills of the southern Canadian Cordil-lera: Chemical harbinger and thermal handmaiden ofCordilleran deformation. Geological Society ofAmerica Bulletin, 112, 929–942.

Everitt, C. W. F. & Clegg, J. A. 1962. A field test ofpalaeomagnetic stability. Geophysical Journal of theRoyal Astronomical Society, 6, 312–319.

Fabian, K. 2003. Some additional parameters to estimatedomain state from isothermal magnetization measure-ments. Earth and Planetary Science Letters, 213,337–345.

Fabian, K. 2006. Approach to saturation analysis of hys-teresis measurements in rock magnetism and evidencefor stress dominated magnetic anisotropy in youngmid-ocean ridge basalt. Physics of the Earth and Pla-netary Interiors, 154, 299–307.

Fisher, D. W. 1987. Lower Devonian limestone, Helder-berg Escarpment. In: ROY, D. C. (ed.) GeologicalSociety of America Centennial Field Guide – North-eastern Section. Geological Society of America,Boulder, 119–122.

Fisher, D. W., Isachsen, Y. W. & Rickard, L. V. 1970.Geologic Map of New York State (1:250000)New York State Museum Map and Chart Series No. 15.

Frankel, R. B. 1987. Anaerobes pumping iron. Nature,330, 208.

Freeman, R. 1986. Magnetic mineralogy of pelagic lime-stones. Geophysical Journal of the Royal AstronomicalSociety, 85, 433–452.

Fruit, D., Elmore, R. D. & Halgedahl, S. 1995. Remag-netization of the folded Belden formation, northwestColorado. Journal of Geophysical Research B: SolidEarth, 100, 15009–15024.

Garven, G. 1995. Continental-scale groundwater flow andgeological processes. Annual Review of Earth and Pla-netary Sciences, 24, 89–117.

Gee, J. & Kent, D. V. 1995. Magnetic hysteresis in youngmid-ocean ridge basalts: dominant cubic anisotropy?Geophysical Research Letters, 22, 551–554.

Gee, J. & Kent, D. V. 1999. Calibration of magnetic gran-ulometric trends in oceanic basalts. Earth and Plane-tary Science Letters, 170, 377–390.

Gong, Z., Dekkers, M. J., Heslop, D. & Mullender,T. A. T. 2009. End-member modelling of isothermalremanent magnetization (IRM) acquisition curves: anovel approach to diagnose remagnetization. Geophy-sical Journal International, 178, 693–701.

Goree, W. S. & Fuller, M. D. 1976. Magnetometersusing r-f driven SQUIDs and their application in rock




magnetism and paleomagnetism. Reviews of Geophy-sics and Space Physics, 14, 591–608.

Graham, J. W. 1949. The stability and significance ofmagnetism in sedimentary rocks. Journal of Geophysi-cal Research, 54, 131–167.

Guyodo, Y., LaPara, T. M., Anschutz, A. J., Penn,R. L., Banerjee, S. K., Geiss, C. E. & Zanner, W.2006. Rock magnetic, chemical and bacterial commu-nity analysis of a modern soil from Nebraska. Earthand Planetary Science Letters, 251, 168–178.

Hirt, A. M., Banin, A. & Gehring, A. U. 1993. Thermalgeneration of ferromagnetic minerals from iron-enriched smectites. Geophysical Journal International,115, 1161–1168.

Jackson, M. J. 1990. Diagenetic sources of stable rema-nence in remagnetized Paleozoic cratonic carbonates:a rock magnetic study. Journal of GeophysicalResearch B: Solid Earth, 95, 2753–2761.

Jackson, M. J. & Worm, H.-U. 2001. Anomalous unblock-ing temperatures, viscosity and frequency-dependentsusceptibility in the chemically-remagnetized TrentonLimestone. Physics of the Earth and PlanetaryInteriors, 126, 27–42.

