LITHOGEOCHEMISTRY IN MINERAL EXPLORATION...chimie de la roche favorable et de La nature des...

15. LITHOGEOCHEMISTRY IN MINERAL EXPLORAnON

G.J.S. GovettUni versi ty of New Brunswick,Fredericton, New Brunswick

Ian NicholQueen's University, Kingston, Ontario

Govett, G.J.S. and Nichol, Ian, Lithogeochemistry in mineral exploration; in Geophysics andGeochemistry in the Search of Metallic Ores; Peter J. Hood, editor; GeolOgical Survey ofCanada, Economic Geology Report 31, p. 339-362, 1979.

Abstract

Lithogeochemistry, as used in this presentation, is defined as the determination of thechemical composition of bedrock material with the objective of detecting distribution patterns ofelements that are spatially related to mineralization.

Mineralogical alteration zones in host rocks around mineral deposits have long beenrecognized and used as indicators of ore. Such alteration zones are the visible manifestations ofphysical and chemical changes in the host rocks resulting from either primary reactionsassociated with ore formation or subsequent secondary reactions between the ore and the hostrocks. Chemical alteration halos may be more intense, and therefore detectable, over greaterdistances than mineralogical halos, since the lattice substitution of elements may be detectedchemically without having any mineralogical representation. The scale and intensity of changesin the chemistry of the host rock is a function of the genesis of the ore, the chemistry of the hostrock, and the nature of the secondary processes. Appreciation of these factors is fundamental tothe successful application of lithogeochemistry to mineral exploration.

Lithogeochemistry has application at three levels of exploration: identification ofgeochemical provinces, favourable ore horizons, plutons or volcanic horizons on a regionalreconnaissance scale; recognition of local halos related to individual deposits on a localreconnaissance or follow-up scale; and wall-rock anomalies related to particular ore-shoots on amine scale.

Lithogeochemistry has been applied on a regional scale to the identification of mineralizedareas in granitic and gneissic terrane (Sn, W, Mo, Cu, and U); in areas of basic intrusions (Ni, Cr,and PO and basic volcanic rocks (Cu); and in areas of sedimentary rocks (Cu, Pb-Zn, and Au-U).In other cases favourable geological environments have been recognized on the basis of somediagnostic geochemical parameter of lithological units spatially, and probably genetically,associated with mineralization. For example, productive greenstone belts can be distinguishedfrom nonproductive greenstone belts, and cycles of volcanism containing significant massivesulphide deposits can be distinguished from equivalent nonproductive cyc les on the basis ofenhanced background contents of certain elements; areas of black shale horizons enriched inchalcophile elements may be readily identified, and these horizons may be spatially associatedwith more favourable horizons - such as quartzite and limestone - for the emplacement ofmineralization.

On a more local scale, geochemical dispersion halos of both major and minor elements havebeen shown to be associated with porphyry copper deposits. Similarly, depletion of Na and Ca,and enhancement of Fe, Mg, and K - together with characteristic distribution patterns of a widerange of trace elements - appear to be a general local-scale feature associated withvolcanogenic massive sulphide deposits. Wall-rock dispersion halos have proved useful in detailedexploration, although their nature is very much dependent on local geological conditions.

Various techniques have been employed to detect bedrock anomalies related tomineralization; these include whole rock analysis, measurement of water-soluble elements(especially halogens), and analysis of mineral separates. In some cases multivariate statisticaltechniques have been used to identifY otherwise unrecognized, subtle features of the data or as ameans of enhancing very weak anomalies.

The results cited are examples where lithogeochemistry has been or could have been usedsuccessfully for the location of particular types of mineralization. They are also indications ofthe diversified role lithogeochemistry could have in mineral exploration.

Exploration requirements in the future will demand techniques to find mineral deposits insituations where current methods have failed. In particular, presently available routinegeochemical methods are unsuitable, or have limited application, for the location of deeplyburied or blind deposits - these are the types of deposits that probably will become commontargets for mineral exploration as the more easily detected near-surface targets are exhausted.The subtle lithogeochemical anomalies and element zoning found in the cap-rocks of some typesof blind deposits offer scope for their detection, and may also be useful in drill-hole control.

The application of lithogeochemistry is largely a function of the availability of adequatesamples either from suitable bedrock exposure or drill core; obviously, it must have limitedpotential in areas of deep weathering and areas of extensive overburden or soil cover. However,within these constraints, recent developments offer convincing evidence of the potential powerof lithogeochemistry as an exploration method.

340 G.J.S. Govett and I. Nichol

Re!limp

Dans ce rapport, on definiL la lithogeochimie comme la determination de la composil.ionchimique de la roche en place; par cette methode, on peut dpceler Ie mode de distribution deselemenLs dans l'espace, dans la mesure ou la distribution est liee Q la mineralisation.

On a reconnu il y a longtemps l'existence de zones d'alteration mineralogique dans les rochesfavorables qui renferment les gz1es mineraux; on les utilise comme "indicateurs" du minerai. Ceszones d'alteration sont les manifestations visibles des modifical.ions physiques et chimiques quesubissent les roches favorables par suite des reactions secondaires ulLerieures qui ont eu lieu entreIe minerai et les roches favorables. Les aureoles d'alteration chimique sont parfois plus intenses,par consequent plus facilement d(;celables, et cela, sur de plus grandes distances que les aureolesmineralogiques, puisque toute Slbslitlltion d'elements apparaz1 Q l'analyse chimique, sans qu'il soiLnecessaire d'effectuer des observations mineralogiques. L'etendue et l'intensite des variations dela composition chimique de la roche favorable dependent du mode de genese du minerai, de lachimie de la roche favorable et de La nature des processus secondaires. nest essentiel d'evaluerchacun de ces facteurs, pour pouvoir appliquer avec succes la liLhogeochimie Q la rechercheminiere.

La lithogeochimie s'applique Q Lroix niveaux d'exploration: l'identification des provincesgeochimiques, des horizons mineralises favol'ables, des plutons au des horizons volcaniques, a uneechelle de reconnaissance I'egionale; l'identification des aureoles locales liees a des g[1esparticuliers, a l'echelle de la reconnaissance locale ou des travaux d'exploration detaillee; etl'etude des anomalies qui caracterisenL la roche encaissante et sont li(;es a l'existence de colonnesmineralisees, cela a l'echelle de l'exploitation miniere.

On a applique, a l'echelle regionale, la lithogeochimie a l'identification de zones mineraliseesdans des terrains granitiques et gneissiques (Sn, W, Mo, Cu et U); dans des zones d'intrusionsbasiques (Ni, Cr et PO et de roches volcaniques basiques (Cu); et enf'in dans des zones de rochessedimentaires (Cu, Pb-Zn, et Au-U). Dans d'autres cas, on a idenLifiP" des milieux geologiquesfavorables, en fonction de certains parametl'es gP"ochimiques diagnostiques caracterisant desunites lithologiques spatialement et sans doute genetiquement associees a une mineralisation. Parexemple, on peut distinguer les zones de roches vertes productives des zones equivalentes nonproductives, et distinguer les cycles volcaniques pendant lesquels se sont constitues d'importantsgz1es sulfureux massifs, des cycles equivalents non productifs, apres accentuation des valeurs defond de cerLains elements; on peut facilement reconnaz1re les secteurs Q horizons d'argile liteenoire enrichis en elements chalcophiles, lesqllels peuvent etre spatialement associes a des horizonsplus favorables - tels que quartzites et caLcaires - pour la mise en place de mineralisations.

A une echelle plus locale, les aureoles de dispersion geochimique des elements importants etsecondaires sont parfois, comme on l'a demontre, associees a des gz1es porphyriques de cuivre. Dememe, l'appauvrissement en Na et Ca, et L'enrichissement en Fe, Mg et K - en meme temps que Iemode de distribution typique d'une vasLe gamme d'elements-traces - semblent constituer uncaractere local du a la presence de gz1es suL{ureux massifs, de types volcanogenique. Les aureolesde dispersion dans la roche encaissante se sont averees utiles pour l'exploration detaillee, bien queleur nature depende beaucoup plus des conditions geologiques locales.

On a employe diverses techniques pour deceler les anomalies que prP"sente la roche en placedu point de vue de la mineralisation; parmi ces techniques, citons l'analyse de la roche entiere, lamesure des eLements solubles dans l'eau (en particulier les halogenes), et l'analyse des diversesfractions minerales. Dans certains cas, on a employe des methodes utilisant plusieurs variablesstatistiques, pour mettre en relief des rP"sult.ats moins evidents, jusque La ignores, ou bien desanomalies t/'es faibles.

Les /'esulta ts cites representent des cas ou la lithogeochimie a ou aumit pu donne/' desresultats satis(aisants pour localiser certains types de mineralisation. ns indiquent aussi Ie rolemultiple que pourrait jouer la lithogeochimie dans l'exploration miniere.

A l'avenir, il sera necessaire d'employer de nouvelles techniques d'exploration la ou lesmethodes habituellement ont echoue. En particulier, les mPthodes geochimiques courammentemployees actuellement sont inadequates, ou n'ont que des applications resLreintes, pour lalocalisation des gz1es profondement enfouis ou tout simplement dissimules - c'esL-Q-dire les sortesde gz1es qUi probablement constitueront les principaux objectifs d'exploration, a mesure ques'epuiseront les mineralisations proches de la Slrface, plus facilement detectables. Les anomalieslithogeochimiques de faible intensite, et la zonalite des elements qui caracte/'isent Ie chapeau(cap-rock) de certains types de gz1es dissimulP"s ravorisent la detection de ces gz1es, et peuventsans douLe pe/'mett/'e un meilleur choix de l'emplacement des trous de forage.

L'usage de la lithogeochimie depend largement de la mesure dans laquelle on peut obteni/'des echantillons adeqllats, provenant soit d'affleurements favorables de la roche en place, soit decarottes de forage; il est evident que la lithogeochimie offre moins de possibilites, dans les zonesd'alteration profondes, et celles cachees par de vastes terrains de couverture ou une impo/'tantecouverture de sol. Cependant, malgre ces limitations, les developpements recents demontrent defaqon convaincante Ie potentiel de la lithogeochimie en tant que methode d'exploration.

Lithogeochem istry 341

INTRODUCTION

Lithogeochemistry, or rock geochemistry, as used inthis paper, is defined as the study of the chemicalcomposi tion of bedrock with particular reference to thesearch for ore deposi ts. Geochemical patterns in bedrockthat can be related to mineralization are referred to asprimary patterns or primary dispersion, irrespective ofwhether the reactions causing the element distributionpatterns are syngenetic or epigenetic (James, 1967).

Mineralogical alteration zones in host rocks aroundmineral deposits have long been recognized and used asindicators of ore. Such alteration zones are the visiblemanifestations of physical and chemical changes in rocksassociated with mineralization. Chemical alteration halosma y be more extensi ve and therefore detectable over greaterdistances than mineralogical halos, since lattice substitutionof elements may occur without any mineralogical change.Recognition of this possibility and the need to developtechniques capable of detecting deeply buried and blinddeposits and to improve exploration capabili ties have led tothe recent widespread interest in lithogeochemistry.

The improved understanding of ore genesis during thelast decade is also a significant factor in the increasedinterest in lithogeochemical techniques. The general featuresof porphyry ore deposi ts have been characterized in terms ofpre-ore host rock, igneous host rocks, the orebody, hypogenealteration, mineralization and occurrence of sulphides (Lowelland Guilbert, 1970). Whereas a considerable degree ofsimilarity exists among the deposits, aspects in the detailedgeological history of a particular deposit (e.g., nature offluids, degree of fracturing and postmineralization history)might be expected to give rise to significant differencesbetween geochemical responses and mineralization.Understanding of the genesis of volcanic-sedimentary massivesulphide deposits has also vastly improved over the pastdecade; they are generally regarded as having originated frommetal-bearing fumarolic exhalations on the sea floor thatwere coeval with the associated volcanic or sedimentaryactivity (Sangster, 1972). Notwithstanding a common genesis,detailed variations in the character of the host rock,tectonic, and postmineralization history of di fferent depositscould be expected to result in variations between thegeochemical responses in the wall rock and mineralization.

