1 Identification and interpretation of ferruginous rock in tropical environment

1.3 Expression of sulfidic rocks in tropical zone

In tropical environment, where elements from the landscape are ruled by the importance of weathering and ironcrust formation, there is obviously a problem of identification and typology of ferruginous rocks.


After Routhier (1963), the discrimination between ironcrust and gossan was impossible, both being neogenetic residual ferruginous rocks. Despite the further improvements of studies on tropical weathering, this assertion was little contested by prospectors.


In Australia and elsewhere (Butt and Smith; 1980 ; Ryall and Taylor, 1981 ; Taylor and Thornber, 1995), an approach through landscape hierarchized classification permitted to point out a typology taking in account the different tropical environments (Fig. 7). Generally gossans are distinct from the overlying ironcrusts, despite some exceptions like at Golden Grove (Smith and Perdrix, 1983). On that account similar occurrences were observed in West Africa, like at Perkoa and Kwademen (Burkina Faso), where gossans developed in recent weathered formations at the foot of ironcrust relieves. The situation is partially similar at Pagala (Togo) where gossans developed in the erosion surface of the Adélé plateau bearing ironcrusts or not. In the environment of Pagala, the weathering of sulfides is partly precocious as compared to general weathering and in gossans area, it is posterior to erosion.

Fig. 7 

Typology for ferruginous outcrops of Australia after Butt, Smith, Ryall, Taylor, Thornber…


There are so many ferruginous outcrops in West Africa, that any of the latter, if distinct from an ironcrust, may be a possible mineralization marker, should there be oxidized, carbonated, phosphatic or sulfidic weathered mineralizations or specific concentration phenomena inducing iron mobilization and oxidation (for example, the “fault ironstone” of Butt and Smith, 1980). This approach is a preliminary to any more thorough use of ferruginous rocks in diagnoses of prospections, but this is already an important way of discrimination. In fact, West-African region is hardly comparable to the tropical mining districts of Australia, Brazil and South Africa since the participation of ancient weathering to landscapes is very important. This predominance obliterates the major part of non-lateritic ferruginous rocks which were also affected by strong agressions. Beside the practical aspect of the discrimination of gossans in ironcrust environment (Blot, 2004), it remains to define the distinctive features of gossans in general from those specific of gossans from tropical zone, without consideration for rarely accessible non-ferruginous weathering horizons.


a Present - day gossans of Togo and Burkina


In absence of mine working, gossans are not studied in Burkina, even if they are known partial approach, in as much as exploitation drillings are little available and badly adapted to the studies on supergene weathering. In Togo, a mine-devoid country too, the study of the gossans of Pagala is prior to exploitation drillings and keeps restricted to outcrops and to some suboutcrops, to which information is thus limited, setting an unaccostumed problem to prospectors. Of course, some sections could be used, but exploitation drillings were implanted in relation to an investigation of the primary mineralization, so that the whole of the weathering horizons remains little known.


In Togo as well as in Burkina, gossan is a ferruginous rock with high contents of iron oxides and hydroxides, without known accessible rooting. Its specific features are the following : chemical composition with prevalent iron, with or without manganese, with more or less subordinate silica and no other well preserved major elements ; the alumina content is rarely higher than 5% (table 10). The elements of the mineralizations are well represented, but contents higher than 1% are exceptional (table 11). No secondary minerals could be identified except ferriferous and manganiferous ones, as well as some secondary quartz. And, this is quite different from what is observed in the most mature formations described by Scott et al. (2001) : in West Africa no secondary mineral is detected on outcrop in association with iron oxides and hydroxides.


Perkoa (VMS Zn-Ag-Ba), Kwademen (VMS Ni-Cu), Bonga (lateritic Ni), PAG = 4 gossans from the area of Pagala (Togo), Tin Akoff : composite mean


Table 10 

Average composition of gossans from Togo and Burkina


Perkoa on sulfidic Zn pocket (Ag, Pb), Ni-Cu index of Kwademen, lateritic Ni ore of Bonga ; average contents of the 10% maxima in Tin Akoff (Burkina) and Pagala (Togo) areas.

For Burkina: Blot’s samples ; for Togo Anglo American’s samples.


Table 11 

Average contents (in ppm, ppb*) of gossans from the West African tropical zone 

Before giving a comparison between gossans of Central Togo and North Burkina (section 2), we present the state of the art on the existence of secondary ferruginous rocks as related to primary sulfidic, carbonated or eventually phosphatic mineralizations.