Jackson, M. J. & Sølheid, P. 2010. On the quantitativeanalysis and evaluation of magnetic hysteresis data,Geochemistry, Geophysics. Geosystems, 11, doi:10.1029/2009GC002932.

Jackson, M. J., McCabe, C., Ballard, M. M. & Van derVoo, R. 1988. Magnetite authigenesis and diageneticpaleotemperatures across the northern AppalachianBasin. Geology, 16, 592–595.

Jackson, M. J., Worm, H.-U. & Banerjee, S. K. 1990.Fourier analysis of digital hysteresis data: rock mag-netic applications. Physics of the Earth and PlanetaryInteriors, 65, 78–87.

Jackson, M. J., Sun, W.-W. & Craddock, J. P. 1992. Therock magnetic fingerprint of chemical remagnetizationin midcontinental Paleozoic carbonates. GeophysicalResearch Letters, 19, 781–784.

Jackson, M. J., Rochette, P., Fillion, G., Banerjee, S.K. & Marvin, J. A. 1993. Rock magnetism ofremagnetized Paleozoic carbonates: low-temperaturebehavior and susceptibility characteristics. Journalof Geophysical Research B: Solid Earth, 98,6217–6225.

Karlin, R. & Levi, S. 1983. Diagenesis of magnetic min-erals in recent haemipelagic sediments. Nature, 303,327–330.

Katz, B., Elmore, R. D. & Engel, M. H. 1998. Authigen-esis of magnetite in organic-rich sediment next to adike: implications for thermoviscous and chemicalremagnetizations. Earth and Planetary ScienceLetters, 163, 221–234.

Katz, B., Elmore, R., Cogoini, M., Engel, M. & Ferry,S. 2000. Associations between burial diagenesis ofsmectite, chemical remagnetization, and magnetiteauthigenesis in the Vocontian trough, SE France.Journal of Geophysical Research, 105, 1.

Kent, D. V. 1979. Paleomagnetism of the DevonianOnondaga limestone revisited. Journal of GeophysicalResearch B: Solid Earth, 84, 3576–3588.

Kent, D. V. 1985. Thermoviscous remagnetization insome Appalachian limestones. Geophysical ResearchLetters, 12, 805–808.

Kirschvink, J. L. 1978. The Precambrian-Cambrianboundary problem: paleomagnetic directions from theAmadeus Basin, Central Australia. Earth and Plane-tary Science Letters, 40, 91–100.

Kligfield, R. & Channell, J. E. T. 1981. Widespreadremagnetization of Helvetic limestones. Journal ofGeophysical Research, 86, 1888–1900.

Kopp, R. E. & Kirschvink, J. L. 2008. The identificationand biogeochemical interpretation of fossil magneto-tactic bacteria. Earth-Science Reviews, 86, 42–61.

Lanci, L. 2010. Detection of multi-axial magnetite byremanence effect on anisotropy of magnetic susce-ptibility. Geophysical Journal International, 181,1362–1366.

Lanci, L. & Kent, D. V. 2003. Introduction of thermalactivation in forward modeling of hysteresis loops forsingle-domain magnetic particles and implications forthe interpretation of the Day diagram. Journal of Geo-physical Research, 108, doi:10.1029/2001JB000944.

Lanci, L. & Kent, D. V. 2006. Meteoric smoke falloutrevealed by superparamagnetism in Greenlandice. Geophysical Research Letters, 33, doi:10.1029/2006GL026480.

Lanci, L., Kent, D. V. & Biscaye, P. E. 2007. Meteoricsmoke concentration in the Vostok ice core esti-mated from superparamagnetic relaxation and someconsequences for estimates of Earth accretion rate.Geophysical Research Letters, 34, doi: 10.1029/2007gl029811.

Lanci, L., Delmonte, B., Maggi, V., Petit, J. R. &Kent, D. V. 2008. Ice magnetization in the EPICA-Dome C ice core: Implication for dust sources duringglacial and interglacial periods. Journal of Geophysi-cal Research, 113, D14207, doi:14210.11029/12007JD009678.