The chemical composition of a sample of rock reflectsthe entire geological history of the rock. For example, thecomposition of a sample of a particular granite may reflectthe composition of the primary magma (possibly modified byassimilation of country rock), metasomatic and postmagmaticprocesses, metamorphism, and, in the case of surfacesamples, the effects of weathering. The extent to which theeffect of a mineralizing event is recognizable in thecomposition of the host rock depends upon the actualprocesses that affected the rock. It is thus of criticalimportance in the evaluation of any Ii tho geochemical data toconsider the nature and effects of all processes that mayhave affected the sample.

In view of the importance of geological environment onthe nature of geochemical dispersion associated withmineralization the main body of this review is divided intotwo parts: (1) geochemical response in intrusi ve rocks; and(2) geochemical response in extrusive and sedimentary rocks.Within each geological group, geochemical response isconsidered in terms of exploration scale:

Regional - large-scale, geochemical responses that arecapable of discriminating between productive andbarren terrain.

Local and Mine Scale - geochemical responses aroundindividual deposits that can be detected up to 1 to 2 kmfrom a deposit, and the geochemical responses that arelimited to the immediate wall rock of a deposit.

This classification is necessarily somewhat arbitrary, andthere is some overlap between the two scales ofinvestigation.

The objective of the paper is to present a state-of-theart review of Ii thogeochemistry; the source of much of thedata, therefore, is published papers, augmented by ourpersonal experiences. The works that we cite were carriedout in many di fferent ways - in terms of sampling, analysis,data presentation and, indeed, with many differentobjectives. Our lack of first-hand knowledge of many of theareas makes comparisons and evaluation difficult; tominimize some of the problems and to achieve moremeaningful comparisons, some of the data have beenreplotted from the published work and, in so doing, we trustthat neither the character of the original data nor the intentof the authors have not been misrepresented.

A number of investigations have been concerned withthe distribution of stable isotopes (sulphur, oxygen, hydrogen,and lead) associated with various types of deposits andmineral provinces. The results of these investigations are notconsidered here, although stable isotope studies may providea better understanding of the genesis of mineral deposits information that would be useful in selecting areas forexploration. In a limited number of cases, isotope zoningaround mining districts has been noted.

Li tho geochemistry has recei ved consi derable attentionwithin the Soviet Union. Although reports indicate notablesuccess and some innovative approaches, the work, in general,is difficult to evaluate and discussion is limited to aconsideration of some of the principles of the Soviet work.

The technical viability of lithogeochemistry in mineralexploration is dependent on the existence of somegeochemical signature in the rocks associated withmineralization which normally requires a geneticrelationship between the geochemical patterns and the oreforming processes. A major factor affecting the applicabili tyof lithogeochemistry in exploration is the extent to whichfeatures are of general occurrence. The economic viabili ty isrelated to the cost effectiveness of the procedure in relationto other techniques, and is dependent on the nature of theappropriate sampling, analytical and interpretationalprocedures.

GEOCHEMICAL RESPONSE TO MINERAL DEPOSITSIN INTRUSIVE ROCKS

Regional Scale

Prior to the present upsurge of interest in lithogeochemistry, geologists and geochemists have sought criteria todiscriminate between productive and barren intrusions on aregional scale. Tin seems to be the only element thatgenerally - but not universally - shows enrichment inintrusions associated with tin mineralization in many parts ofthe world. Quartz-cassiterite-bearing granite bodies have Snvariations in the range 20-30 ppm, compared to 3-5 ppm inbarren intrusions (Barsukov, 1967). Granite associated withtin deposits in Transbaikaliya (U.S.S.R.) has 16-32 ppm Sn,compared to barren granite that has less them 5 ppm Sn(Ivanova, 1963).

Tin-bearing granjte of northeast Queensland has beenshown to have significantly higher Sn contents thannonstanniferous granite (Sheraton and Black, 1973). A granitewith a high proportion of samples with Sn contents abovebackground values appears a criterion of potential tin


Table 15.1

Variations in composition of mineralized andunmineralized intrusives - Allen et aJ., 1976

Mineralized Unminerallzed

KzO x 2.31 1.47a 0.80 1.12I~ 1.12 - 2.77 0.28 - 5.36

Cu x 1450 57a 1785 73F<. 17 - 7400 o - 350

CaO x 3.58 4.85a 1.17 1.80R 2.0 - 3.62 0.50 - 9.28

NazO x 2.27 3.67a J .40 1.82R 0.50 - 4.52 0.22 - 6.48

MnO x 0.12 0.18a 0.04 0.09R 0.07 - 0.21 0.06 - 0.42

Zn x 3G 92a 14 67R 17 - 70 3D - 270

n 2(J 2/

20 ppm have associated tin mineralization (Fig. 15.0; thisfinding is consistent with most of the work carried out in theU.S.S.R.

Many workers have observed that, aside from tin, thereis no noticeable correlation between the abundance ofelements in granitic rocks and the presence or absence Dfmineralization; the lack of correlation is documented for leadand zinc (Tauson, 1967), and for lead, zinc, molybdenum, andtungsten (B2rsukov, 1967). In the investigation of mineralizedqranitic rocks of northeast Queerlsland referred to earlier, norelation between the lead, zinc, and copper contents of thegranite and lead/zinc and copper minerali7ation was noted,(Sheraton and Bl<Jck, 1973). This feature was ascribed to themineralization rresulting from processes unrelated to thedifferentiation of the intrusive.

Un the other hand, in si tuations where there has beenhydrothermal alteration, significant differences in meanelement content have been observed between mineralized andbarren intrusions. For example, the intrusion that hosts theCoed-y-Brenin copper porphyry-type deposit in North Wales(U.K.) is enriched in KzO, Cll, and Mo, and is depleted inMnO, NazO, CaO, and Zn relative to similar but barrenintrusions in the same region on the basis of a comparison ofmean element content (Allen ct al., 1976). With theexception of Cu and Zn, the differences in mean content areslight, and there is considerable overlap in the ranges of themineralized <Jnd unmincralized populations (see Table 15.1).The authors drew attention to the fact that the distributionof the data within the data sets is not norm31 and thus thestatistical dota should be treated with caution. Theenhancement and depletion effects are interpreted as areflection of the effects of mineralization and nssociatedhydrothermal alteration. The KzO, NazO, and CaO variationsare associated with phyllic alteration, whereas the Zndepletion possibly reflects Zn migration to the pyrite andpropylitic zone.

In certain cases, the form of the frequency distributionof an elrment in a region or a particular rock unit is morediagnostic of mineralization than the mean value of theelement (Be us, 19(,9; Bolotnikov and Kravchenko, 1970;

n-84

Unmineralized60

~ 40>-vCQI~

tT~ 20

L.L.

60 Mineralized

n-34

~ 40>-...CQI~

l1' 20QI......

o -t'--,--,,-.'-I-'--'1---"---'---l'------

o 4 12 20Sn ppm

mineralization. The variation in tin contcnt of granite withmineralization is ascribed to a primary um"ven distribution oftin in the crust or mantle that was the magma source.Although the areal extent of the granite bodies is nut givr-nthe Sn content of individual intrusions was characterized onthe basis of an 3verage of some 25 samples per intrusive butranging from 6 to 73 samples for individual bodies. None ofthe minerallzed granite is uniformally hilJh in tin and thecharacteristic feature is a very variable tin content amongstsamples necessitating that sampling must be reasonablyextensi ve if the tin-bearing potential is to be reallsticallyiJssessed. On the basis of a wider study of granitoids from theTasman geosyncline in Eastern Australia, granitic rocksassociated with cassiterite mineralization have a higher meanSn corltent (26 ppm) rclativr~ to similar but barren granitoids0.4 ppm) (Hesp and Rigby, 1975). The samples werecollected from outcrops some distance from locations wherecassiterite mineralization was observed to avoid the effect ofthe latter on Sn values. However, thl3 extent to which thehigher Sn values 8re characteristic of the mineralized graniteas a whole is not known as the sampling was biased to someextent towards areas of Sn mineralization rather than beingrepresentative of the rjranitoid rocks as a whole (Hesp andRigby, 1975). Although there are some exceptions to theg[~neralizatiDn, particularly in the low range of Snconcentration, all the granitoid samples that have more than

4 12 20Snppm

Figure 15.1. Frequency distribution of Sn in mineralizedand barren granitic intrusions, Tasman Geosyncline, EasternAustralia (data from Hesp and Rigby, 1975).

Li thogeochemistry 343

Backgroundn-126

Porphyry-type mineralizationn=24

2 5 10 20 >20ppm

2 5 10 20 >20ppm

O¥--"--i~-'-r'~+-''-'-T---''---',.--"---'-'

o

oo

20

40

80

60

>-

~ 20:::l0CD.....

0.fJ>u

; 40:::lITCD.....

Skarn - type mineralizationn-53

60

80

20

stock was used to characterize specific stocks. In the case ofbiotite, higher mean Rb and Sn contents and lower Srcontents are characteristic of mineralized granite relative tobarren granite. In feldspars, mean contents of Rb, Zn, and, toa lesser extent, K, Mn, Pb and Sn are higher and Ca and Srare lower in mineralized than in nonmineralized stocks. Themean contents of Pb and Zn in feldspar from the granite showa clear distinction in Zn content between barren andmineralized intrusions (Fig. 15.4). The mineralized andbarren intrusions, however, are not strictly comparable, sincethe mineralized intrusions are more highly fractionated thanthe barren intrusions; only the Weardale granite (identified onFig. 15.4) is petrochemically similar to the mineralizedgranite - but has an even lower Zn content than other barrengranite. The study was based on relatively few samplRs (atotal of 55 samples from the fi ve mineralized granitc bodies

2 5 10 20 >20ppm

Figure 15.2. Frequency distribution of W in mineralizedand barren granitic intrusions, Northern Canadian Cordillera(data from Garrett, 1971).

*>..uc~ 40ITCD.....

Tauson and Kozlov, 1973). The form of frequencydistributions for W has been shown to di ffer bet.ween barrenand mineralized acidic plutons in the Yukon and NorthwestTerritories in Canada (Garrett, 1971). The distribution of Win barren plutons (Fig. 15.2) is only slightly skewed, whereasthe distribution in mineralized plut.ons has a more markedposi tive skew; this is interpreted as reflecting the occurrenceof a tungsten minerali7inf) event due to magmatic processessuperimposed upon the primary tungsten distribution. Plutonsthat are host. to porphyry-type disseminated tungstenmineralization, and therefore presumably were closely alliedto primary maf)mat.ic processes, are more clearlydistinguishable from barren plutons than those associatedwith more local skarn-type mineralization. Garrettconcluded that. an "average" size pluton could begeochemieally characterized by collecting duplicate samplesfrom 15 samples si tes in an unzoned pluton.

Attempts have also been made to use variations inelement content of a particular mineral phase in a rock todiscriminate between barren and productive plutons. Most ofthe work has been done on biotite (although trace elementdistribution in feldspar and magnetite have also beeninvestigated). The rationale behind the investigations is thefact that the structure of biotite can readily accommodateminor elements in different states within the lattice and thatbiotite in productive stocks might contain characteristicminor element contents as a result of hydrothermal alterationassociated with the mineralization. Some early work on thedistribution of Cu and Zn in biotite from acidic intrusions inthe Basin and Range Province of the U.S.A. (an area of majorbase and precious metal production) by Parry and Nackowski(1963) showed an apparent relation between the contents ofCu and Zn in biotite and the presence of base metalmineralization within a particular stock. The high coppervalues in biotite from the Basi n and Range intrusions wereascribed to the deposi tion of sulphides in biotite byhydrothermal solutions. Subsequent investigations on biotitefrom the Basin and Range Province, however, have indicatedthat there is no clear relation between minor elementcontents of biotite in intrusions hosting porphyry coppcrdeposits and unmineralized stocks (Jacobs and Parry, 1976);this is attributed to the existence of three genetic types ofbiotite (magmatic, replacement, and hydrothermal), thecomposition of which reflects the physical and chemicalconditions of crystallization and the postmafJmat.ic hist.ory ofa stock. Systematic variations in biotite composition of thethree types of biotite may be characterist.ic of individualporphyry copper deposits and thus moy serve t.o distinguishintrusions genetically related to porphyry copper depositsfrom barren intrusions.

