Geological Survey of

Bulletin 366

Glaciofluvial transport of elasts and heavy minerals from the Sokli carbonatite complex, Finnish

by Marjatta Perttunen and Heikki Vartiainen

Geologian tutkimuskeskus Espoo 1992 Geological Survey of Finland, Bulletin 366

GLACIOFLUVIAL TRANSPORT OF CLASTS AND HEA VY MINERALS FROM THE SOKLI CARBONATITE COMPLEX, FINNISH LAPLAND

by

MARJATT APERTTUNEN AND HEIKKI VARTIAINEN

with 10 figures and 6 tables

GEOLOGIAN TUTKIMUSKESKUS ESPOO 1992 Perilunen, Marjalla & Vartiainen, Heikki, 1992. Glaciolluvial transport of elasts and heavy minerals from the Sokli carbonatite complex, Finnish Lapland. Geologi­ cal Survey 01 Finland, Bulletin 366. 21 pages, 10 figures and 6 tables. The esker crossing the Sokli carbonatite complex in Finnish Lapland affords an ideal test site for the study of glaciolluvialtransport in the area of the Pleisto­ cene ice divide. The Devonian Sokli carbonatite complex is surrounded by Precam­ brian bedrock. Sam pIes were collected from 22 sites along the esker course over a distance of 16 km. The lithology of elasts and heavy minerals were studied using sam pIes from the surficial parts of the esker. The first appearance in the esker of elasts derived from the carbonatite com­ plex occurs within one kilometre of its proximal contact whereas heavy minerals appear in the esker 2 to 3 kilometres from the proximal comact. Peak abundances in the esker are in the distal portion of the source area. The transport distance in the esker for the majority of carbonatite complex elasts and carbonatite heavy miner­ als is only 4-5 kilometres. 1ndicator elasts and soft heavy minerals of surficial esker material are useful for mineral exploration in the areas of the last deglaciation during the Weichselian glaciation.

Key words (GeoRef Thesaurus, AG/): glacial transport, eskers, elasts, heavy minerals, gamet group, ilmenite, rutile, electron probe data, provenance, carbonatites, Sokli, Savukoski, Finland

Marja((a Per((unen Geological Survey 01 Finland. SF-02150 Espoo, Finland Heikki Vartiainen Kemira Gy, Lainaankatu 8, SF-96200 , Finland

ISBN 951-690-481-5 ISSN 0367-522X

Vammalan Kirjapaino üy 1992 CONTENTS

Introduction ...... 5 Geological setting and the source rocks of the Sokli esker ...... 6 Quatemary geological setting ...... 8 Methods ...... 9 Sampling ...... 9 Laboratory work ...... 9 Provenance of elasts ...... 10 Fenite ...... 10 Phosphorus ore ...... 12 Carbonatites ...... 12 Other rock types ...... 13 Heavy mineral contents and evaluation of mineral transport ...... 13 Electron microprobe analyses of gamet, ilmenite and rutile and discussion on their source ...... 14 Garnet ...... 14 Ilmenite ...... 17 Rutile ...... 18 Discussion ...... 18 Acknowledgements ...... 20 References ...... 21

INTRODUCTION

Mineral exploration, as weH as bedrock map­ out the field and laboratory works as weH as the ping, has been carried out by Rautaruukki Oy glacial geology, Heikki Vartiainen carried out the and Kemira Oy in the Sokli carbonatite complex, stone counts and microscopic mineral analyses. and its surroundings, since its discovery in 1967 ; Both authors are responsible for the results and most of the results of these investigations have conclusions of the study. been published (e .g. Vartiainen and WooHey The esker system from the last deglaciation 1976, Vartiainen and Paarma 1979 and Vartiai­ during the Weichselian glaciation trends from nen 1980). northwest to southeast and crosses the Caledo­ Abundant detailed petrological and mineralog­ nian (Devonian) Sokli carbonatite complex, ical studies permit closer examination of the which differs greatly in lithology from the lithology of the esker system. The fjeld work was Precambrian surroundings. Therefore the area is done in 1988-1989. Marjatta Perttunen was ideal for a study of glaciofluvial erosion, trans­ responsible for planning the study, for carrying port and deposition.

24 ' / / r...... _""1 1 I /- I / NORWAY I <", .' I I I I ~, RUSSIA

68 ' 68'

SWEDEN

Fig. 1. Location of the study area in northern Finland (black areal. ice divide zo ne (shaded areal during last deglaciation ca. 10,000 years B.P. (Ig­ -- - natius, Korpela and Kujansuu 1980) and preglaciall y weathered bedrock 100km R•OVANIEMI 30 ' zone (broken line) (Hirvas 1983) in Finnish Lapland. 6 Geological Survey of Finland, Bulletin 366

The relative transport distances of elasts and deglaciated areas of the Weichselian glaciation, heavy minerals in the Sokli esker were measured. especially in the regions of ice divide (Fig. 1), ap­ Both the influence of the proximal increase in plied to mineral exploration. In addition, the pos­ abundances of lithologies as weil as distal de­ sible incidence of kimberlites in the area was in­ crease were studied. The objective was to assess vestigated through the study of the heavy miner­ the use of glaciofluvial sediments, in the als.