Li, Y.-L., Pfiffner, S. M. et al. 2009. Degeneration ofbiogenic superparamagnetic magnetite. Geobiology,7, 25–34.

Lowrie, W. 1990. Identification of ferromagnetic mineralsin a rock by coercivity and unblocking temperatureproperties. Geophysical Research Letters, 17, 159–162.

Lu, G., McCabe, C., Hanor, J. S. & Ferrell, R. E. 1991.A genetic link between remagnetization and potassicmetasomatism in the Devonian Onondaga formation,northern Appalachian basin. Geophysical ResearchLetters, 18, 2047–2050.

Maher, B. A. & Taylor, R. M. 1988. Formation ofultrafine-grained magnetite in soils. Nature, 336,368–371.

Maloof, A. C., Kopp, R. E. et al. 2007. Sedimentary ironcycling and the origin and preservation of magnetiza-tion in platform carbonate muds, Andros Island,Bahamas. Earth and Planetary Science Letters, 259,581–598.

McCabe, C. & Elmore, R. D. 1989. The occurrence andorigin of Late Paleozoic remagnetization in the sedi-mentary rocks of North America. Reviews of Geophy-sics, 27, 471–494.

McCabe, C. & Channell, J. E. T. 1994. Late Paleo-zoic remagnetization in limestones of the Cravenbasin (northern England) and the rock magnetic finger-print of remagnetized sedimentary carbonates. Jour-nal of Geophysical Research B: Solid Earth, 99,4603–4612.




McCabe, C., Van der Voo, R., Peacor, D. R., Scotese,C. R. & Freeman, R. 1983. Diagenetic magnetite car-ries ancient yet secondary remanence in some Paleo-zoic sedimentary carbonates. Geology, 11, 221–223.

McCabe, C., Van der Voo, R. & Ballard, M. M. 1984.Late Paleozoic remagnetization of the Trenton lime-stone. Geophysical Research Letters, 11, 979–982.

McElhinny, M. W. & Opdyke, N. D. 1973. Remagneti-zation hypothesis discounted: a paleomagnetic studyof the Trenton limestone, New York State. GeologicalSociety of America Bulletin, 84, 3697–3708.

Menabreaz, L., Thouveny, N., Camoin, G. & Lund,S. P. 2010. Paleomagnetic record of the late Pleisto-cene reef sequence of Tahiti (French Polynesia): a con-tribution to the chronology of the deposits. Earth andPlanetary Science Letters, 294, 58–68.

Mitra, R., Tauxe, L. & Gee, J. 2011. Detecting uniaxialsingle domain grains with a modified IRM technique.Geophysical Journal International, 187, 1250–1258,doi: 10.1111/j.1365-246X.2011.05224.x.

Molina Garza, R. S. & Zijderveld, J. D. A. 1996.Paleomagnetism of Paleozoic strata, Brabant andArdennes Massifs, Belgium: implications of prefoldingand postfolding Late Carboniferous secondary magne-tizations for European apparent polar wander. Journalof Geophysical Research, B, Solid Earth and Planets,101, 15799–15818.

Moskowitz, B. M., Frankel, R. B., Bazylinski, D. A.,Jannasch, H. W. & Lovley, D. R. 1989. A compari-son of magnetite particles produced anaerobically bymagnetotactic and dissimilatory iron-reducing bac-teria. Geophysical Research Letters, 16, 665–668.

Moskowitz, B. M., Frankel, R. & Bazylinski, D. 1993.Rock magnetic criteria for the detection of biogenicmagnetite. Earth and Planetary Science Letters, 120,283–300.

Moskowitz, B. M., Jackson, M. & Kissel, C. 1998. Low-temperature magnetic behavior of titanomagnetites.Earth and Planetary Science Letters, 157, 141–149.

Moskowitz, B. M., Bazylinski, D. A., Egli, R.,Frankel, R. B. & Edwards, K. J. 2008. Magneticproperties of marine magnetotactic bacteria in a sea-sonally stratified coastal pond (Salt Pond, MA, USA).Geophysical Journal International, 174, 75–92.