The halogen distributions in biotite within individualplutons of the Basin and f{anye area are distinctive, but thereis no systematic distinction between barren and mineralizedintrusions (Parry, 1972; Parry and Jacobs, 1975). The meanCI content of biotite from intrusions in western NorthAmerica and the l:aribhean was shown to be similar in barrenand mineralized plutons, ond there is a greater range ofvalues in the former than in the latter (Fig. 15.3) (Kesleret aI., 1975a). Similarly neither water-soluble nor total CIand F cont.ent in whole rock is consistently different. in barrenand mineralized intrusions (Kesler et aI., 1973, 1975b). Thelack of a clear relation between halogen content andmineralizat.ion is attributed to the halogen content of thebiot.ites also being related to magmatic processes.

Some differences have been noted in major and traceelement contents in feldspar, biotite, and muscovite fromfive granite bodies associated with tin, copper, zinc, and leadmineralization in Devon and Cornwall (southwest England)compared to nonmineralized granite elsewhere in England andScotland (Bradshaw, 1967). An average of eleven samples per

344 G.J.5. Govett and I. Nichol

and a totnl of 94 from the eight barren granite bodies);although there are clear differences in the mean elementcontents, there is also considerable overlap between the twogroups of granite in terms of element content of individualsamples. The differences in mean contents of the ore metals5r1, Pb and Zn betwe~n mineralized and unmineralized stocksare attributed to effects of mineralization whereas variationsin the mean contents of K, I<b, 51' and Ca are related tovarying degrees of fr8ctionation of the granite.

A rather different approach, based on determining themetal content of sulphid(~ minerals, has been adopt~d todiscriminate significantly mineraliLed from unmineralizedultramafic bodies of the Canadian Shield (Cameron et at.,1971). A total of 1079 samples from 61 locations wereinvestigated and each location classi fied according toimportance of mineralization: an ORE group contains oregrade deposits with more than 5000 tons nickel-copper, aMINORL group contains deposits of ore grade with less than5000 tons nickel-copper, and a RARREN 'troup contains onlyminor amounts of sulphide. The samples collected variedaccording to restraints on sampling irnposed by sampleavailabili ty blJt attention was focused on obtaining a suite ofsamples as representative as possible of the individual areas.A sulphide-selective digestion was used to determinc the Cu,Ni, and Co contents present in sulphide minerals, based on theassumption that for an ultramafic magma to give rise tosignificant corper-nickel mineralization, the m<lgma must beenriched in sulphur to such an extent that the solubilityproducts of the metal sulphides will be exceeded, causingmetal sulphides to separate. The content of sulphur or metalsheld as sulphides should be indicative of this condition 'mdhence of possible lTlineralization. The frequency distributionsof sulphide-held Cu, Co, and Ni, <lnd of total sulphur, are all

pDsilively skewed; the results are summarized in Tabl~ 15.2.Ore-bearing 10cClIities (ORE group) are enriched in suJphideheld Ni and Cu (and, to a lesser extent, Co) and total 5relative to mineralized localities (MINORE group) and tobarren locnlities (BARREN group). The MINORE andBARRfN localities are only distinguishable from one anotheron the basis of higher sulphide-held Ni content of th~

MINORE group.

The importance of recognizing the mineralogical site oftrace elements has been demonstrated by variations in themineralogical form of uranium in granite bodies in the U.K.(Simpson et at., J977); delrtyed neutron analysis and Lexanplastic fission track studies were carried out to determine thecontent and form of the uranium. The geometric mean of Ucontent of the Caledonian granite is 3.9 ppm U (118 samples,range 0.5-17.2 ppm), compared to Hercynian granite ofsouthwest England that has a geometric mean of 10.8 ppm U(66 samples, range 3.2-35.5 ppm). The difference betweenthe two granite suites is significarlt at the 99 per centconfidence level (bas~d on the two-tailed KolmogorovSmirnov test). In the Caledonian granite, U occurs inpostmagmatic secondary minerClls, such as hematite andchlorite, in alteration zones associated with major faults. Insouthwest England, uranium occurs as mainly primaryaccessory minerals such as zircon, rtpatite, and sphene invein-type deposits closely associated with the granite. It wasconcluded thClt the LJ vein-type mineralization of theHercynian granite is co magmatic with the uranium-enrichedgrani te and is not attributable to later kaolinization orweathering. Analyses of granite can serve to identifyprim<lry uranium provinces with associated vein-typemineralization. In the CClledonides, where uranium isprincipally associated with later faults and molasse faciessediments, granite sampling can help to identi fy theenrichment processes and thus aid in the definition ofeconomic targets.30

10

little or no mineralizationn-53

300FELDSPAR o

20ppm Zn

o

o

Key:

o Mineralized• Mineralized averageo Barren• Barren average

o

10

o

/Weardale

oO+--------r-------r----

100

200

EQ.Q.

0.40.3

Cu Porphyryn-51

CU,Pb,Zn,Ag,Au depositsn-63

0.2

CI%

0.1o

10

30

~>- 10ucCIl:;)

tr 50CIl..u.

30

Figure 15.3. Frequency distribution of Cl in biotite frommineralized and barren granitic intrusions. Basin and RangeStructural Province. U.S.A. (redrawn from Parry and Jacobs.1975).

Figure 15.4. Mean contents of Pb and Zn in biotite {rommineralized granitic intrusions of southwest England andbarren granitic intrusions from northern England and Scotland(data from Bradshaw. 1967).

Lithogeochemistry 345

Table 15.2

Geometric mean content of sulphur and sulphide-held Cu, Ni, and Co in ore-be<Jringmineralized, rmu barren ultramafic intrusions in the Canadian Shield

(from Cameron et al., ] 971)

proximity to mineralizationwithin the altr-ration zone andmay be useful in exploration forblind deposi ts.

There have been a variety of investiqations of geochemical patterns associated with specific mirleral deposits within,or closely associated with, intrusi ve rocb. Porphyry-type oredeposits constitute an important ore type where in theexploration stage it is frequently of significance todistinguish arCilS of mineralization within intrusions (Lowelland Guilbert, 1970). An evaluation of the nature ofgeochemical dispersion associated with porphyry copperdeposits as a whole is difficult due to variations in samplingprocedures and in the elements that have been analyzed byvarious worker,;. In view of the importance of porphyrycopper deposits as sources of Cu and Mo, it is somewhatsurprising that relatively few lithogeochemical studies havebeen reported. Available data for the distribution of majorand trace elements in rocks around copper porphyry depositsare given in Table 15.3. Perhaps the most comprehensiveinvestigations are those carried out on the I iighland Valleydeposits of British Columbia (Brabec and White, 1971; Oladeand Fletcher, 1975, 1976a, 1976b; and Olade, 1977). A totalof 1860 samples were used in the study, comprisirlg 1800 fromthe mineralized areilS and 60 from background areas of theGuichon Creek. Surface outcrop samples consisted of 4 kgcomposites of severed fist-sized chip samples Clnel drill coresamples comprising composites of 5 cm lengths of drill coretaken over 3 m. Analysis was carried out for 3D clements orforms of elements (e.g., total and water extractable C\). Inthe Highland Valley, geochemical dispersion in proximity toporphyry copper deposits is related to primary lithology,hydrothermal alteration, and mineralization (Fig. 15.S) Oladeand Fletcher (1976a). Ancmalous halos of Cu, S, cHId Bvariously extend beyond the alteration zones and thusconstitute larger explor<Jtion targets than the visiblealteration halo (Table 15.4), (Fig. 15.5). Rb, Sr, and Badistributions are closely rclated to alteration and thus provideuseful guides of porphyry Leming. Element ratios were foundto be more consistent indicators of mineralization than sin<Jleelement patterns, which show erratic trends due tumineralogical variations.

In a preliminary investig<Jtion of geochemical dispersiona,;sociated with the Kalamazoo deposit (southwestern UnitedStates) whole rock samples of core and cuttings from 3 mintervals, spaced 15 m apart from two drill holes wereanalyzed for 60 elements (Chaffee, 1976). Trace elementdistributions, for the most part, are not related to lithologyand are thus more significant in terms of mineral explorationthan major element distributions. Increases in Cu, Co, B, S,or Se and decreases in Mn, Tl, Rb, and Zn occur with

0.03J

0.036

0.05(;

S, %

D.] (;657.4

25.2

31.3

37.9

Co, ppmThe Copper Canyon

porphyry copper deposi t inNevada occurs in fractured andaltered sedimentary rocksassociated with <J <]ranodioriteintrusion (Theodore and Nash,1973). One kilogr<Jfll compositerock chip samples from areas of10 m 2 were collected over a16 km 2 zone and <In<Jlyzed for20 elements. Highest copperconcentrations occur over thenonproductive qrrmodioriterather than associated with theorebodies in the adjacentmetasediments. The copperdistribution within the sedimenthas also been affected by

premineralization structures 'md supergene enrichment. Theabundance of metal dispersed through rocks at CopperCanyon rather than occurrirlrJ as fracture fillings of coatingsis a more specific indicator 01 ore deposits. The distributionof high salinity fluid inclusions outlines the orebodies in themetasedimentary rocks more precisely than the pyrite halo.The distribution of the high salinity fluids is considered toreflect the circulation limits of the fluids. These data drawattention to the limited effectiveness of exploration basedsolely on a consideration of ge()chemical data and the needfor the additional consideration of the petrography of thefluids and alteration.

In the Copper Mountain area of British Columbia whereporphyry-type mineralization in volcanic rocks surrounds theintrusi ve, patterns exist on various scales (Gunton and Nichol,1975). An orientation survey was c<Jrried out to establish thenature of geochemical zoning associated with the depositsand identify appropriate geochemical exploration parameters.On the basis of these results 5 Ib composi te chip sampleswere collected as outcrop permitted from 500 foot grid cellsand the s0mples were analyzed for 18 elements or forms ofelements. High contents of Na, K, P, Rb, Sr, total Cu andsulphide-held copper are characteristic of the volcanicssurroundirHJ the stock relative to background areas thusprovidin<J a broader exploration tarrJet. than the discretedeposits. In significantly mineralized meas surrounding thestock area the host rocks are characterized by 8-fold higherCu and 1.5-fold higher Sr contents over three-fold moreextensive areas than the economic minerali/iltion. Thevolcanic host rocks associated with the individual Ingerbelleand Copper Mountain deposits are characterized by high sodaand potash contents reflecting di fferences in the nature ofmetasomatic alteration associated with the two deposits. Onthe basis of these results it was concluded that inreconnaissance level tOxploration samples should be taken at adensi ty of ten samples per square mile and in detailed levelexploration at 80 samples per square mile with the samplesbeing analyzed for P, Sr, Na, K, and Cu. During samplingattention should be given to recording dominant Cllterationtypes, degree of fracturing, and SUlphide content.