GEOLOGICAL SETTING AND THE SOURCE ROCKS OF THE SOKLI ESKER

The geological setting of the Sokli esker is from previous work done on the Sokli carbona­ shown on the geology map (Fig. 2), compiled ti te complex (Vartiainen and Woolley 1976, Var-

~ GRANITE GNEISS ~ GRANITE ~ AMPHIBOLITE ~ ULTRAMAFITE ROCKS OF THE CARBONATITE COMPLEX ~ CARBONATITE AND METACARBONATITE L±J FENITE

(-:-:-:-:-:J PHOSPHORUS ORE ..r ...... SOKLI ESKER ITJ SAMPLE SITES

Fig. 2. Geological map of the Sokli carbonatite complex and surrounding bedrock with the sampling sites located on the Sokli esker. Table I. Average mineral contents of basement rocks, carbonatites and phosphorus ore (weathered crust), and heavy mineral fractions of the Sokli esker sampies.

Esker Sampies Basement Carbonatite Located over Distance from SE-contact of carb. massif massif basement carbon. 0.5-2.0 3.0-4.0 5.0-6.5 7.5 Sampie numbers 1-4,22 6-10 11-16 17-18 19-21 22 (in Fig. 2) 070 % ' % % % % % %

Minerals from the basement rocks Gamet + 28 17 11 3.5 12 30 Rutile + 4 3 3 I 3 2 Quartz, feldspar 75 + 2.5 4 2 0.5 Minerals from the carbonatite massif 0 Apatite 0.2 12.0 3.5 15 30 35 25 4 "0 Ö Francol.-goeth. 4 7 3 ()Q 3.0 5 ;S ' Pyrochlore 0.2 + + + + a Carbonate + 55 I 5 2.5 0.5 Vl=; <: Common minerals '<" Amphiboles 16 8 46 35 27 35 30 43 0..., 'TI Opaques 3 11 12 17 12 12 17 15 5' öl Zircon + 0.1 4 5 0.5 4 6 p-::l Micas 5 10 0:1 =-~ + observed 5' - not detected w 0-. 0-. 8 Geological Survey of Finland, Bulletin 366 tiainen 1980). The head of the Sokli esker lies ally variable amphibole and mica rich metasoma­ on Precambrian basement rocks, traverses south­ tites and metacarbonatites. Because of the strong east over the fenite aureole, the carbonatites and weathering, no elasts of these rocks have been their regolithic phosphorus ore, the fenite aure­ observed in the esker. The magnetic carbonatite ole and terminates on the basement again. The co re occupies 6 km2 in the central part of these esker on the Sokli carbonatite complex is eight rocks. These carbonate rich magmatic carbona­ kilometres long. tites have been more resistant to weathering than The Archean basement rocks, in the vicinity the former rocks, and they form a blocky of the esker on both si des of the carbonatite com­ weathered surface (regolith) which has yielded plex, comprise gneissose granites and granodi­ so me rock fragments. oritic gneissose granites, as weil as a variety of The regolithic phosphorus ore was formed by amphibolites. Serpentinized ultramafite bodies the complex weathering and recrystallisation near the esker are over one kilometre in length. processes. Accordingly it contains relict carbona­ Pegmatites of variable width are common in the tite minerals such as apatite, and recrystallized basement. The basement surface is unweathered. minerals such as the phosphorus mineral fran­ The generalized average mineral composition of colite (Table 1). About 10 % of the volume of the basement in the vicinity of Sokli is estimated the phosphorus ore is composed of a hard ore in Table 1. type capable of yielding rock fragments. The 363 Ma carbonatite complex, consisting The average mineral composition of the of the fenite aureole and the carbonatite massif, weathered crust of the carbonatite massif is esti­ has a strong lithological contrast relative to the mated in Table 1. The mineralogical difference basement. The fenites are easily identifiable and between the rocks of the basement and the car­ originated through Na-metasomatism of the bonatite massif shown in Table 1 reflects the dis­ basement rocks. The fenite is composed prin­ tinct lithological dissimilarity between these cipally of albite, sodium amphiboles and pyrox­ rocks. enes with the proportion of relict primary miner­ Consequently, the Sokli area, at the site of the als diminishing with increasing degrees of feniti­ Sokli esker, offers two types of source rock. Be­ zation. Because the fenite aureole is extensively cause the basement and carbonatite complex tectonized, but not weathered, it readily yielded lithologies differ so strikingly from each other, rock fragments during glaciation. both petrographically and mineralogically, the The upper part of the carbonatite massif is ex­ Sokli esker is eminently suitable for the study and tensively weathered, particularly in a 1-2 evaluation of the transport of elasts and heavy kilometres wide zone in contact with the fenite mineral grains. aureole. This zone is composed of composition-