Mullins, C. E. 1977. Magnetic susceptibility of the soiland its significance in soil science – a review.Journal of Soil Science, 28, 294–306.

Muttoni, G. 1995. ‘Wasp-waisted’ hysteresis loops froma pyrrhotite and magnetite-bearing remagnetizedTriassic limestone. Geophysical Research Letters, 22,3167–3170.

Muttoni, G. & Kent, D. V. 1994. Paleomagnetism ofLatest Anisian (Middle Triassic) Sections of the PrezzoLimestone and the Buchenstein Formation, SouthernAlps, Italy. Earth and Planetary Science Letters, 122,1–18.

Nagata, T. & Carleton, B. J. 1987. Magnetic remanencecoercivity of rocks. Journal of Geomagnetism andGeoelectricity, 39, 447–461.

Neel, L. 1949. Theorie du traınage magnetique des ferro-magnetiques en grains fins avec applications aux terrescuites. Annales de Geophysique, 5, 99–136.

Oldfield, F., Hunt, A., Jones, M. D. H., Chester, R.,Dearing, J. A., Olsson, L. & Prospero, J. M. 1985.

Magnetic differentiation of atmospheric dusts.Nature, 317, 516–518.

Oliva-Urcia, B. & Pueyo, E. L. 2007. Rotational base-ment kinematics deduced from remagnetized coverrocks (Internal Sierras, southwestern Pyrenees). Tec-tonics, 26, 1.

Oliva-Urcia, B., Pueyo, E. L. & Larrasoana, J. C. 2008.Magnetic reorientation induced by pressure solution: apotential mechanism for orogenic-scale remagnetiza-tions. Earth and Planetary Science Letters, 265,525–534.

Oliver, J. 1986. Fluids expelled tectonically from oro-genic belts: their role in hydrocarbon migration andother geological phenomena. Geology, 14, 99–102.

Ozdemir, O. & Banerjee, S. K. 1984. High temperaturestability of maghemite (g-Fe2O3). GeophysicalResearch Letters, 11, 161–164.

Ozdemir, O. & Dunlop, D. J. 2006. Magnetic memoryand coupling between spin-canted and defect magnet-ism in hematite. Journal of Geophysical Research-Solid Earth, 111, 1.

Ozdemir, O. & Dunlop, D. J. 2010. Hallmarks of maghe-mitization in low-temperature remanence cycling ofpartially oxidized magnetite nanoparticles. Journal ofGeophysical Research B: Solid Earth, 115, 1,doi:10.1029/2009JB006756.

Ozdemir, O., Dunlop, D. J. & Moskowitz, B. M. 1993.The effect of oxidation of the Verwey transition inmagnetite. Geophysical Research Letters, 20,1671–1674.

Potter, D. K. & Stephenson, A. 1988. Single-domainparticles in rocks and magnetic fabric analysis. Geo-physical Research Letters, 15, 1097–1100.

Reynolds, R. L. 1990. Paleomagnetism – a polished viewof remagnetization. Nature, 345, 579–580.

Ripperdan, R. L., Riciputi, L. R., Cole, D. R., Elmore,R. D., Banerjee, S. & Engel, M. H. 1998. Oxygenisotope ratios in authigenic magnetites from theBelden Formation, Colorado. Journal of GeophysicalResearch, 103, 1.

Roberts, A. P., Cui, Y.-L. & Verosub, K. L. 1995. Wasp-waisted hysteresis loops: mineral magnetic char-acteristics and discrimination of components inmixed magnetic systems. Journal of GeophysicalResearch B: Solid Earth, 100, 17909–17924.

Roberts, A. P., Chang, L., Rowan, C. J., Horng, C.-S. &Florindo, F. 2011. Magnetic properties of sedimen-tary greigite (Fe3S4). An update: Reviews of Geophy-sics, 49, RG1002.