In a study intu the nature of primary dispersionassociated with an early Precambrian porphyry-typemolybdenum-copper mineralization in northwestern Ontario,fifty 500 g to 1 kg composite samples of rock chips werecollected 5-15 m apart from the 5.7 km 2 area of the porphyritic granodiorite-quartz monzonite pluton (Wolfe, 1974). Thesamples were analyzed for Cu, Zn, Pb, Mn, and Mo. Stronglycontrasting Cu and Mo dispersion within the stock partly

715

560

354

469

Ni, ppm

67.8

6.8

6.9

15.2

Cu, ppm

372

91

616

1079

Number ofSamples

5

40

61

16

Number ofbodies

or.zE groups:(deposi tswith >5000tons Ni-Cu)

MINURE group:(deposi tswith <,-,DODtons Ni-Cu)

BARRlN group:

All depnsi ts:

Local and Mine Scale


Table 15.3

Halos of selected elements around some copper porphyry and porphyry-type deposi ts

Deposit Na K Ca Mg Fe FeS S B Cu CuS In Mn Mo Hg Au Ag Rb Sr

Bethlehem JA(1,2,3,4) - + - - - + + + + + - - + + + -

Valley Copper(1,2,3,4) - + - - - + + - + + m - + 0 + -

Lorne)2) na na na na na na na + + na m 0 + na na -

Highmount(2) na na na na na na na + + na + - + na na -

Ingerbelle(s) +. + 0 0 0 0 na na + + na 0 na na na na + +

Granisle(6) na na na na na na + na + na m na + + + 0 0 0

Bell Copper na na na na na na + na + na m na + + + 0 0 0

Setting Net Lake(7) na na na na na na na na + na 0 0 + nd na nd na na

Kalamazoo(6) - +? 0 0 + na + +7 + na m7 - + 0 + + - 0( )

Copper Canyon 9 na na na na na na na 0 + na m 0 + + + + na +

Bio Blancoe0) na 0 0 na na na na na na na na na na na na na + -

El Teniente(10) na 0 0 na na na na na na na na na na na na na + 0

na =not analyzed; nd =not detected; + =positive halo; - =negative halo; m =positive halo at margin of deposit. Interpretedin part by the writers from the following sources: Olade and Fletcher, 1975 1, 1976a 2, 1976b 3; Olade,19774; Gunton andNichol, 1975 s ; Jambor, 19746; Wolfe, 1974 7; Chaffee, 1976 6

; Theodore and Nash, 19739; Oyarzun, 1975 10 .

coincides with the known mineralization and also gives someindication of the presence of hitherto unknownmineralization.

Mineralized andesite associated with granodioritedacite complex of the Rio Blanco (Chile) deposit have a3-fold Rb enrichment and an 8-fold Sr depletion relative tofresh andesite (Oyarzun, 1975); at EI Teniente a 2-fold to4-fold enhancement of Rb is characteristic of the alterationzone relative to unaltered andesite (Dyarzun, 1971). Nosignificant variation in Sr content was noted, probably due tothe abundant anhydrite present in the potassic zone. At theEl Abra porphyry deposi t in Chile, no significant anomalousrubidium dispersion is associated with mineralization probablydue to lack of chemical contrast between intrusive andintruded rocks (Page and Conn, 1973). At Chuquicamatahighly anomalous rubidium contents have been reported(Oyarzun, 1975).

Vein deposits associated with intrusive rocks representthe smallest target for exploration; geochemical halos arealso correspondingly small, although narrow wall-rock halosare commonly well defined. Au and Ag anomalies occur inwhole rock and ferromagnesian minerals in the Marysville(Montana, U.S.A.) quartz diorite-granodiorite up to 300 mfrom gold and silver vein deposits within and adjacent to thestock (Mantei and Brownlow, 1967; Mantei et al., 1970). Theanomalous contents of Au and Ag in the mineralized areas ofthe stock were considered to have resulted from thepenetration of the wall rock by mineralizing solutions. Thepresence of Au and Ag and lack of base metal anomalies ispossibly related to a higher mobility of Au and Ag or highercontrast in content of Au and Ag of the mineralizing fluidsand the adjacent rock than in the case of base metals.

An even more extensi ve anomalous halo of Mn in biotitein the granodiorite-quartz monzonite Philipsburg Batholith(Montana, U.S.A.) has been described by Mohsen andBrownlow (1971); the Mn content of biotite increases two-foldover 1500 m towards the margins of the intrusion wheremanganese deposi ts occur in adjacent sedimentary rocks.

The general feature of geochemical halos around veintype deposits is illustrated in Figure 15.6 based on the

Table 15.4

Comparison of anomaly extent and contrast for selectedelements (from Olade and Fletcher, 1976a)

Bethlehem ValleyJA Copper Lornex Highmount---

Element E 1 C 2 E C E C E C

Cu *** 5 *** 5 *** 5 *** 5

S *** 5 *** 5 nd nd nd nd

Mo * 2-5 * 2-5 * 5 * 5

B * 1-2 0 *** 2-5 *** 2-5

Hg ** 5 0 nd nd nd nd

Rb ** 1-2 ** 1-2 nd nd nd nd

Sr3 ** 1-2 ** 1-2 ** 1-2 ** 1-2

nd: Not determined

1Extent: 0: No anomaly*: Anomaly confined to ore zone

**: Anomaly within alteration envelope***: Anomaly beyond alteration envelope

2Contrast: Approximate anomaly to background ratio

3Negative anomaly

collection and analyses of chip samples, which shows thedistribution of Cu, Pb, In, Ag, and Mn in sandstone adjacentto lead-zinc-silver and copper lode deposits in the Park CityDistrict of Utah (Bailey and McCormick, 1974). The decaypattern in metal values in wall rock away from the veins isapproximately logarithmic - a common pattern that has ledmost observers to regard this type of dispersion as diffusioncontrolled migration of mineralizing fluids from the vein.The anomalous halo has a restricted extent; anomalous Cuand Mn contents are detectable for about 10 m, andanomalous Pb, In, and Ag contents are detectable for about20 m from the vein.

Lithogeochem istry 347

Regional Scale

Attention has been focused by Coope (1977)on the composition of exhalative horizons as aregional indicator of mineralization. Exhalativehorizons (i.e., essentially chemicrtl precipitates)commonly occur over a wide area at the samebroad stratigraphic horizon as volcanic-sedimentary sulphide deposi ts. The metalliferoussediments bein'l deposited in the Red Sea representa sedimentary hydrothermal deposit in the processof formation and show anomalous contents of Hg,Cu, Zn, and Mn in the exhalite sediments extendingup to 10 km from the Atlantis 1I Deer with 6-foldcontrasts between mean metal contents in samplesadjacent to the deep relative to mean b3ckgroundcontents in samples remote from the deep.Anomalous contents of Cu, Zn, and Mn also extend9 km from the Nereus Deep (Bignell et al., 1976).In both cases the anomalous elements show aqradient of increasinfJ concentration towards theexhalative centres lor8ted in the Deeps.

A number of anomalous dispersion patterns insedimentary rocks around massi ve sulphides inpredominantly sedimentary environments have alsobeen reported. In the Ii mestone overlying theTynafJh base metal deposit in Ireland there is a Mnanomaly in limestone, identifiable in surface andcore samples, extending at least 7 km, with up to a20-folrl contrast (Fig. 15.7) (Russell, 1974). Up to a20-fold range in Mn content exists in samples nearmineralization requiring that an adequate numberof samrles are taken for representativeinformation to be obtained. It is considered thatthe anomalous concentrations result fromdeposition of metal from "spent" mineralizingsolutions following their eSC8pe into the seasubsequent to the major phase of deposition of oremetals in the underlying rocks. High Ba, Pb, Zncontents also occur in the iron formationassociated with the mineralization (Russell, 1975).At the Meggen lead-zinc-barium deposi t inGermany there is a Mn anomaly in limestone withup to 7-fold contents extending at least 5 km fromthe deposi t (Gwosdz and Krebs, ] 977) (Fig. 15.8).The observations based on the analyses of342 samples again display a 3 to 4-fold range inmanganese concentration (particularly near themineralization). Peak Mn contents lie lateral to,and not immerliately over the deposit which isconsistent with the model for Mn dispersion in anaqueous environment around an active brine source2S predicted by Whi tehead (1973).

EAST

J 19ppm

At the McArthur River (Australia) Zn, Pb, Ag deposit,based on the analyses of samples from six drill holes,anomalous contents of Zn, Pb, As, and Hg extend at least20 km from the deposit in shale which is the lateralequivalent of the unit hosting the deposit (Lambert and Scott,1973). Anomalous zinc contents in metasediments in theregion of the Kidd Creek deposit have also been recorded(Cameron, 1975). These types of extensive anomalies(Tynagh, Meggen and McArthur River) are interpreterl 8S

syngenetic with the deposits and 10 have resulted fromdeposition from solutions in euxinic basins in nearshoreregions of shallow seas.

The cupriferous pyrite massive sulphides of Cyprusoccur in a sequence dominated by theoleiitic basaltic pillowlavas with only minor local sedimentary rocks associated withthe deposits. From 20 traverses across the strike of theTroodos volcanic belt 2000 samples were collected and were

8990 ppm

OREZONE

K::::~ 1 22 ppm

I 7 ppm II II II II I

I12 500 ppm I

Phase

Arg ill i c: '------'- ----'--p.:..:h-Ly--C1.:..:1i--=c__---J'--~I.:-A:..:.r...911TIL:..:i:.:..:i c

RegionalBackground

Limit of samPlin g",-

Zn~1 ~~I

of sampling

Sr

S

Cu

o 200m...............Scale

I L"I.¥""" Im,t

III

f------------l-------------l---~1322IIII

~~ : ~:588Ppm

I 198 ppm I II I II I II I I

'b~"".I I I

M~~Oppm ~ Ir----------~L.--_!_--=:J>~------~_=_____t___"'.c.:.-+~~----l12ppm

I II

Bethsaida

WEST

Figure 15.5. Distribution of Cu, S, Mo, Rb, Sr, and Zn in granitic rocksalong a traverse over the Valley Copper porphyry copper deposit. BritishColumbia, Canada (redrawn from Glade and Fletcher, 1976a),

GEOCHEMICAL RESPONSE TO MINERAL DEPOSITS INEXTRUSIVE AND SEDIMENTARY ROCKS

In this section, the lithogeochernical characteristics ofthe economically important stratabound sulphide occurrences- variously referred to as "massive", "volcanogenic", or"volcanic-sedimentary" - are considered. Included in thiscategory are deposits that appear to have formed in asubaqueous environment, probably from ascending hot,metalliferous brines, and that lie conformably within maficvolcanic sequences (e.g., Cyprus), felsic volcanic sequences(e.g., Bathurst, New Brunswick, Canada), or in predominantlysedimentary sequences (e.g. McArthur River, Australia).Stratabound deposi ts of an apparently sedimentary origin(e.g., the Zambian Copperbelt, Kupferschiefer) and vein-typedeposits of volcanic association (e.g. Kirkland Lake,Kalgoorlie, Australi8) are specifically excluded.


Distance from vein in metres

Figure 15.6. Smoothed average distribution of Ag, Cu, Pb,An, and Mn in sandstone adjacent to mercury-copper-leadzinc vein mineralization, Park City District, Utah, U.S.A.(redrawn from Bailey and McCormick, 1974).

analyzed for up to eleven elements (Govett and Pantazis,1971). As virtually no fresh surface rocks occur in Cyprus, allthe samples have undergone weathering to various degrees,although attention was given to obtaining samples asunweathered as possible. The data indicated that thecontents of Cu, Zn, Ni, and Co vary as a function ofproximity to mineralization, stratigraphic position, petrology,rock type, and secondary processes. Mineralization occurscharacteristically in areas of low Cu and high Zn and Co. Nosulphide deposit occurs in a region with a Cu/Zn ratio inbasalt which exceeds 1.2; all but two of the deposits occur inzones where the Cu/Zn ratio is less than 1.0, as shown inFigure 15.9 (Govett, 1976).

The salient features of Precambrian volcanogenicmassi ve sulphide deposi ts in Canada have been described bySangster (1972). Regional-scale lithogeochemical studies ofmassive sulphide deposits in the Canadian Shield indicate thatthey occur in calc-alkaline rather than theoleiitic mafic tofelsic volcanic sequences, and that they generally occur infelsic rather than mafic differentiation sequences(Descarreaux, 1973; Cameron, 1975; Wolfe, 1975). Themassi ve sulphide deposi ts in the Bathurst district of NewBrunswick (Canada) similarly occur in calc-alkaline felsicvolcanic rocks (Pwa, 1977). Distinctive geochemical featuresof productive and nonproductive cycles of volcanism havebeen noted in various areas of the Canadian Shield (Davenportand Nichol, 1973; Cameron, 1975; Nichol et al., 1975; Wolfe,1975). Productive cycles of volcanism are generallycharacterized by slightly higher contents of Fe, Mg, and Znand by lower Na20 and CaD, contents as illustrated inhistograms of Fe203 and Zn contents from the productiveand nonproductive cycles at Uchi Lake (Fig. 15.10)(Davenport and Nichol, 1973). These data are based on theanalyses of 1 kg composite chip samples collected at afrequency of 4 samples per square mile over an area of72 square miles. There is an almost complete overlap in thetwo populations for Fe203 and Zn, but the form of thefrequency distributions is quite different and the meancontents for Fe203 and Zn are higher in the productivecycle. The differences between the two cycles, are moreclearly seen if the degree of fractionation, as represented bythe Si02 content is taken into account (Fig. 15.11, 15.12)(Nichol, 1975). The Fe203 and Zn contents of rocks from theproductive cycle are clearly higher than those from thenonproductive cycle for equivalent Si02 contents. In thelatter case these data relate to 1 kg samples of individual

Pb

--Zn

-

\"~ Mn

"........,...