QUATERNARY GEOLOGICAL SETTING

The Quaternary deposits in the area were (Fig. 2) . The Sokli esker is part of an approxi­ mapped in 1966-1970 and described by Hirvas mately 35 km long esker system. and Kujansuu (1972) at a scale of 1 : 400 000. The The deglaciation occurred in the area ca. esker system within the study area was remapped 10,000 years B.P. (lgnatius, Korpela and Kujan­ at the scale of 1 : 20000 by Marjatta Perttunen suu 1980) . According to till fabric analyses the Geological Survey of Finland, Bulletin 366 9 upper till shows a northwest-southeast-trending Generally , the coarse fractions dominate in the ice flow direction, which correlates with stage 11 sampled surficial portions. in Lapland (Hirvas 1977). The esker system The Sokli esker is subglacial in origin. In the trends from northwest to southeast and thus is supra-aquatic area it was, in places, formed by parallel to the direction of the last ice flow . The streams under high hydraulic press ure (Ilvonen earliest ice flow recorded in the area is from the 1973). The esker occurs in both supra-aquatic and west or southwest (Ilvonen 1973). subaquatic setting. In subaquatic areas the esker The esker system consists of sinous ridges with ridges are covered by glaciolacustrine sediments beads 200 m to 2 km in length. Proximal ridges deposited by a previously unrecorded glaciallake. mainly displaya sharp crest and fairly steep sides. The glaciolacustrine sediments consist of ho­ The distal parts consist of a broad esker complex mogeneous silt overlying silt with rhythmites. partly with kames. The heights of the ridges range In the surroundings of the esker system the from 5 to 25 m and the widths from 40 to 400 m. dominant Quaternary deposit is ground moraine, According to Ilvonen (1973) material in the which only occasionally attains a thickness of Sokli esker shows great horizontal and vertical 15 m. The total thickness of Quaternary sedi­ variation. In the proximal part of the esker ments, based on drilling, exceeds 38 m. Ilvonen ridges, layers of cobbles, pebbles and gravel al­ (1973) described the Eemian Interglacial diatom ternate with sand. In the distal parts of the esker deposit found at Sokli . ridges sand is dominant. The study area is situated in the ice divide re­ There are two gravel pits in the Sokli esker. gion, where the weathered crust is commonly The one situated over the Sokli carbonatite mas­ preserved. The preglacial weathering profile over sif shows sandy material with cross-bedding and the Sokli carbonatite complex has been found to layers with concentrates of garnets. The other pit be on average 10-20 m thick. In the fenite aure­ at the Ainijärvi border station has coarser mate­ ole only the carbonatite dike rocks are weathered, rial such as cobbles and pebbles with gravel and in pI aces down to 20 m. sand. Garnet concentrations occur as weil.

METHons

Sampling

SampIes were collected from 22 sites along the At each pit 100 elasts of the 2-10 cm fraction esker course over a distance of 16 km . The pits were collected for stone counts. Each sampIe, were dug by shovel to one metre depth. The sur­ totaling 10 litres of < 4 mm fraction material was ficial parts of two existing gravel pits were used taken for heavy mineral analysis. These sampIes for sampling. were panned in the field .

Laboratory work

Panned concentrates were sieved into subsam- mm, 0.125-0.250 mm, 0.07-0.125 mm and pies: 2-4 mm, 1-2 mm, 0.5-1 mm, 0.250-0.5 < 0.07 mm. Grain-size analyses were made on 10 Geological Survey of Finland, Bulletin 366

100 % 90

70

~ <3'" - ~ 50 u ~ 0 '";>. 30

10

0 _002 20 60mm Grain size mm)

Fig. 3. Graph showing the grain-size analyses of 22 sampIes from the Sokli esker. The shaded area indicates the grain-size limits for the sampIes . The black line indicates the median curve.

sampIes « 60 mm) representing the original Semiquantitative mineral contents of the heavy material from the sampling sites of the 22 sam­ mineral fractions were determined optically from pIes studied. The analyses show that the medi­ thin sections under the polarizing microscope us­ um grain-size (M;z:) was sand for 19 of the sam­ ing grain density charts and the chemical com­ pIes and gravel for 3 of the sampIes (Fig. 3) . All position of the sampIes. The accuracy of this sampIes, with the exception of number 6, con­ method is estimated to be about ± 5-10 %, tain cobbles and pebbles. which is acceptable for the purposes of this study. For the study in question the 0.07-0.125 mm For each of the sampIes, mineral abundances of fraction was treated in the laboratory with the heavy mineral fractions, from wh ich the mag­ bromoform (SO 2.810-2.830) to separate the netic fraction was separated, were calculated for heavy minerals. The magnetic grains were then the 0.07-0.125 mm fraction. separated by hand magnet and the sam pIes were Identification of the 2-10 cm elasts was made prepared as polished thin sections. in the laboratory.