Rochette, P., Fillion, G., Mattei, J.-L. & Dekkers,M. J. 1990. Magnetic transition at 30–34 Kelvin inpyrrhotite: insight into a widespread occurrence ofthis mineral in rocks. Earth and Planetary ScienceLetters, 98, 319–328.

Schlinger, C. M., Veblen, D. R. & Rosenbaum, J. G.1991. Magnetism and magnetic mineralogy of ashflow tuffs from Yucca Mountain, Nevada. Journal ofGeophysical Research B: Solid Earth, 96, 6035–6052.

Scotese, C. R., Van der Voo, R. & McCabe, C. 1982.Paleomagnetism of the Upper Silurian and LowerDevonian carbonates of New York State: evidencefor secondary magnetizations residing in magnetite.Physics of the Earth and Planetary Interiors, 30,385–395.




Smirnov, A. V. & Tarduno, J. A. 2001. Estimating super-paramagnetism in marine sediments with the timedependency of coercivity of remanence. Journal ofGeophysical Research, 106, 16135–16143.

Soto, R., Villalain, J. J. & Casas-Sainz, A. M. 2008.Remagnetizations as a tool to analyze the tectonichistory of inverted sedimentary basins: a case studyfrom the Basque-Cantabrian basin (north Spain). Tec-tonics, 27, doi: 10.1029/2007TC002208.

Stoner, E. C. & Wohlfarth, E. P. 1948. A mechanism ofmagnetic hysteresis in heterogeneous alloys. Philoso-phical Transactions of the Royal Society of London,Series A, 240, 599–602.

Suk, D.-W. & Halgedahl, S. L. 1996. Hysteresis proper-ties of magnetic spherules and whole-rock specimensfrom some Paleozoic platform carbonate rocks.Journal of Geophysical Research B: Solid Earth,101, 25053–25076.

Suk, D.-W., Peacor, D. R. & Van der Voo, R. 1990a.Replacement of pyrite framboids by magnetite in lime-stone and implications for paleomagnetism. Nature,345, 611–613.

Suk, D.-W., Van der Voo, R. & Peacor, D. R. 1990b.Scanning and transmission electron microscopeobservations of magnetite and other iron phases inOrdovician carbonates from east Tennessee. Journalof Geophysical Research B. Solid Earth, 95,12327–12336.

Suk, D., Voo, R. V. D. & Peacor, D. R. 1991. SEM/STEM observations of magnetite in carbonates ofeastern North America: evidence for chemical remag-netization during the Alleghenian Orogeny. Geophysi-cal Research Letters, 18, 939–942.

Suk, D.-W., Van der Voo, R. & Peacor, D. R. 1993.Origin of magnetite responsible for remagnetizationof early Paleozoic limestones of New York State.Journal of Geophysical Research B: Solid Earth, 98,419–434.

Sun, W.-W. & Jackson, M. J. 1994. Scanning electronmicroscopy and rock magnetic studies of magnetic car-riers in remagnetized Early Paleozoic carbonates fromMissouri. Journal of Geophysical Research B: SolidEarth, 99, 2935–2942.

Symons, D. T. A., Kawasaki, K. & Pannalal, S. J. 2010.Paleomagnetic mapping of the regional fluid flowevent that mineralized the Upper Mississippi ValleyZn-Pb ore district, Wisconsin, USA. Journal of Geo-chemical Exploration, 106, 188–196.

Tarduno, J. A. 1995. Superparamagnetism and reductiondiagenesis in pelagic sediments: enhancement ordepletion? Geophysical Research Letters, 22,1337–1340.

Tarduno, J. A. & Myers, M. 1994. A primary magnetiza-tion fingerprint from the Cretaceous Laytonville Lime-stone: further evidence for rapid oceanic platevelocities. Journal of Geophysical Research B, SolidEarth, 99, 21691–21703.