\ Cu" ,

I \

':1 \.,.,.,.,.>._._._,_'_'~'-'-'-'-'---~J!a 20 40 60

1000

-cCIlECIl

w

E~ 100

.5co-c...-cCIluCou

10000 8o

o

......

o Drill core samples... Surface samples

1000

Mnppm

100 .........

...... ...

a15 10 5 a 5 10

WNW MINEFigure 15.7. Distribution of 0.2 m acetic acid-soluble Mn in limestone that is host to theTynagh deposit (Pb-Zn-Cu-Ag-BaS04) deposit, Ireland (redrawn from Russell, 1974).

15km

ESE

hemistryLithogeoc 349

Mnppm

2000 .

1000

'i~' .'. .. ".

,: /;:.. ,·/<:..i:

/

.:J .....::.. . . .': . :: ":.:

i-:--j,.: .

o I!l~-----.l.....-

3km,

KYRENIA

-Kambia

\A

-Nicosia

rrI=I=JC~;;=O~~~~·:B:a:50:~:-'-- e an lead-zinc-Ba 504h t is host to the M gg

. total Mn in lim~s~~~eir~bs, 1977).DistributIOn of from GwosdFigure 15.8.. " Germany (redrawn. deposlt,banum

Sulphide mines

Geochemical traverses

Cities and towns

L1MASSOl

tory - Troodos~ Contact.sedimen,~ .. _0<_". .Cu I

Z

1

0

12 oct Troodos vo.lcolllcI; 0 0,,~ 118. eon, . Troodos IgtIl!llUS-" . 10 L~-,~~ "ho roLo < . ~p

~ ,~p"" Gowtt. 1976)..., du"d f""m• • (yp,",r"p"

~

..::i°llllll~!~Mil 'Uow lava,•• SCALE IN MILES t' in basaltic Pl

. of Cu/Zn ra 10DistributIOnFigure 15.9.


Zn

Zn

The elements that display anomalous dispersion patternsin the wall rock associated with massive sulphide depositsshow some variation amongst deposits and the composition ofthe wall rock also reflects the effects of fractionation. Inorder to identify a more generally diagnostic indicator ofmineralization, a function based on the Fe, Mg, Ca, and Nacontents related to mineralization was determined for wallrock associated with a number of massive sulphide deposits inthe Canadian Shield (McConnell, 1976). The component ofthe Fe, Mg, Ca, and Na content due to fractionation wasestimated on the basis of the SiOz content and subtractedfrom the observed concentrations, the difference or residualcontent then possibly being related to mineralization. Bycombining the residuals of Mg, Fe, Ca, and Na in the form ofa standardized net residual a more consistent reflection ofproximity of mineralization was obtained. In general thestrength of this factor increases towards mineralization andtowards the presumable conduit of mineralizing solutions.The distribution of the "standardized net residual" at theSouth Bay Deposit extends beyond the mapped alteration zone(Fig. 15.18) but at the East Waite deposit it is only asextensive as the alteration zone. The variation in Mg and Cain footwall felsic volcanic rocks with proximity to theBrunswick No. 12 deposit is shown in Figure 15.19; the Mg/Caratio is still three times greater than mean background 175 mfrom the ore zone (Govett and Goodfellow, 1975).

Concn.109 ppm

Productive

1.05

Non-productive

.45 .60 .75 .90

Concn. 109 0/0

i = 3.7%

.3

0+------,--.--+----+-...,--1--,

30

30

>uc:G>

~ 10~u.

>uc:G>:lIT 10~

u.

textural types rather than composite chip samples. Largelysimilar differences have been shown to exist betweenproductive and nonproductive cycles of volcanism in a numberof mining areas within the Superior Province of the CanadianShield (Lavin, 1976; Nichol et al., 1975, 1977; and Sopuck,1977). However, the geochemical distinction of productivecycles is not apparent in all textural types. For example, atNoranda a relatively high Zn content is apparent only inspheroidal flows of the productive cycle and does not appearto exist in the presumably less permeable massive flows(Fig. 15.13) (Sopuck, 1977). Similarly variations in chemicalcomposition with textural type give rise to significantvariations in composition. In pyroclastic units of thenonproducti ve Kakagi Lake area the fragments are moresiliceous and soda-rich than the matrix whereas the matrix isenriched in MgO, Fez03 (3-fold), and Zn (5-fold) (Fig. 15.14)(Sopuck, 1977). Since these are some of the most usefulelements indicative of mineralization the nature of thesample needs to be considered during interpretation.

"" 20

Local and Mine Scale

"" 20

distribution, although the mean value and alsothe modal value of the anomalous group ishigher, as seen in Figure 15.15 (Govett andPantazis, 1971; Govett, 1972). This type of

i= 58 ppm overlap in element content between back-ground and anomalous populations of rocksamples has led to the use of computer-basedstatistical techniques of data interpretationin an attempt to clarify any differencesbetween populations. The application ofdiscriminant analysis to this type of problemis illustrated with reference to data from atraverse across the Mathiati Mine in Cyprus(Fig. 15.16) (Govett, 1972; Pantazis andGovett, 1973; Govett and Goodfellow, 1975).The difference between a background and ananomalous population has been maximized bycalculating the discriminant score thatexpresses the combined effect of increasingZn and Co content and decreasing Cu content

i-126ppm in pillow lava with proximity to mineraliza-tion. Anomalous discriminant scores extendabout 2 km from the deposit.

Major element patterns in volcanicrocks around massive sulphides appear tohave some common characteristics, regardless of age, economic metals, or thegeological environment of the deposits(Table 15.5). Proximity to mineralization isgenerally indicated by an enrichment in Feand Mg and a depletion in Na and Ca. The

1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4 2.6 behaviour of K is variable; it is generallyenriched relative to Na close to sulphidemineralization but in some cases it is

Figure 15.10. Frequency distributions of Zn and Fez 03 contents in productive depleted; there are also some cases where Naand nonproductive cycles of volcanism. Uchi Lake (Davenport and Nichol, 1973). shows an absolute enrichment. These

patterns are best developed in footwall rocks,but, in some situations, they also occur in thehanging wall. Mg enrichment and Na deple-tion occur adjacent to the Jay Copper Zone

(Abitibi, Canada); the very distinct Na anomaly extendsup to 500 m from the mineralized zone, but the Nacontent beyond these limits is still anomalously low(Fig. 17) (Descarreaux, 1973).

Investigations of geochemical variations in rocksassociated with individual massive sulphide deposits are farmore numerous than regional studies (Tables 15.5 and 15.6).Again trace element anomalies are typically quite subtle andare generally most readily detected as population differences.Comparison of the distribution of Zn and Co in backgroundbasaltic pillow lavas with that in similar rocks within about1 km of the Skouriotissa (Cyprus) cupriferous pyrite depositshows an almost complete overlap in the range of


60 90

•

•

oo

•

•

80

•••·••.. .••

•

•••..••••••

••••

•

• 0o

70

•

•

••••.: ..... -.•• •••~. -=u·:o ·i.::~ .:-II... .. ,.,....

• I •• •.. ( ..o

o ,.

o o.• 0 cr.,• 0 Co

•

••• ·0

••

o

•••

o

o

o 0

=-" 00 0oo

••

50

9

3

6

o40

12

15

Confederation Lake • Woman Lake o

Figure 15.11. Relation between Fez 03 and SiOz contents of productiveand nonproductive cycles. Uchi Lake (Nichol, 1975).

300 • • • • •• • •• • ••

• •• • •

240 ••• • •• • •

0 .. •0 • • • •• • • •• • • • •• •180 0 • ..... . • • •••• •• • • •Zn • • • • •0 0 ... .. • •• •..". . •ppm 0 ".. • ~ ••• • • . ,::a • •• •• 00 • .... I •120 •• • • 0 •• • • • •00 0 • .,.

0 00 . ... : • •• •0. Co 0 • c' • ..

0 •• •• • • • •00 00;. , •• 0 ••• 0 o.

~~ .0 0 • eo• 0 • •

oOo.~oo~ 0 • • • 0~ •• t o 8. • • •

60 \;8 o~ d6 ••0 •~.~ 0 i· 0

• 0 0 -'?• o~o 0 100 •:~ 00 eo.

o 0 0 0 0cfJ no 0 0000

040 50 60 70 80 90

Si02 o~0

Confederation Lake • Woman Lake 0

Figure 15.12. Relation between Zn and SiOz contents of productive andnonproductive cycles, Uchi Lake (Nichol, 1975).


MASSIVE FLOWS SPHEROIDAL FLOWS

mineralized •unmineralized 0

Znppm

200

100

200

Znppm

•100

•

•• • •

,• ••

0

0

Si02 %

Figure 15.13. Relation between the composition and texture ofproductive and nonproductive cycles, Noranda (Nichol, 1975).

75 85

706050O-!--,----------,-----.,.--

706050O-l--,-------r----,--

6 120 o Matrix• Fragment

4 80MgO Zn

% ppm

2 40

7050O-!---r-------.,.-----..,---

8

4

12

7050

2

6

4

0-1--,----------,-------.---

Figure 15.14. Relation in composition of fragment and matrix of pyroclastics at Kakagi Lake (Sopuck, 1977).


Table 15.5

Summary of major element dispersiuns in relation tovolcanogerlic massive sulphide deposits

Alteration Elements Elements Elements AgeDepusit mineralogy enriched depleted LHlchanged

Kuroko, Japan Mon, Ser, Chi, K, Mg, Fe, Si Ca Ceno7lJicLambert & Sato Kaol(1974)

Kuroko, Japan Ser, Qtz, Mg,K Na,Ca, Fe Al CenuzoicTatsumi & Clark Cal(J 972)

Hitachi, Japan Cord, Anthuph. Mg, Fe, Sa Na, Ca, Sr CenozoicKuroda (1961)

Buchans, Canada Chi, Ser, Mg, Fe, Si Na, Ca, K P;clleozoicIhurlow et al. Qtz

(J 975)

Heath Steele, Chi, Ser Mg Na, Ca PaleozoicCanadaWahl et al.(1975)

Brunswick No. 12 Chi, Ser, Mg, Fe, (Mn), Na, Ca Al PaleozoicCanada Qtz (K) (Mn), (K)Goodfellow(1975)

Killingdal, Chi, Bin, Mg, K, Mn Na, Ca, Si AI, Ti, PaleozoicNorway Qtz Fe (total)Rui (1973)

Skorovass, Norway Chl,S"r Mg Na,Ca PaleozoicGjelsvik (1968)

Boliden, Sweden Chi, Ser, Mg, K, Al Na, Ca ProterozoicNilsson (1968) Qtz, Andal

Mattabi, Canada Qtz, Carb, Fe, Mg Na,Ca ArcheanFranklin et al. Ser, Chid,(1975) Chi, Andal,

Gar, Kyan,Bio

Millenbach, Canada Chi, Ser, Mg, Fe Na, Ca, Si ArcheanSimmons et Lll. Anthoph,(1973) Cord.