As I discovered the originality of these ferruginous outcrops, first in Togo and then in Burkina, I had emphasized the importance of the facies as distinctive features from the widespread lateritic ironcrusts.


In Togo, the first “labelled” gossan dates from 1982, the first tenth from 1983 (Blot, 1983) and the first hundred from 1984 (Magat and Blot, 1989), which gave rise to doubt and mistrust, in the mind of the prospectors a priori convinced of the small mining interest of surficial covers (Bessoles and Trompette, 1980). The Bureau de la Recherche géologique et minière (BRGM) (Deschamps, 1991) as well as Anglo American (Napon et al., 2002 ; Presseau et al., 2002) record more than 300 sites in the district of Pagala during their prospectings.


The primary mineral phase was represented by quartz and often barytic muscovite, and the secondary phase by large-sized and often well-crystallized oxidized or hydroxyle iron minerals, sometimes associated with clear euhedral quartz and Mn-oxides.


The first semi-quantitative and partial geochemical analyses permitted to highlight the importance of Zn and Mn, then Ba as well as abnormal Co contents and high Ni amounts. These observations were complemented by a first inventory (Magat and Blot, 1984, 1987 ; Godonou et al., 1986) which fined the definition, composition and geochemistry of these ferruginous outcrops. From that time, the mining companies in fine did not really discuss the definition admitted for the principal metallic concentrations. Massive kaolinic (associated kaolinite and metahalloysite) or gibbsitic clays observed at the contact of gossan nr 3 seem to be typically hydrothermal. Alunite, identified in the borings of Anglo American did not jam in the different sections and seems to be anecdotic. The absence of secondary minerals except the ferriferous ones was mentioned in his turn by Theron (2002) who admitted that this concerned “tropical” gossans specific of West Africa.


The forecast of the polymetallic nature of a stratiform mineralization was reinforced by, Pb, Ba, Ni, As and sometimes Cu. The origin of gossans seemed to be sulfidic (frequent residual pyrite), carbonated (importance of siderite ghosts) and locally phosphatic from the phosphorites of Pagala (turquoise, wavellite and gorceixite associated with iron hydroxides). Fig. 8.


Fig. 8 

Peculiar gossan: geochemical compositions of ferriginous surficial rocks developed from phosphorites, sideritite and iron oxides from Pagala area (Togo). Clarkes of concentration. As and Pb were not analyzed.

In Burkina, gossans are known in the Birrimian, in the Boromo-Houndé furrow from the UNDP geochemical prospections between 1976 and 1982, at Perkoa and Kwademen, together with a lateritic Ni ore at Bonga or in the Po area at Tiébélé ; the latter was found during the prospecting of basic metal by Anglo American (Robb, 2000).


Ores or indicators found during geochemical prospecting whose reply is very variable according to the prospecting, likely attest a different mobility between Pb and Zn, the principal elements of prospectings. The outcrops do not mark the landscape, because they are below the ironcrust-bearing surfaces, eventually with blocks of massive facies. The outcrops of Perkoa (Franceschi and Ouédraogo, 1982) about 250m long and 20m wide gives rise to a metric ripple on the plain. Its mineralogical composition is the following : prevailing goethite or hematite or both, with sometimes magnetite, as well as thin veins of gorceixite, barytine and Mn minerals.


Kwademen is known above all by its gossan and the variety of the successive prospections targets, i.e. Zn, Cu then Au. The only mineralization identified by drilling is massive and composed of pyrrhotite, pyrite and magnetite in which Ni and Cu contents are 0.4 and 0.5% respectively but these elements are not found inside minerals. The gossan is well oriented 500m long and 10m wide, and does not mark the landscape. Beside goethite, hematite and quartz, asbolane generated from spessartine can be identified. Geochemical analysis highlights the importance of Ni and Cu, as well as clear Te, Pt and Au contents. Goethite is characterized by notable Cu and Ni contents (Fig. 9).


Fig. 9 

Analyses of goethites from Kwademen by means of microprobe (in atoms per 100 Fe atoms).


In the Birrimian of Burkina, it seems that no specific studies have been performed on gossans as by-products of geochemical prospectings. There are only scarce informations on these surficial residual rocks and the author disposes only of his own collection to outline their principal features.