PROVENANCE OF CLASTS

Fenite

The mechanical properties of fenite, wh ich disrupted, fragmentary and jointed nature, as a forms an aureole surrounding the carbonatite result of which it has contributed more material massif (Figure 2), are comparable to those of to the Sokli esker than have the other lithologies granite. The fenite has, however, a tectonically of the complex (Figure 4). Fenite elasts first ap- Geological Survey of Finland. Bulletin 366 11

20 PHOSPHORUS ORE ClASTS % 0 o o 0 o 10 o 0 o 0 'I, o 0 0 0 o 0 0 0 o 0 0 0 0 0000000 000 000 000

40 CARBONATITE ClASTS 'I,

20 U1 'I, (J; u oC u C ::J .D o 60 FENITE ClASTS 'I, (J; + > + + + + + T + + -t- + + T + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + -t- + + + + + + + + 10 + + + + + + + + + + + + 'I, + + + + + + + + + + + + +

Sampie slle numbers 2 3 4 5 6 7 8 9 10 11,12 ,13,14,15,16 17 18 19,20 21 22

Bedrack profile

B d k 1"1 loOloo ca rbonat i te e ro c pro I e : ~ massil I carbonatite ) compl ex 1+ + +1 lenite

basement

Fig. 4. Relative abundances of elasts derived from the carbonatite complex in the Sokli esker. pear in the esker very elose to the proximal con­ side of the complex (typically aegirine rock and tact, with abundances rapidly increasing and at­ light green fenites) and the outer (distal) contact taining maximum values of around 60 070 within of the aureole on the southeastern side 5-6 km from the distal contact of fenite. Thus (amphibole-rich dark green fenites). This indi­ the fenites elearly indicate the location of the car­ cates that fenite elasts still comprise 30-40 % bonatite complex. Figure 5 shows the dispersal of elast populations at a distance of 3-4 km of fenite elasts measured from both the inner (dis­ from their source. tal) contact of the aureole on the northwestern 12 Geological Survey of Finland, Bulletin 366

Phosphorus ore

Although the bulk of the phosphorus ore is traverses both rock types over equal distances; soft and weathered, about 10 070 by volume con­ this difference is therefore attributed to differ­ sists of more indurated material that is highly ences in mechanical properties, ineluding fractur­ resistant to mechanical crushing. Thus, elasts of ing and jointing. The occurrence of a distinct phosphorus ore are also present within the Sokli peak in phosphorus ore elast abundances within esker, showing the same kind of dispersal be­ the esker southeast of the carbonatite massif may havior as the fenite elasts. However, phospho­ even indicate the presence of a previously un­ rus ore elasts are only about half as abundant as detected ore body (Figure 4). those of fenite, in spite of the fact that the esker

Carbonatites

The surficial parts of the carbonatite massif are elasts in the esker, as is readily apparent from pervasively weathered, with only sporadic occur­ Figure 4. Although weaker, variations in distri­ rences of fresh carbonatite having been exposed bution and abundance of elasts nevertheless fol­ from beneath Quaternary overburden during the low the same trends as for the fenites and phos­ excavation of a total of 30 km of exploration phorus ore. trenches. This explains the dearth of carbonatite

% 90

80 Cf) f- Cf) -< 70 TRANSPORT OF FENITE ClASTS -l Ü X Iro m northwest ern dis t a l c o ntac t 01 l e nite aure ole w 60 f- • Irom southeas t ern dis t a l cont ac t 0 1 l e nite aureole Z W 50 • LL LL 0 - - 40 • w CD x -< • -- .... f- 30 • Z W • Ü x 0: 20 w Q. • "- 10 " \ • 1.0 2 3 4 5 6 km DISTANCE FROM DISTAL CONTACTS OF FENITE AUREOLE

Fig. 5. Transport distances for fenite elasts derived from northwestern and southeastern distal contacts of aureole. Geological Survey of Finland, Bulletin 366 13

Other rock types

Mica-rich mafic lithologies (metasomatites, tite massif, but are so weathered that they have metaphosphorites and lamprophyres) are present not contributed any coarse cIastic material to the throughout roughly two thirds of the carbona- esker.

HEA VY MINERAL CONTENTS AND EV ALUA nON OF MINERAL TRANSPORT

The average mineral abundances of the non­ 0.125 mm fraction from esker sampies are given magnetic heavy mineral fractions in the 0.07- in Table 1. Minerals are cIassified into two

5 PYROCHLORE % FRANCOLITE - GOETHITE 10 '!.

0 0

CARBONATES 10 ~ c!- °/.

U)

Cl

10 °/.