Mines de Poirier Chi, Ser, Mg, K Nel, Ca Si ArcheanCa,nadaDescarreallX(1973)

Lac Duf8ult, Canada Chi, Ser Mg, Fe, Mn Na, Ca AI, Ti ArcheanSakrison (1966) K, Si

East Waite, Mobrun, Qtz, Ser, Chi, Mg, Fe Na, Ca Si, Al ArcheanJoute!, Poirier, Carb, Sauss,Agnico-Eagle, LpidoteMattabi,Sturgeon Lake,South [J;JyMcConrmll (1976)

601 Co - backgroundIt

I

I40i Jl

n-85 Differences from the averagex- 30.7 curves of Abitibi for a given

I Si02 value

II I + +~ '" 0 '" ~*

201~ ~ I

z l ~

IQ \ co

t-) ~ °0 \

20

+00Nl\ I

o+----L----.--'---.~ \

r0 50 100 .'601 II>•• C, • '"Co - anomalous I ..,

/ »nI

/ m/

40i 12+00N

1

/ II>

/ » C"Jn.59 • ~ LI

x-43.8 / .".... !n• zI

{. Gl C"JI • 0

<201 / p C'l.....

I P .....

I 1;::t ~

·----e-_ :::l

In Co

4+00N c: ii -, :-'. ~ , z0 I z I (.___e ri'--. m '7

0 50 100 V') \I\\\ ••••••••••::::;:::~\': e.ppm \~I\'\\~ _:.~t...·i.,·i,"·;,\l\\\\\\\"\ ...- - - •tI, •.• \\\\\\ J

n.59x' 83,4

n-85x-59.7

-----.i--+==1--------.----U100 150

Zn - background

.. "I

i 10

1~...

00 50

Zn - anomalous

~ "1>-

H... I

oL0

ppm

Figure 15.15. Frequency distributions of Zn and Co in basaltic pillow lava from backgroundand anomalous areas within 1 km of the Skouriotissa cupriferous pyrite deposit, Cyprus (datafrom Govett and Pantazis, 1971; Govett, 1972).

Figure 15.17. Distribution of "residual" MgO and NazO infelsic volcanic rocks across the Jay copper deposit, AbitibiDistric t, Canada. " Residual" represents the differencebetween analyzed content and average value in the AbitibiDistric t for corresponding SiOz content (redrawn fromDescarreaux, 1973).

p

~..... ,

....\

\

I

I

,,"4+ OOS

12 + 005

East

5000

MeanBackground

-,_~_--

20- WestII..0u

II>

CCc'E 10..u

'"0 .. _.

~''C

ojP~

~I

1000 3000Metres

Figure 15.16. Distribution of anomalous discriminant scores for Cu, Zn, and Co acrossthe Mathiati cupriferous pyrite deposit. Cyprus (Govett and Goodfellow, 1975).


Table 15.6 SOVIET LITHOGEOCHEMISTRYSummary of trace-element dispersion in relation to

volcanogenic massive sulphide deposits

Variations in the nature of footwall and hanging wallanomalies have been noted according to the spatial relationof the deposit to the vent (Wahl, 1978). Extensive Ca and Mghalos occur in both the footwall and hanging wall of depositslocated proximal to the vent (A) (Fig. 15.20) whereas adeposit located distal to the vent (B) has an associatedfootwall anomaly but only a very restricted hanging wallanomaly. The extensive dispersion associated with theproximal deposit is attributed to deposition from acidicfumarolic brines whereas the more restricted dispersionassociated with the distal deposit is related to depositionfrom metal-rich brines that have flowed down the slope fromthe vent.

The same elements show the same types of trends onboth a regional and a local and mine scale in an Archeanmassive sulphide environment. At the Joutel-Poirier area inthe Canadian Shield, relative enrichment of Fe, Mg, and Znwith depletion of Ca, Na, and K, occurs around the deposit ona scale of hundreds of metres; these element distributionpatterns also occur regionally on a kilometre scale(Fig. 15.21, Sopuck, 1977). The same situation is evident inthe Bathurst district of New Brunswick, where enrichment ofZn and Mg and depletion of Ca and Ma occur both regionallyand in the immediate vicinity of indiVidual deposits(Whitehead and Govett, 1974; Goodfellow, 1975; Pwa, 1977;Wahl, 1978). It is reasonable to suppose that similarprocesses were operative at both the local and regional scale.

sph, ga, cpy, Zn, Pb, Ba,(py, tet, bo, (Ag) (Cu)cov)

py, sph, ga, po, Zn, Pbcpy, tet, bo

py, sph, ga, cpy Pb, Zn

py, sph, cpy Cu, Zn

cpy, py, sph Cu, Zn, S

cpy, py, sph Zn

ph, sph, cpy (po) Zn, Cu, S

Archean

It may be surmised from translated articles andother Soviet contributions over the past decade thatgeochemists in the U.S.S.R. have developed techniquesthat enable them to successfully detect blind orebodiesat depths of hundreds of metres; moreover, theyappear to be able to assess whether halos reflectprobable economic mineralization or merely the rootzone of deposits that have had the economic portioneroded away. The basis of the technique is theestablishment of zoning patterns of different elementsin halos (Fig. 15.22). Investigations of hydrothermalorebodies have shown that, despite a marked variationin composition and in local geological conditions, thehalos conform essentially to a general zoning pattern(Grigoryan, 1974); for steeply dipping bodies thesequence from top to bottom (supra-ore elements tosub-ore elements) is:

Ba-(Sb,As,Hg)-Cd-Ag-Pb-Zn-Au-Cu-Bi-NiCo-Mo-U-Sn-Be-W.

Ovchinnikov and Baranov (1972) reported a generalzonal sequence in halos around a subvolcanichydrothermal copper-pyrite deposit of Ba, Ag, Pb, andZn in the upper zones, and Cu, Mo, Co, and Bi in thelower zones; this is consistent with Grigoryan'ssequence.

Soviet geochemists generally use multi-elementratios to reduce the effects of local reversals in zoningsequences, and also to reduce analytical variations(most of their analyses are semi-quantitative spectrographic). Additive halos (addition of different elementvalues standardized to respective background),multiplicative halos (multiplication of element values),and ratios of supra-ore to sub-ore elements arefavoured rather than some of the techniques such asregression and discriminant analyses that have foundwider use in the Western world. Western geochemistsare more concerned with interpreting the absolutecontent of an element in a halo to the exclusion of its

other defining parameter - dimension. Soviet geochemists,on the other hand, use the dimension of halos as an effectiveinterpretive aid by calculating the linear productivity of ahalo; this is simply the product of the average content of anelement and the width of an anomaly (areal producti vities arealso used). It has been stated (Grigoryan, 1974) that not onlydo different deposits have similar zoning patterns, but thattheir extent, as defined by the change in linear productivitywith depth, is also similar.

The application of the procedures of multiplicative andlinear productivities is illustrated schematically inFigure 15.22. Elements (a) and (b) give supra-ore halos;elements (c) and (d) give sub-ore halos. The enhancement ofanomalies - and the clear distinction between above-ore andbelow-ore halos - using multiplicative ratios of supra-ore tosub-ore halos is obvious when compared to profiles ofelements. Changes in linear productivities with depth forsupra-ore element (a) and sub-ore element (d) are alsoillustrated. The trend of linear productivities of ratios withdepth is particularly striking (this schematic trend is typicalof those illustrated in the Soviet literature). The practicalimportance of this type of approach is the apparentpossibility of determining the location of blind deposits fromsurface samples, of predicting the likelihood of the depositsbeing economic or not, and of accurately locating mineralization from limited drillhole data. The size of the halos(according to Beus and Grigorian (1977), are 200-850 m) arecomparable to those detected by Western geochemists, butinterpretative procedures seem to be rather more advanced.

Paleozoic

Paleozoic

Age

Archean

Paleozoic

Cenozoic

Paleozoic

Archean

Zn, Cu, Pb

Metaldispersion

py, cpy, sph,ga, bar

Ore mineralogyDeposit

Buchans, CanadaThurlow et al.(1975)

Kuroko type,JapanLambert & Sato(1974)

Heath Steele,Canada Whiteheadand Govett (1974)

Killingdal, NorwayRui (1973)

Brunswick No. 12,CanadaGoodfellow(1975)

East Waite Mobrun,Joutel, Poirier,Agnico-Eagle,Mattabi, SturgeonLake, South BayMines, McConnell(1976)

Mattabi, CanadaFranklin et al.(1975)

Millenbach, CanadaSimmons et al.(1973)

356 G.J.S. Gavett and I. Nichol

WALLROCK DISPERSION SOUTH BAY MINES

:::::::::::::::::::::::::::::::::::::.

....oo+SO!

IKey:

...' Alteration• Massive sulphides/-Limit of sampling

...........,'-

ResidualNetStandardized

oo+o

-co -2 -1 +1 +3 ...co oI

200ftI

Figure 15.18. Distribution of "standardized net residual" at South Bay lv1ine (McConnell, 1976).

Mg

.~

a 6

au"-

'"::l:4

verylow

:'1~

low~

background hi g h •veryhig h

50 100 150Metres from ore zone

_ A,S massivesulphides

ometres

3000

Figure 15.19. Distribution of Mg/Ca ratio in footwall felsicvolcanic rocks, Brunswick No. 12 deposit, New Brunswick(reproduced from Govett and Goodfellow, 1975).

Figure 15.20. Schematic distribution of Ca and Mg in felsicvolcanic rocks around proximal (A) and distal (B) massivesulphide deposits, New Brunswick (data from Wahl, 1978).


CoO

JOUTEL -POIRIER, Local

Zn-ppm

x <50o 50-100• >100

CoO

JOUTEL -POIRIER. Regional

Zn-ppm

x <50o 50-100• >100

o o

o

o

•

•

Figure 15.21. Relation between geochemistry and the degree of alteration on thelocal and regional scale at Jautel-Poirier (Sopuck. 1977).

oI

500I

Q.

element A element B element C

---r<.--- -- Level I

-Levell!

-Level III.

-Level IV.element 0

b.10

Level I.~Level IV.

1.O{\ 1\0.5

500m

c.

o

0.1

Levell.

500m

Level II.

o

Level III.

vLevel IV.

v0.05 -"------

d.Level I.

--Leveili.---.sf---- ..,

.0./ I." I." .

0.10.01f--------r--------r--~---r'----.,---------"

1.0 10 1000.001

Figure 15.22. Schematic illustration of supra-ore and sub-ore lithageochemical halos, and anomaly enhancement through the use of multiplicativehalos and linear praductivities (after Govett, 1977).


identify pDtentiallys"dimentary horizonsare detectable over

Evaluation of State-of-the-Art and Conclusions

The existence of diagnostic lithogeochemical signatureshave been recognized on a number of scales associated withdifferent types of deposits. These have a variety ofapplicotions in explorotion.

a. anomalous metal contents of mineralized relative tounminerali7ed intrusions e.g., (1) Sn (Australia), W(Canada), Sn (United Kingdom) and Ni, Cu (Canada).These features are rlttributed to thf' high metal content ofthe parent magma that gave rise to the intrusives ,md theassociated deposits. (2) Cu, Mo (United Kingdom)resulting from post-magma hydrothermal activity in thehost intrusi ons.

b. Flnomalous metal contents of mineralized zones ofintrusions relative to unmineralized zones e.g., pDrphyrycopper deposi ts in general resulting from the permeationof hydrothermal fluids into the intrusion from the centreof mineralization.

c. localized anDmalous metal contents in the wall rockadjacent to vein-type mineralization e.g., Pb, Zn, AgminerFdization in Utah as a result of mineralizing fluidsdiffusing into the wall rock.

d. anomalous metal cDntents in favourable areas Dfsedimentary basins for mineralization e.g., McArthurRiver, Australirlj Tynagh, Ireland; Meggen, Germany.

These patterns are related to the exhalation of metal-richsDlutions into the sea and subsequent oeposition with andoecay of certain metal cDntents over cDnsiderabledistances from the depDsits.

e. ,momalous met.al contents in prDductive relative tononproductive cycles, and Dn more local scales associat edin genernl with vDlcanogenic massive ~1ulphide depDsit.s.These fpatures apparently result from hydrotherrTlalactivity associated with the phase of mineraliz;tiDn.

These diagnostic l]eochemical parameters havevariously been identified on the basis of whole rDckcomposition, analysis of separate mineral phases or thedistinction of subtle variatiDns in the form of populations.