In the example of the lateritic Ni ore of Bonga, the ironcrust which takes the outlines of the hillslope contains very iron-rich blocks without rooting, whence a denomination of paleo-gossan (Lavaud, 2002 ; Lavaud et al., 2004). Goethite is the prevalent mineral of the massive facies, but Ni is also found in numerous residual primary minerals as well as in secondary ones (table 12)


Table 12 

Ni, Cr, Ti, Fe contents in weight percent of some minerals from Bonga (hematite and goethite here in the form of weathering plasma) – Lavaud, 2002.


Last in the Neoproterozoic of North Burkina, special ferruginizations were described, among which a good part would be gossans, close to primary ferriferous mineralizations with little remobilization of goethite. These formations will be detailed in the second part of this study. Here the only principal features of likely sulfidic mineralizations of the tropical zone will be specified. The mineralogical composition is the most simple, including largely prevalent goethite, and hematite, siliceous components like quartz, quartzite, silexite, chert…. and kaolinite in cavities and inclusions to be probably related to weathering of pelitic rocks. No other mineral has been observed on surface outcropping. There are only three principal chemical elements that is, Si, Fe and H2O, more rarely manganese. The total amount of minor and trace elements is low. However, some elements whose population has a median higher than the clarke display rather regularly high contents, as Be, V, Co, Zn, As, Mo, Sb, Bi, Re, Au, U and by definition Fe. Another group of elements is characterized by medians lower than the clarke, but by a mean higher than the clarke, as Mn, Cu, Se, Cd, Te, Y, Pd and Pt.


In these West-African cases, the chemical composition’s characteristics are little different from what is known elsewhere, with small amounts of alumina, alkaline and alkaline-earthy elements whilst manganese oxide accumulations are frequent. The trace elements contents are rather low as compared to Australian and Arabian references and rarely reach 1%. The comparison between geochemistry of gossans and that of parental mineralizations, does not indicate the same evolutions at Pagala and Perkoa, in a state of great maturation ,i.e. without any other secondary minerals than goethite and hematite (table 13). Identical elemental behaviours in the ore deposit and gossan are rare at both sites; in fact, similar Zn and Cd impoverishment and Ba and Mn enrichment are observed, whilst As and Mo behave differently from one site to the other. Thus, here and elsewhere (table 7) and corresponding all the elements, no evolutive law between a given ore and the corresponding end-product of weathering can be established because of the multiplicity of the parameters involved.


Contents in ppm and percent. Collections Billiton (Perkoa), Anglo American (Pagala), Blot (Tin Akoff). Dispersed facies of ores from Perkoa and Pagala, sampled by drilling. Average contents and means of the 10% maxima for surficial rocks from Pagala and Tin Akoff


Table 13 

Average contents of the principal minerals of ores and gossans from Perkoa and Pagala, and of ferruginizations from Tin Akoff. 


b Other occurrences of mineralizations weathering in tropical zones


We attempted to find elsewhere what is the specificity of gossans, as compared to ironcrust formations, i.e. ferruginous facies particularly rich in iron oxides and hydroxides. Then, we obviously risk eventual confusions with primary ferriferous mineralizations (BIF, hematitites….)


This refers to different observations upon concentrated goethite (and hematite) enrichments, exact definition of blocks and existence of goethitic string patterns in ironcrusts, usually accompanied by a thickening appearance of the ironcrust horizon.


Very iron-rich, quasi holo-hematitic and above all goethitic blocks are frequent everywhere in the tropical zone, in ironcrusts and residues of ironcrust dismantling and a role of witnesses for a detrital origin of ironcrust formation is thus ascribed to the latter.


In Togo the very numerous gossans of Pagala, oriented in meridian direction for the most part are observed in the dissection of the Adélé plateau by east-west retrogressive erosion. This eventually ironcrust-bearing plateau, displays in several loci blocks of gossan embedded in the true ironcrust or in the soft one, and prospections in this area may point out aureoles with Zn anomalies, without any corresponding indicatory outcrop.