Tauxe, L., Mullender, T. A. T. & Pick, T. 1996. otbel-lies, wasp-waists, and superparamagnetism in mag-netic hysteresis. Journal of Geophysical Research B,Solid Earth, 101, 571–583.

Till, J. L., Jackson, M. J., Rosenbaum, J. G. & Solheid,P. 2010. Magnetic properties in an ashflow tuff with

continuous grain size variation: A natural referencefor magnetic particle granulometry. Geochemistry,Geophysics, Geosystems, 12 (Q07Z26), doi: 10.1029/2011GC003648.

Tohver, E., Weil, A. B., Solum, J. G. & Hall, C. M. 2008.Direct dating of carbonate remagnetization by 40Ar/39Ar analysis of the smectite–illite transformation.Earth and Planetary Science Letters, 274, 524–530.

Trindade, R. I. F., D’Agrella, M. S., Babinski, M.,Font, E. & Neves, B. B. B. 2004. Paleomagnetismand geochronology of the Bebedouro cap carbonate:evidence for continental-scale Cambrian remagnetiza-tion in the Sao Francisco craton, Brazil. PrecambrianResearch, 128, 83–103.

Van der Voo, R. 1990. The reliability of paleomagneticdata. Tectonophysics, 184, 1–9.

Verwey, E. J. & Haayman, P. W. 1941. Electronic con-ductivity and transition point in magnetite. Physica,8, 979–982.

Walton, D. 1980. Time-temperature relations in the mag-netization of assmblies of single-domain grains.Nature, 286, 245–247.

Walz, F. 2002. The Verwey transition – a topical review.Journal of Physics: Condensed Matter, 14, R285–340.

Weil, A. B. & Van der Voo, R. 2002. Insights into themechanism for orogen-related carbonate remagnetiza-tion from growth of authigenic Fe-oxide: A scanningelectron microscopy and rock magnetic study of Devo-nian carbonates from northern Spain. Journal ofGeophysical Research, 107, 1, doi:10.1029/2001JB000200.

Wohlfarth, E. P. & Tonge, D. G. 1957. The remanentmagnetization of single-domain ferromagnetic par-ticles. Philosophical Magazine, 2, 1333–1344.

Wolfers, P., Fillion, G., Ouladdiaf, B., Ballou, R. &Rochette, P. 2011. The Pyrrhotite 32 K magnetictransition. Solid State Phenomena, 170, 174–179.

Woods, S. D., Elmore, R. D. & Engel, M. H. 2002.Paleomagnetic dating of the smectite-to-illite conver-sion: testing the hypothesis in Jurassic sedimentaryrocks, Skye, Scotland. Journal of GeophysicalResearch, 107, doi: 10.1029/2000JB000053.

Worm, H.-U. 1998. On the superparamagnetic-stablesingle domain transition for magnetite, and frequencydependence of susceptibility. Geophysical JournalInternational, 133, 201–206.

Worm, H.-U. & Jackson, M. 1999. The superparamagnet-ism of Yucca Mountain Tuff. Journal of GeophysicalResearch B: Solid Earth, 104, 25415–25425.

Xu, W., Van der Voo, R. & Peacor, D. R. 1998. Electronmicroscopic and rock magnetic study of remagnetizedLeadville carbonates, central Colorado. Tectonophy-sics, 296, 333–362.

Zegers, T. E., Dekkers, M. J. & Bailly, S. 2003. LateCarboniferous to Permian remagnetization of Devo-nian limestones in the Ardennes: role of temperature,fluids, and deformation. Journal of GeophysicalResearch, 108, doi: 10.1029/2002JB002213.

Zwing, A., Matzka, J., Bachtadse, V. & Soffel, H. C.2005. Rock magnetic properties of remagnetizedPalaeozoic clastic and carbonate rocks from the NERhenish massif, Germany. Geophysical Journal Inter-national, 160, 477–486.



Documents

Rock magnetism of remagnetized carbonate rocks: another lookgilder/Re... · 1976), carbonate rocks represented a new fron-tier for palaeomagnetic research. These and other weakly