Sampie site numbers 1 2 3t. 5 6 789 10,11.12.13.14.15.16 17 18 19.20 21 22

Bedrock prof ile

roöloo carbonatite Bedrock profi le ' L.c!....J massif l carbonatite \ complex 1+ ++1 fenite basement

Fig. 6. Variation in the concentration of carbonatite massif heavy minerals (0.125-0.07 mm fraction) in the Sokli esker. The carbonatite heavy minerals are concentrated in this fraction. A minor portion of apatite (m .a. 4070) is derived from basement. 14 Geological Survey of Finland, Bulletin 366 groups; those from the basement and those from fied into four groups in Table 1 in relation to dis­ the carbonatite massif. For reference, approxi­ tance from the contact of the carbonatite mas­ mate modal mineral compositions for basement sif. The two largest groups show a uniform pat­ rocks and the carbonatite massif are also given tern of heavy minerals. The apatite content of in Table 1. Figure 6 shows the variation in the 30-35 0J0, francolite of 5-7 0J0 and carbonates concentrates of carbonatite heavy minerals in the of 2.5-5 0J0. The first two are higher than the Sokli esker. corresponding amounts within the carbonatite The sampies taken from the esker over base­ massif proper. This indicates that signifant trans­ ment represent background values of non-rn ag­ port of minerals has taken place over a distance netic heavy minerals with no contribution from of at least 4 kilometres. the carbonatite massif. They contain on average Sampies 5.0-6.5 kilometres from the distal 3.5 % apatite, which corresponds to an enrich­ contact of the carbonatite massif still show a dis­ ment factor of more than 10 times with respect tinct effect of mineral transport, but beyond 7.5 to the source rocks. Other minerals present in the kilometres of the contact, all minerals from the carbonatite massif are absent from the back­ carbonatite massif are absent and apatite is at ground sampies. background level. Basement-derived minerals are present in all Non-magnetic opaques are unformly concen­ the esker sampies. Garnet, the main basement trated between 12 and 17 0J0 in all esker sampies. mineral is naturally less abundant over the car­ Although the quantity of opaque minerals in the bonatite massif and on the distal side of the mas­ carbonatite massif is much higher than in the sif relative to the basement. Rutile and light basement (Table 1), the non-magnetic opaques minerals (quartz, feldspars) are fairly evenly dis­ are most probably derived from the basement be­ persed in all the sampies. cause magnetite is the dominant opaque mineral The distribution of the carbonatite massif known from the carbonatite massif. Also, the minerals in the esker sam pies allows an evalua­ even distribution of opaques reflects their base­ tion of the transport distance of heavy minerals ment origin. in the Sokli esker. The distal sampies are c1assi-

ELECTRON MICROPROBE ANALYSES OF GARNET, ILMENITE AND RUTILE AND DISCUSSION ON THEIR SOURCE

Several minerals have particular ranges of the potential of the source rocks so me electron composition which may be indicative of the pres­ microprobe work has been done on garnet, il­ ence of kimberlites or even diamondiferous kim­ menite, rutile and zirconium-rich rutile. berlites. In order to check this and to evaluate

Garnet

A total of 33 garnet grains were analyzed from grouped as Mg-almandine, almandine and Fe­ three esker sampies (1,9 and 16, Figure 2). With grossular (Table 2). Mg-almandine is prevalent slight compositional variations the garnets can be (64-82 0J0), but only one grain of Fe-grossular Geological Survey of Finland, Bulletin 366 15

Table 2. Electron microprobe analyses of ga mets from three Sokli esker sampies.

Sampie 9 16 Gamet Mg- Alman- Mg- Alman- Mg- Alman- Fe- group almandine dine almandine dine almandine dine grossular 070 % % % % % %

Si02 38 .80 37.20 38 .10 37.00 39 .30 38.20 37.80 Ti02 0.08 0.06 0.06 0.03 0.05 0.09 0.08 AlPl 22 .20 20.60 21.90 20.70 22 .20 21.00 20.60 'FeO 25 .30 27 .20 25 .10 29.20 24 .90 26.60 23.90 MnO 0.38 1.38 0.30 0.95 0.41 0.87 0.51 MgO 11.30 4.70 11.20 5.60 11.40 1.50 1.50 CaO 1.60 6. 11 1.20 5.40 1.50 6.90 15.40 Cr2O) 0.05 0.02 0.05 0.04 0.05 0.01 0.05

• Total iron calculated as FeO.

occurs in population. The most probable source rocks for the gar­ The distinct and consistent degree of correla­ nets in the Sokli esker occur within the granulite tion between Mg-almandine and almandine in all area of Inari, which extends to within about 40 the esker sam pIes reflects a common source rock km of Sokli. According to published gamet ana­ containing two compositional gamet types. On lyses the gamets of the Inari area can be defined the other hand an increasing proportion of Mg­ as those from the granulite complex and those almandine can be observed over a distance of ten from the surrounding rocks. The MgO-FeO-CaO kilometres (SampIes 1-16, Figure 2), suggesting diagram (Figure 7) illustrates dose affinities be­ Mg-rich gamets are more resistant during tween the Mg-almandines and almandines of the glaciofluvial transport. Sokli esker and the Inari granulite area. If the

Table 3. The similarities between minor components of comparative garnets sug­ gests a dose relationship between the Mg-almandines and almandines of the Sokli esker and the lnari granulite area.