Equally impDrtant in terms of drawing attention to thelimitations of lithogeochemistry in cxploration has been thpdemDnstration that the compositiDn of the whole rock orminerrll phases may nDt be diagnDstic Df prDximity tominernlizatiofl. Since the compositiDn Df rocks is the resultof thre overrlll geDlogical history of the sample thecompositiDn represents the effects of a number Df processesthat hnve affected the sample. Only in cases where themineralizing episode is the Dver-riding proccss will effectsdue to mineralization be readily identifiable. In Dther caseseffects on cDmpDsition due to other causes must be taken intDaccount. before any respDnse due tD mineralization can berecDgnized. This feature is illustrated by thc need to takeintD account the effecU1 of fractionation in the interpretationof data relating tD igneDus rocks if a variety Df Ii thologictypes are sampled in order t.o identify respDnses due tomineralization. It is thus vital at the interpret<ltion stage totake into account those aspects of the geologic<11 history thathave a bearing on the geDchemistry in order to identifymeaningful responses due to mineralizntion. The lack ofgeneral acceptance Df Ii thDgeochemistry as an aid inexploration is undoubtedly due to a lack Df recogni tion ofdiagnostic patterns relatpd to in part petrogenesis, tectonic,and overall geolDl]ical history.

In many Df the examples cited in the review theinvestigations have been carried out over areas of knownmineralization where the geolDgical knowledge is relativelyhigh. Whilst thRse locat.ions are logical focal pDints for

establishing the nature of dispersion assDciated withmineralizatiDn it may be that in view Df the suhtlety of manyof the respDnses, they have only been recDgnized asanomalous responses becrluse of the relattvely gODd geologicaldata base.

The diversity of geochemical responses relatRd tDmineralization in terms of anDmalous elements, lateralextent, and cDntrast from different examples of the sametype of depDsit clearly is an adverse faetDr affecting theapplicability of lithogeochemistry in explDration. Thisferlture is illustrated by the div"rsity in elements showinganomalous responses, extent, and magnitude of the anomalDusdispersions assDciated with four porphyry-type depDsits in theHighland Vnlley, British Columbia.

In explDratiDn over relatively unexplDred t.errain t.henecessary level of geological information tD <lidinterpretation is not availFlble or can Dnly be Dbtained atcDnsiderable expense which must cDnstitute a furthercDnstraint on the applicabili ty of the technique. It isprobrlble thnt a reconnaissance survey r:arried out overrelatively unexplDred ground will generate a multitude ofsubtle anDmalies similar to those assDciuted with knDwnmineralization but in the absence of adequate geolDgicalinformation it will be impossible to identify those anomaliesreflecting mineralization from those related to other causes.

Notwithstanding these limitations, within the past tenyears in the Western Vlorld lithogeochemistry has advancedfrom an essentially academic involvement to iJ technique usedsparingly by t.he mining industry in mineral exploration.However, we iJre not aware of any mineral discDvery to datethat can be attributed tD lithogeochemistry.

The use of lithDgeochemistry on pre-drilling explorationis dependent upon adequate surface outcrDp or expensivesurface drilling thrDugh Dverburden to bedrDck. Sampling,therefore, is generally mDre costly them in stream sedimentor soil geochemical surveys; crushing and grinding of srlmplespreparatory to analysis is at lerlst five times more expensivethan sieving soil Dr stream sediments.

Nevertheless, there are a number Df explDrationsituations where lithogeDchemistry is useful and may beexpected tD become increasingly important in the future.These arc:

1. RegiDnal scale explDration tDprDductive plutons ano volcanic or(anD maiD us gcochemical patternsmany kiIDmetres).

2. LDcal scale exploration to locate deeply-buried and blinddepDsits (geDchemical halos have dimensiDns of mDre t.han100 m rlnd can range up tD 1 km).

3. ExplDration drilling tD ussist in drill hole location.

4. Mine sCrlle exploratiDn and undergrDund mappinl].

As indicated previously, the naturp of geDchemicalresponses in bedrock associated with mineralizatiDn variesconsiderably in terms of anDmalous element assDciations,extent and magnitude of anomalies amongst differentexamples of the same type of deposit. This situation relatedin part to conditiDns in the geological environments preventsthe definitiDn of any standardized procedures. Perhaps morethan in any Dther branch of geochemical explDration, it isessential tD carry out an orientation survev over knDwnmineralization 'to establish the nature Df' geochemicaldispersiDn <1ssociated with known mineralization andappropriate sampling, analytical and interpretatiDnalprocedures. These operational prDcedures may Dnly beapplicable in a precisely similar geological environment.

Lithogeochemistry 359

1969: Geochemical criteria for assessment of the mineral

REFERENCES

Beus, A.A.

Descarreaux, J.1973: A petrochemical study of the Abitibi volcanic belt

and its bearing on the occurrences of massivesulphide ores; Can. Inst. Min. Met., Bull., v. 66,no. 730, p. 61-69.

Franklin, J.M., Kasarda, J., and Poulsen, K.H.1975: Petrology and chemistry of the alteration zone of

the Mattabi massi ve sulphide deposit; Econ. Geo!.,v. 70, no. 1, p. 63-79.

Garrett, R.G.1971: Molybdenum, tungsten, and uranium in acid

plutonic rocks as a guide to regional exploration,S.E. Yukon; Can. Min. J., v. 92, p. 37-40.

Gjelsvik, T.1968: Distribution of major elements in the wall rocks

and the silicate fraction of the Skorovass pyritedeposit, Grong area, Norway; Econ. Geol., v. 63,p.217-231.

Goodfellow, W.D.1975: Major and minor element halos in volcanic rocks at

Brunswick No. 12 sulphide deposit, N.B., Canada; inGeochemical Exploration 1974, Elliott, I.L. arldFletcher, W.K., ed., Elsevier Pub I. Co.,Amsterdam, p. 279-295.

Brabec, D. and White, W.H.1971: Distribution of copper and zinc in rocks of the

Guichon Creek Batholith, British Columbia; inBoyle, R.W. and McGerrigle J.I., ed., Geochemic31Exploration, Can. Inst. Min. Met., Spec. Vo!. 11,p. 291-297.

Bradshaw, P.M.D.1967: Distribution of selected elements in feldspar,

biotite, and muscovite from British granites inrelation to mineralization; Trans. Inst. Min. Met.,London, v. 76, p. B117-B148.

Cameron, E.M.1975: Geochemical methods of exploration for massive

sulphide mineralization in the Canadian Shield; inElliott, I.L. and Fletcher, W.K., ed., Geochemic31Exploration 1974, Elsevier Pub!. Co., Amsterdam,p. 21-49.

Cameron, E.M., Siddeley, G., and Durham, C.C.1971: Distribution of ore elements in rocks for evaluating

ore potential: nickel, copper, cobalt, and sulphur inultramafic rocks of the Canadian Shield; inBoyle, R.W. and McGerrigle, 5.1., ed., Geochemic31Exploration, Can. Inst. Min. Met., Spec. Vol. 11,p. 298-313.

Chaffee, M.A.1976: The zonal distribution of selected elements above

the Kalamazoo porphyry copper deposit, SanManuel District, Pinal County; J. Geochem.Explor., v. 5, no. 2, p. 145-165.

Coope, J.A.1977: Potential of lithogeochemistry in mineral

exploration; Paper presented at AIME AnnualGeneral Meeting, 7 March 1977, Atlanta, Georgia.

Davenport, P.H. and Nichol, I.1973: Bedrock geochemistry as a guide to areas of base

metal potential in volcano-sedimentary belts of theCanadian Shield; in Geochemical Exploration 1972,Jones, M.J. (Editor), Inst. Min. Met., London,p. 45-57.

duringMines

seriesSch.

rockColo.

In terms of sampling the nature of the optimum sampleshould be established i.e., grab or composi te chip sample, andsample interval or density to ensure adequate representivityin terms of the target population. Sample preparationprocedure involving preparation of whole rock or mineralseparates should be considered. With regard to analyticalprocedures attention should be given to identifying elementsthat display diagnostic distributions and ways of analyzingthese in terms of obtaining the necessary sensitivity,accuracy, and precision and in some cases the metal held inparticular mineral phases e.g., sulphide-held meta!. In orderto identify the necessary interpretational procedures it willbe essential to identify those features of the geologicalenvironment that have a bearing on geochemical dispersion.

At the end of 1977 there is abundant evidence thatgeological events - including those that give rise to mineraldeposits are detectable by lithogeochemistry, givenappropriate geological interpretation. Future developmentand success will depend upon improved understanding ofgeological processes and the mechanisms of elementmigration and fixation. The fact that lithogeochemicalsurveys are more costly to conduct than stream sediment orsoil surveys means that the major application of rockgeochemistry in the near future is likely to be restricted toareas that have already been well-prospected by moretraditional techniques. Moreover, since multi-element dataare generally required and interpretation is correspondinglydifficult, lithogeochemical techniques are likely to be limitedby the availability of advanced analytical technology andhighly-trained geochemists. To the extent that the miningindustry is prepared to invest in renewed exploration in thewell-prospected areas of North America and Europe,lithogeochemistry provides a promising new technique.

potential of the igneousreconnaissance exploration;Quart., v. 64, no. 1, p. 67-74.

Beus, A.A. and Grigorian, S.V.1977: Geochemical Exploration Methods for Mineral

Deposits; Applied Publishing Co., Wilimette,Illinois, 287 p.

Allen, P.M., Cooper, D.C., Fuge, R., and Rea, W.J.1976: Geochemistry and relationships to mineralization

of some igneous rocks from Harlech Dome, Wales;Trans. Inst. Min. Met., London, v. 85, p. BI00-BI08.

Bailey, G.B. and McCormick, G.R.1974: Chemical halos as guides to lode deposit ore in the

Park City District, Utah; Econ. Geo!., v. 69, no. 3,p. 377-382.

Barsukov, V.L.1967: Metallogenic specialization of granitoid intrusions;

in Vinogradov, A.P., ed., Chemistry of the Earth'sCrust, Vol. II (Jerusalem: Israel Program forScientific Translation), p. 211-231.

Bignell, R.D., Cronan, D.S., and Tooms, J.S.1976: Metal dispersion in the Red Sea as an aid to marine

geochemical exploration; Trans. Inst. Min. Met.,London, v. 85, p. B274-B278.

Bolotnikov, A.F. and Kravchenko, N.S.1970: Criteria for recognition of tin-bearing granites;

Akad. Sci. USSR Dok!. Earth Sci. Sect., v. 191,p. 186-187.

360 G.J.5. Govett and I. Nichol

Govett, G.J.S.1972: Interpret.ation of a rock geochemical exploration

survey in Cyprus statistic<ll and graphicaltechniques; J. Geochcm. Explor., v. 1, p. 77-102.

1976: The development of geochemical explor<ltionmethods and techniques; in Govett, G.J.S. andGovett, M.H., ed., World- Min~ral SuppliesAssessment and Perspecti ve, Els"v ier Pub!. Co.,Amsterdam, p. 343-376.

1977: Presidenti<ll Address to the Annual GeneralMeeting of the Association of ExplorationGeochemists, Vancouver, B.C., April 1977; J.Geochem. Explor., v. 8, p. 591-599.

Govett, C.J.5. and Goodfellow, W.U.1975: IJevelopmenl of rock geochemical techniques for

detecting buried sulphide deposits - a discussiorl;Trans. Inst. Min. Met., London, v. 84, p. B134-B140.

Govett, G.J.S. and PanLJzis, Th.M.1971: Uistribution of Cu, Zn, Ni, and Co in the Troodos

Pillow Lava Series, Cyprus; Trans. InsL Min. Met.,London, v. 80, p. B27 -B46.

Grigoryan, S.V.1974: Primary geochemical halos in prospecting and

exploration of hydrothermal deposits; Int. Geo!.Rev., v. 16, no. 1, p. 12-25.

Cunton, J.E. and Nichol, I.1975: Ch<:mical zoninl) associated with the Ingerbelle

Copper Mountain mineralization, Princeton, BritishColumbia; in Elliott, I.L. and Fletcher, VI.K., ed.,GeochemicaT Exploration 1974, Elsevier Publ. Co.,Amsterdam, p. 297-312.