In Burkina, the conglomeratic aspect of ironcrust including iron minerals-rich blocks is frequently observed on Birrimian formations with similar ferruginized levels in the rocks of the near neighbourhood (Perkoa, Kwademen, Rissian, Diana, Piéla, figure 10). Bauxitic and hologoethitic facies developed obviously before embedding ironcrust formation and are witnesses of paleo-weatherings. Their equivalence may be considered whatever the model used to explain the formation of ironcrust, that is autochthonous by integrating into the weathering zone of volcano-sedimentary silicated rocks, some residual blocks originating from previous weatherings, or allochthonous by iron embedding of detrital materials transported during an intermediary step. The paleo-weatherings of these sulfidic mineralizations only occur in the form of blocks detached from their weathering profile, and surely emptied of their matter facing the tremendous present - day but above all past climatic stresses. These are paleo-gossans whose roots were inevitably hydrolyzed and probably obliterated from landscape. Distinct steps can be discriminated in the gossan formation as follows: an ancient step-may be contemporaneous to bauxitisation- maintained by blocks (or elements) with very ferruginized facies of some “conglomeratic” ironcrusts - and a more recent more or less actual step preserving the morphologic features of mineralization but developing its own faciologic features at once.


The whole weathering profile of sulphidic ore, never observed in West Africa, is theoric.


Fig 10 

Scheme of the junction between alteration of sulfidic formations and ironcrust tropical weathering from silicated rocks.


The enrichment of goethite in the mass of the weathering cover and of ironcrusts through goethitic strings pattern – corresponding to that of sulfidic ones in the gold ore of Diouga and Gangaol – led me to propose the denomination of “crypto-gossan” in order to discriminate this facies from the usually more massive ones. The geochemical picture of these secondary goethite – rich strings is obviously well defined, with some diffusion into the surrounding ironcrust (table 14)


Collections Parisot and Blot. Major elements : %, others : ppm, Au : ppb


Table 14 

Comparison between the composition of ironcrust from Gangaol (BF) and that of goethitic strings included.


What are specific informations brought by blocks of paleo-gossans, when there is no apparent connection with any parental material ?


Their mineralogical composition is very simple: they are principally holo-goethitic, more or less hematitic, with high contents of manganese oxi-hydroxides. Their texture is inherited and neoformed at once, organized according to features of the parental mineralization preserved such as, boxworks, scoriaceous or micro-sceriaceous facies, inherited orientations, recrystallizations of goethite in the form either of microcrystals or of millimetric to centimetric combs. Chemical composition shows the extreme prevalence of iron followed by silica, whilst alumina content is very low unlike the iron-richest hardened facies, as those observed in the ironcrusts of the intermediary surface of geomorphologists. These goethitic bodies display frequently high contents of metals.


This is clear in the case of weatherings developed from nickeliferous ultrabasic rocks from Bonga (table 15) where the hyperferruginous blocks embedded in ironcrust are well individualized (Lavaud, 2002 ; Lavaud et al., 2004). However, weathering of ultrabasic rocks must be considered as noticeably particular, and ironcrust as probably generated by weathering of the surrounding gabbroid rock (amphibolite) of the ultra-basic body


Table 15 

Composition in percent and ppm and mineral assemblage in supergene ferruginous rocks from the Bonga massif, Burkina Faso (Collection Blot quoted in Lavaud).

Goethite (Goet), Hematite (Hem), Kaolinite (Kaol), Gibbsite (Gibbs)


Identifying the rooting of these facies is obviously a challenge, since it seems little probable that carbonated and sulphated horizons could be preserved during generalized strong weatherings. At the very most, a specific distribution of iron oxi-hydroxides can remain in the mass of the weathering cover.


c Consequences of mineralizations’weathering in the tropical zone of West Africa


Without being denied, gossans were considered as more theoretical and marginal than real in the whole West Africa. And so, the gossans pointed out in Togo by the author from a process of identification of the ferruginous outcrops were listed some years long by prospectors due to their specific characteristics, different from what is known elsewhere. In truth, there are no secondary minerals, the mineralogical suite is rather poor and atypical, and this extreme maturation seems to have not been described elsewhere. It is also why the study of the gossans by Togbé (1991) has been almost exclusively performed on outcrops, mineralizations being not yet identified. Such a problem arises now in Extreme North Burkina, where many ferruginous outcrops are gossans. In both cases, the early diagnosis, delicate as a matter of fact, is hardly taken in account for prospecting of mineral resources and never as direct mineralization index. But if accepted, it could however result in development of more targeted prospections. And consequently, the study of outcropping ferruginous rocks should get a place in prospection in tropical zones, what was never the priority for prospectors.