Mg-almandines Almandines Sokli Granulite Sokli Rock zones esker complex esker around granulites % % % %

Ti02 0.06 0.05 0.04 0.05 Cr2O) 0.05 0.05 0.03 0.01 MnO 0.36 0.46 1.17 2.05 16 Geological Survey of Finland, Bulletin 366

MgO

Fig. 7. MgO-FeO-CaO diagram for garnets. Stippled fields for garnets from kimberlites after Dawson and Stephens (1975): Field 11 high-titanium pyrope from kimberlites only. Field V magnesium almandine from kimber­ lites, edogites and diamond indusions. Open square denotes Mg-almandines (23) and open triangle almandines (8) from the Sokli esker. Filled sq uare denotes garnets (17) from the Inari granulite complex. The filled triangle denotes garnets (12) fr"m the rock zones surrounding the Inari granulite complex. The number of analyses is FeO CaO given in parentheses.

CARBONATITES

Fig. 8. Compositional fields of kim­ berlite and carbonatite ilmenites in the system MgTiOj-FeTiOj-MnTiOj ac­ cording after Mitchell (1986). Numbers I, 2 and 3 refer (0 ilmenite groups from the Sokli esker sampies. Geological Survey of Finland, Bulletin 366 17 interpretation that Mg-rich garnets are more composite fields of kimberlite garnets have been resistant du ring transport is correct then it can compiled. Only two fields have been separated; be hypothesized that Mg-rich garnet variants group 11 High-Titanium pyrope, occurring only have been enriched in the Sokli esker compared in kimberiites, and group V Mg-almandine, which to the source rocks (see Figure 9). The similari­ is found in kimberlites, ecIogites and as incIusions ties between the minor components of compara­ in diamonds. The magnesian almandines of the tive garnets also suggests a cIose relationship (Ta­ Sokli esker sampies fall at the margins of the Mg­ ble 3). almandine field and almandines outside of it. The Dawson and Stephens (1975) presented a MgO­ possibility of a kimberiitic origin for the Sokli FeO-CaO diagram in which they statistically esker Mg-almandines is thus very small. The pres­ separated garnets from kimberlites and asso­ ence of the granulite complex Mg-almandines in ciated xenoIiths into 12 groups. In Figure 7 the field V supports this concIusion.

Ilmenite

Thirteen ilmenite grains have been analyzed exsolutions in magnetite (Heinänen and Vartiai­ from the same esker sampies as the garnets. The nen 1981). ilmenites can be separated into three groups as shown in the Table 4. Plotting of the average il­ Table 4. Average compositions of 13 ilmenite grains ana­ menite compositions on the diagram of the sys­ Iyzed by electron microprobe. tem MgTi03-FeTi03-MnTi03 (Figure 8) reveals how groups 1 and 2 (12 ilmenite grains) are 10- Group 2 cated in the field of Fe-rich carbonatite ilmenites Number of grains 9 3 and group 3, those rich in Mn, outside of it. It 0J0 "'0 "'0 is cIearly evident that SokIi ilmenites are composi­ Ti02 49.40 49.40 49.60 tionally distinct from the ilmenites of kimberlites. AIP3 0.03 0.01 0.01 er203 0.05 0.02 0.00 The results in Table 4 and Figure 8 suggest that "FeO 46.10 44.20 35.0 the main source of the SokIi esker ilmenites is the MnO 1.50 4.40 11.20 carbonatite massif, although ilmenite only occurs MgO 0.19 0.65 2.40 in the SokIi carbonatite massif as variably sized " Total iron calculated as FeO.

Table 5. Electron microprobe analyses of rutile from Sokli esker sam pies N normal rutile, Zr = zirconium rutile, N / Zr = number of analyses for N and Zr.

Esker N / Zr Ti02 Zr02 Nb20 S FeO sampie N Zr N Zr N Zr N Zr numbers "'0 "'0 "'0 "'0 "'0 "'0 "'0 "'0 I 1/ 4 95.6 95.7 0.09 0.51 0.85 0.23 0.49 0.06 3 2/ 2 95.6 95.3 0.66 0.39 0.35 0.20 0.05 6 1/ 4 97.2 95.7 0.05 0.40 0.36 0.10 0.10 9 0/ 5 96.1 0.24 0.17 0.07 10 4/ 0 97.2 0.05 0.08 0.36 12 4/ 1 96.3 96.0 0.07 0.49 0.24 0.24 0.08 0.07 14 0/ 5 96.7 0.69 0.37 0.04 20 0/ 4 96.6 0.37 0.21 0.05 18 Geological Survey of Finland, Bulletin 366

Rutile

Rutile is present in all esker sampIes, amount­ varieties are present (Table 5). ing to 1-4 % of heavy mineral fractions. A to­ The compositional averages of the rutile are tal of 37 rutile grains have been analyzed from shown in Table 6. All the rutile comes from the eight esker sampIes. Normal and zirconium-rich basement.