Gwosdz, W. and I<rebs, W.1977: Manganese halo surrounding Meggan ore neposit,

Germany; Trans. Inst. Min. MeL, London, v. 86,p. B73-B77.

Hesp, W.R. and R.igby, D.1975: Aspects of tin metallogenesi,; in the Tasman

geosyncline, Eastern Australia, as reflecled bycluster and factor analysis; J. Ceochem. txplor.,v. 4, no. 3, p. 331-347.

Ivanova, G.F.19113: The content of tin, tungsten, and molybdenum in

granite enclosing tin and tungsten deposits;Geochemistry, v. 5, p. 492-500.

Jacobs, D.C. and Parry, VI.T.19711: A comparison of the l)eochemistry of biotite from

some Basin and Range Stocks; Econ. Geo!., v.71,p. 1029-1035.

Jambor, J.L.1974: Trace element variations in porphyry copper

deposits, Babine Lake Area, B.C.; Geo!. Surv. Can.,Paper 74-9, 30 p.

James, C.H.1967: The use of the terms "primary" and "secondary"

dispersion irl geochemical prospecting; Econ. Geo!.,v. 62, p. 997-999.

Kesler, S.L, Issigonis, M.J., Brownlow, A.H., Damon, P.E.,Moore, W..l., Northcote, K.E., and Prdo, V.A.

1975a: C;eochemistry of biotites from mineralized andbClrren intrusi ve systems; Econ. Geo!., v. 70,p. 559-567.

Kesler, S.E., Issigonis, M.J., and Van loon, J.C.1975b: An evaluation of the use of halogens and water

abundances in efforts to distinguish mineralizedand barren intrusive rocks; J. Geochem. Explor.,v. 4, p. 235-245.

Kesler, S.E., Van Loon, J.e., and Moore, C.M.1973: Evaluation of ore potential of granodioritic rocks

usinl) water-extractable chlDride and fluoride; Can.Inst. Min. Met. Bul!., v. 66, no. 730, p. 56-60.

Kuroda, Y.1961: Minor elements in metasomatic zone re1::lled to a

copper-bearing pyrite deposi t; Econ. Geo!., v. 56,p.847-854.

Lambert, I.B. and Sato, T.1974: The KlJfoko and associated ore deposits of Japan: a

review of their fealures and metallogenesis; Econ.Geo!., v. 69, p. 1215-1236.

Lambert, I.B. and Scott, K.M.1973: Implications of g<:ochemical investigations of

sedimentary rocks within and around the McArthurzinc-Iead-si Ivel' deposi t, Northern Terr itory; J.Geochem. Explor., v. 2, p. 307-330.

Lavin, Cl.P.1976: Lithogeochemical discrimination between

mineralized and unmincralized cyclcs of volcanismin the Sturgeon Lake Clnd Ben Nevis areas of theCanadian Shield; M.Sc. thesis, Queen's University,Kingston, Clntario, Canada, 249 p.

Lowell, J.D. and Guilbert, J.M.1970: Lateral and vertical alteration-mineralization

/oning in porphyry ore deposits; Econ. Geoi., v. 65,p. 373-408.

Mantei, E. and Brownlow, A.1967: Variation in gold cont<:nt of minr'rals of the

Marysville quartz diorite stock, Montana; Geochim.Cosmochim. Acta, v. 31, p. 225-236.

Mantei, E., Bolter, E., ann Al Shaieb, l.1970: Distribution of gold, silver, copper, lead, and zinc

in the productive Marysville stock, Montana;Mineral. Depm;ita, v. 5, p. 1fl4-190.

McConnell, J.W.1976: Geochemical dispersion in wallrocks of Archean

massive sulphide deposits; M.Sc. thesis, Queen'sUniversity, Kingston, Ontario, 230 p.

Mohsen, L.A. and Brownlow, A.H.1971: Abundance and distribution of manganese in the

western part of the Philipsburg Batholith, Montana;Econ. GeoI., v. 66, p. 611-617.

Nichol, I.1975: Bedrock composition as a guide to areas of base

metal potential in the greenstone belts of theCanadian Shield; Queen's UnivNsity, Kingston,Ontario, Unpubl. Repl., 74 p.

Nichol, I., Bogle, E.W., Lavin, D.P., McConnell, J.W., andSopuck, V.J.

1977: Lithogeochemistry as an aid in massive sulphideexploration; in Prospecting in areas of glacialedterrain, Jones,- M.J., cd., Inst. Min. Met., London,p.63-71.

Nichol, I., Lavin, D.P., McConnell, J.W., Hodgson, C.J., andSopuck, V.J.

1975: Bedrock composition as an indicator of Arche~Jn

massi ve sulphide environments (abstract); Can.Inst. Min. Met., v. 68, no. 755, p. 48.

Lithogeochcmistry 361

Nilsson, C.A.196B: Wall rock alteration at the Boliden deposit,

Sweden; Econ. Geol., v. 63, p. 472-494.

Olade, M.A.1977: Major element halos in granitic wall rocks of

porphyry copper deposi ts, Guichon Creek Batholith,British Columbia; J. Geochem. Explor., v. 7, no. 1,p.59-71.

Olade, M.A. and Fletcher, W.K.1975: Primary dispersion of rubidium and strontium

around porphyry copper deposits, Highland Valley,British Culumbia; Econ. Geol., v. 70, p. 15-21.

19760: Trace element geochemistry of the Highland Valleyand Guichon Creek Batholi th in relation toporphyry copper mineralization; Econ. Geol., v. 71,p.733-748.

1976b: Distribution of sulphur, anel sulphide-iron andcopper in bedrock associated with porphyry copperdeposits, f lighland Valley, British Columbia; J.Geochem. [xplor., v. 5, no. 1, p. 21-30.

Ovchinnikov, L.N. and Baranov, E.N.1972: Endogenic geochemical halos of pyritic ore

deposits; Int. Geol. Rev., v. 14, no. 5, p. 419-429.

Oyarzun, M.J.1971: Contribution a l'etude des roches volcaniques et

phutomiques du Chile; These, Univ. de Paris, 195 p.

1975: Rubidium and strontium as guides to coppermineralization emplaced in some Chilean andesiticrocks; in Geochemical Fxplor3tion 1974,Elliott, I.e and Fletcher, W.K., ed., r-lsevier Publ.Co., Amsteruam, p. 333-338.

Page, B.G.N. and Conn, H.1973: Investigacion sobre metodos de prospeccion

geoquimica en el yacimiento tipo cobre porfidico,El I1Ibra, provincia de Antofag8sta-Chile InstitutoInvestigaciones Geologicas, Santiago, 40 p.

Pantazis, Th.M. and Govett, C.J.S.1973: Interpretation of a detailed rock geochemical

survey around Mathiati Mine, Cyprus; J. Geochem.Explor., v. 2, no. 1, p. 25-36.

Parry, W.T.1972: Chlorine in biotite from Ilasin and Range Plutons;

Econ. (-;eol., v. 67, p. 972-975.

Parry, W.T. and Jacobs, D.C.1975: Fluorine and chlorine in biotite from Basin and

Range Plutons; Econ. Geol., v. 70, p. 554-558.

Parry, W.T. and Nackowski, M.P.1963: Copper, lead and zinc in biotites from Basin and

I~ange quartz monzonites; Econ. Geol., v. 58,p. 1126-1144.

Pwa, U. Aung1977: Regional rock geochemical exploration, Bathurst

District, N.B.; M.Sc. thesis, University of NewBrunswick, Fredericton.

Rui, I.J.1973: Structural control and wall rock alteration at

Killi ngdal mlne, central Norwegian Caledonides;Econ. Geol., v. 68, p. 859-883.

Russell, M.J.1974: Manganese halo surrounding the Tynagh ore

deposi t, Ireland: a preliminary note; Trans. Inst.Min. Met., London, v. 83, p. B65-B66.

1975: Lithogeochemical environment of the Tynagh basemetal deposit, Ireland, and its bearing on oredeposition; IMM Sect. B, v. 84, p. 128.

Sakrison, H.C.1966: Chemical studies of the host rocks of the I_ake

Dufault Mines, Quebec; Ph.D. thesis, McGillUniversity, Montreal.

Sangster, D.F.1972: Precambrian volcanogenic m8ssive sulphide

deposi ts in Canada: A Review; CeDI. Surv. Can.,PLlper 72-22, 44 p.

Sheraton, J.W. and Black, L.P.1973: Geochemistry of mineralized granitic rock of

northeast Quecnsland; J. Ceochem. Explor., v. 2,p.331-348.

Simmons, B.D. and the Geological Staff, Lake Dufault Div.,F alconbridge Copper Ltd.

1973: Geology of the Millenbach massive sulphidedeposit, Noranda, Quebec; Can. Inst. Min. Met.,v. (,;6, no. 739, p. 67-78.

Simpson, P.R., Plant, J., and Cope, M.J.1977: Uranium abundance and distribution in some

granites from northern Scotland and southwestEngland as indicators of uranium provinces; Int.Syrnp. Inst. Min. Met., London, p. 126-139.

Sopuck, V.J.1977: A lithogeochemical approach in the search for

are8S of felsic volc;mic rucks associated withmineralization in the Canadian Shield; Ph.D.thesis, Queen's University, Kingston, Ontario,400 p.

Tasumi, T.and Clark, L.A.1972: Chemical composition of acid volcanic rocks

genetically related to formation of the Kurokodeposits; Ceol. Soc. Japan J., v. 78, p. 191-201.

Tauson, L. V.1967: Geochemistry of rare elements in igneous rocks and

metallogenic specialization of magmas; ~

Chemistry of the Earth's Crust, Vinogradov, A.P.,ed., Vol. II (Jerusalem: Israel Program forScientific Translation), p. 248-259.

Tauson, I .V. and Kozlov, V.D.1973: Distribution functions and ratios of trace-element

concentrations as estimators of the ore-bearingpotential of granites; in Geochemical Exploration1972, Jones, M.J., ed:;- Inst. Min. MeL, London,p. 37-44.

Theodore, T.G. and Nash, J. T.1973: Geochemical and fluid zonation at Copper Canyon,

Landor County, Nevada; Econ. Ceol., v. 68,p. 565-570.

Thurlow, J.G., Swanson, E.A., and Strong, D.F.1975: Geology and lithogeochemistry of the Buchans

polymetallic sulfide deposits, Ncwfoundland; Econ.Geol., v. 70, p. 130-144.


Wahl, J.L.1978: Rock geochemical exploration and ore genesis at

Heath Steele and Key Anacon deposits, NewBrunswick; Ph.D. thesis, University of NewBrunswick, Fredericton.

Wahl, J.L., Govett, G.J.S., and Goodfellow, W.O.1975: Anomalous element distribution in volcanic rocks

around Key Anacon, Heath Steele, B-zone andBrunswick No. 12 sulphide deposits (abstract); Can.Inst. Min. Met., Bull., v. 68, no. 755, p. 49.

Whitehead, R.E.S.1973: Environment of stratiform sulphide deposition:

variation in Mn:F e ratio in host rocks at HeathSteele Mine, New Brunswick; Canada; Mineral.Deposita, v. 8, p. 148-160.

Whitehead, R.E. and Govett, G.J.S.1974: Exploration rock geochemistry - detection of trace

element haloes at Heath Steele Mines (N.B.,Canada) by discriminant analysis; J. Geochem.Explor., v. 3, p. 371-396.

Wolfe, W.J.1974: Geochemical and biogeochemical exploration

research near early Precambrian porphyry-typemolybdenum-copper mineralization, NorthwesternOntario, Canada; J. Geochem. Explor., v. 3,p. 25-4l.

1975: Zinc abundance in Early Precambrian volcanicrocks: its relationship to exploitable levels of zincin sulphide deposits of volcanic-exhalative origin;in Geochemical Exploration 1974, Elliott, I.L. andFletcher, W.K., ed., Elsevier Publ. Co.,Amsterdam, p. 261-278.

LITHOGEOCHEMISTRY IN MINERAL EXPLORATION...chimie de la roche favorable et de La nature des...

Documents

Transcript of LITHOGEOCHEMISTRY IN MINERAL EXPLORATION...chimie de la roche favorable et de La nature des...