Table 6. The compositional averages of the rutile.

Number of Ti02 Zr02 NbPs FeO analyses 0/0 % % %

Normal rutile 12 96.5 0.05 0.24 0.23 Zr-rutile 25 96 . 1 0.40 0.27 0.05

DISCUSSION

It is evident from this study that the mechani­ imal contact. In general, however, elasts first ap­ cal properties of given source rocks control the pear in the Sokli esker eloser to proximal con­ abundances and dispers al of elasts within eskers, tacts than is the case with the eskers in southern as Virkkala (1958) coneluded from his investiga­ Finland, where the first appearance of indicator tion of the Hämeenlinna esker in southern Finland. elasts has been found to be 5-8 km (Hellaakoski Figure 2 indicates how the Sokli esker has incor­ 1930, Virkkala 1958, Perttunen 1989). porated material as it traversed approximately Heavy mineral fractions from the Sokli esker equivalent intervals underlain respectively by show the same trends as elasts (Figures 6 and 10). fenite, phosphorus ore and carbonatites. Fig­ Mineral distributions are however more sensitive ure 4 shows that the more resistant fenite elasts to variations in source rocks and have gene rally elearly predominate over the softer phosphorus travelled further . Carbonate remaining in the ore and carbonatite elasts. This difference is fur­ heavy mineral fraction also shows the same dis­ ther accentuated by the tendency for fenites to tribution pattern as apatite (Figure 6) . show extensive tectonic fracturing, and also by The first appearance of heavy minerals derived the more intense weathering of carbonatites, with from the carbonatite massif occurs at a distance only isolated occurrences of fresh material avail­ of 1.5 km from the proximal contact, where they able. The behavior of phosphorus ore elasts is comprise 4 % of the population. Their abun­ intermediate between the above two, in that dances remain high for a greater distance than transport distances are comparable to those of those of elasts (Figures 4 and 6), that is, over dis­ fenite, while abundances are distinctly lower. For tances of 4-5 km. example whereas fenite elast abundances are be­ Apatite predominates in the heavy mineral tween 35-50 % at a distance of 1 km from the fraction because it is present in all lithologies proximal contact of the fenite aureole and 30- throughoui the carbonatite complex, with an ad­ 40 070 at a distance of 3-4 km, phosphorus ore ditional component, amounting to 3- 4 0J0 of the elasts only comprise 5-15 % of the total popu­ total, being derived from the surrounding base­ lation at a distance of 3-4 km from the prox- ment. Due to the mass effect it is transported fur- Geological Survey of Finland, Bulletin 366 19 thest. The sensitivity of the heavy mineral frac­ port. Comparable results have been obtained for tion is indicated by the distribution of francolite­ eskers in southeastern Finland (Perttunen 1989). goethite, wh ich is characteristic of the Loitso In conelusion it can be stated that overall trans­ phosphorus ore body. pyrochlore is present port distances are short, the transport for the throughout the esker lying on the carbonatite majority of carbonatite complex elasts and car­ massif, being on average 0.5 0,10, but because of bonatite massif heavy minerals being only so me its resistance to abrasion, it is transported rela­ 4-5 km (Figs. 9 and 10). The transport mode tively further than other minerals. lt is signifi­ of elasts of the Sokli esker, from the distal con­ cant that garnet is abundant in the Sokli esker tact of the carbonatite complex, corresponds to at distances of 30 to 40 kilo met res from its the transport of esker materials in southern Fin­ source, testifying to its durability during trans- land referred to above, even though they are far

100 %

~ 0 0-

tJ) (lJ u c 0 "0 c 50 :J 'I , n 0

(lJ .....> 0 Qi 0::

10 'I ,

~~~g~r ~lte 2 34 5 6 78 9 10 11.12.13.14,15,16 17 18 19.20 21 22

Bedrock profile 0

~ carbonatite eomplex elasts L>I basement elasts

B ' 1 loOloo carbonat i te e d roc k pro 1I e ' ~ massil l carbonotite \ complex 1+++1 lenite I(J basement

Fig. 9. Variation in the concentration of carbonatite complex elasts (Phosphorus ore, Carbonatites and Fenites) and base­ ment elasts in Sokli esker sam pies. 20 Geological Survey of Finland, Bulletin 366

100 ~------, 0/0

lf) OJ u c a u 50 c :J 0/0 D a OJ > o 0 ,

10 0/0

Sam pie site numbers 2 34 5 6 7 8 9 10 11,12,13,14)5.16 17 18 19,20 21 22

Bed roc k profil e o 16km ~~==~~~==~~==~~~==~~====~==~~======~====

~ heavy minerals in carbonatite massif

lOOloo carbonatite Bedrock profile ' ~ massif I carbonatite ) compl ex 1+ ++1fen ite basement

Fig. IO.Variation in the concentration of heavy minerals derived from the carbonatite massif. A background value, of around 4 070, due to apatite in the Precambrian basement, has been subtracted from the total apatite content of the sampies .

away from the ice divide zone. The first appear­ Evenson 1987, Lilliesköld 1990). ance of the elasts in the Sokli esker is eloser to Indicator elasts and soft heavy minerals with­ the proximal contact of the source rocks than in in surficial esker materials are useful for miner­ eskers in southern Finland or other count ries al exploration in areas of the last deglaciation of (Shilts and McDonald 1975, Bolduc, Klassen and Weichselian glaciation.

ACKNOWLEDGEMENTS

The authors are very gratefu! to Professor materials from the Sokli esker in the context of Heikki Paarma for suggesting examination of their application to mineral exploration. We are Geological Survey of Finland, Bulletin 366 21 also grateful to Professor Reijo Salminen and Dr. vey of Finland, who carried out the electron Bill Coker for reading the manuscript critically microprobe analyses, and to Arto Vitikka, who and for constructive comments. Our thanks go panned the sampies. to Bo Johanson, M.Sc., of the Geological Sur-

REFERENCES

Bolduc, A.M., Klassen, R.A., & Evenson, E.B., 1987. Cob­ glaciofluvial material.ln: R.Kujansuu and M. Saarnisto ble lithologies in eskers of central Labrador. Geological (Editors), Glacial indicator tracing. A.A. Balkema, Rot­ Survey of Canada, Paper 87- IA, 43 -51. terdam, 151-164. Dawson, J.B. & Stephens, W.E., 1975. Statistical analysis MilcheII, R. H., 1986. Kimberlites: mineralogy, geochemis­ of garnets from kimberlites and associated xenoliths. J. try, and petrology. Plenum Press. New York, 442 p. Geol. 83, 589-607. Perilunen, M., 1989. Transportation of garnets in glacioflu­ Heinänen, K. & Vartiainen, H., 1981. Magnetite in Sokli car­ vial deposits in southeastern Finland. In: R.N.W. DiLa­ bonatite massif and in Tulppio olivinite. Bull . Geol. Soc. bio and W.B. Coker (Editors), Drift Prospecting. Geo­ Finland, 53, 83-90. logical Survey of Canada, Paper, 89-20, 13-20. Hellaakoski, A., 1930. On the transportation of materials in Shilts, W.W. & McDonald, B.C., 1975. Dispersal of elasts the esker of Laitila. Fennia, 52, 1-42. and trace elements in the Windsor esker, southern Que­ Hirvas, H., 1977. Glacial transport in Finnish Lapland. bec. Geological Survey of Canada, Paper, 75-1 Part A, Prospecting in areas of glaciated terrain 1977. The Insti­ 495 - 499. tution of Mining and Metallurgy, 128- 137 . Vartiainen, H., 1980. The petrography, mineralogy and Hirvas, H., 1983. Correlation problems of imerglacial deposits petrochemistry of the Sokli carbonatite massive, north­ in Finnish Lapland. In : A. Billard, O. Conchon & F.W. ern Finland. Geological Survey of Finland, Bull. , 313, Shotlon (Editors), Quaternary Glaciations in the North­ 126 p. ern Hemisphere, Project 73 / 1/ 24, Paris, 129-139. Vartiainen, H. & Woolley, A.R., 1976. The petrography, Hirvas, H. & Kujansuu, R. , 1972. Quaternary Deposits, Sheet mineralogy and chemistry of the fenites of the Sokli car­ 47, Talkkunapää. General Geological Map of Finland, bonatite intrusion, Finland. Geological Survey of Finland, I : 400 000. Geological Survey of Finland. Bull ., 280, 87 p. Ignatius, H., Korpela, K. & Kujansuu, R. 1980. The deglaci­ Vartiainen, H. & Paarma, H., 1979. Geological Characteris­ ation of Finland after 10,000 B.P. Boreas, 9, 217-228. ti es of the Sokli Carbonatite Complex, Finland. Econ. llvonen, E., 1973. Eem-imerglasiaalinen kerrostuma Savu­ Geol., 74, 1296-1306. kosken Soklilla, Pohjois-Suomessa, orgaanisten kerros­ Virkkala, K., 1958. Stone coums in the esker of Hämeenlin• tumien ja glasiaaligeologisen tutkimuksen valossa. Un­ na, southern Finland. Bull. Comm. Geol. Finlande, 180, published Licentiate-thesis, 144 p . 87-103. Lilliesköld, M., 1990. Lithology and transport distance of