Research article

Geochemical exploration for tantalum in coltan-rich pegmatites at
Bewadze-Mankoadze area of the Kibi-Winneba Belt, southern
Ghana: Constraints from exploratory data analysis

Emmanuel Daanoba Sunkari a,b,*, Joshua Nkansah a, Salaam Jansbaka Adams c,d

a Department of Geological Engineering, Faculty of Geosciences and Environmental Studies, University of Mines and Technology, P.O. Box 237,
Tarkwa, Ghana
b Department of Chemical Sciences, Faculty of Science, University of Johannesburg, P.O. Box 524, Auckland Park 2006, Johannesburg, South Africa
c School of Natural and Environmental Sciences, University of Environment and Sustainable Development, PMB –Somanya, Eastern Region, Ghana
d School of Nuclear and Allied Sciences, College of Basic and Applied Sciences, University of Ghana, P.O. Box LG 80, Legon – Accra, Ghana

A R T I C L E I N F O

Keywords:
Tantalum
Pegmatites
Multivariate statistical analysis
Geochemical anomalies
Pathfinder elements

A B S T R A C T

The Bewadze-Mankoadze pegmatites in the Kibi-Winneba Belt of Ghana host several columbite
group minerals (WGM) and wodginite group minerals (WGM) as well as other rare and radio-
active elements such as uraninite and cesium. In this study, petrographic studies of rock samples
from pegmatite outcrops and statistical analysis of the major and minor elements were conducted
to identify the pathfinder elements of a new tantalum deposit in the area. Ten samples were
obtained from each town for whole-rock geochemistry and thin sections were prepared from some
of the samples taken for petrographic analysis. The petrographic analyses showed the presence of
quartz, K-feldspars, plagioclase, muscovite, spodumene, albite, tourmaline, columbite group
minerals, and montebrasite, which indicate that the studied samples are granitic pegmatites. The
geochemical data of the 10 samples obtained from each town showed high concentrations of Cr,
Cs, Rb, Sm, and Ta. The Ta concentrations ranged from 10.5 to 773 ppm with an average value of
260 ppm. Q-Q plots showed outliers and variations from the dataset’s normal distribution, which
were fixed by centred log-ratio transformation and demonstrated to be normal by the
Kolmogorov-Smirnov test and Shapiro-Wilk test for normality. Spearman correlation revealed
that Sc, Ga, Nb, and Cs showed a moderate to strong correlation with Ta. Factor analysis indicated
elemental association of Ta with Cs, Zn, Nb, MgO, Sc, Ga, and V. Three (3) multi-element re-
lationships were discovered by hierarchical cluster analysis: (1) As, La, Hf, CaO, U, Co, Pb, Ce, Ba
and Na2O, (2) V, Nb, Ta, Ga, Sc, Cr, Cu, Nb and MgO, and (3) Ni, SiO2, Cs, Zn, Sm, Rb, K2O, Y, Th,
Al2O3 and Fe2O3. Hence, a comparison of the results of the multivariate statistics established Sc,
Ga, Nb, and Cs as the pathfinders of tantalum in the Bewadze-Mankoadze area. Geochemical
anomalies involving these elements can be observed in the south-western portion of the study
area, according to single and multi-element halo mapping. It is therefore recommended that
exploration activities for tantalum mineralization should focus on the south-western part of the
study area, where the anomalies of the pathfinder elements are located.

* Corresponding author. Department of Geological Engineering, Faculty of Geosciences and Environmental Studies, University of Mines and
Technology, P.O. Box 237, Tarkwa, Ghana.

E-mail address: edsunkari@umat.edu.gh (E.D. Sunkari).

Contents lists available at ScienceDirect

Heliyon

journal homepage: www.cell.com/heliyon

https://doi.org/10.1016/j.heliyon.2024.e38176
Received 3 June 2024; Received in revised form 18 September 2024; Accepted 19 September 2024

Heliyon 10 (2024) e38176 

Available online 21 September 2024 
2405-8440/© 2024 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license 
( http://creativecommons.org/licenses/by-nc-nd/4.0/ ). 

mailto:edsunkari@umat.edu.gh
www.sciencedirect.com/science/journal/24058440
https://www.cell.com/heliyon
https://doi.org/10.1016/j.heliyon.2024.e38176
https://doi.org/10.1016/j.heliyon.2024.e38176
https://doi.org/10.1016/j.heliyon.2024.e38176
http://crossmark.crossref.org/dialog/?doi=10.1016/j.heliyon.2024.e38176&domain=pdf
http://creativecommons.org/licenses/by-nc-nd/4.0/


1. Introduction

Tantalum (Ta) and other critical elements such as Be, Cs, Li, Nb, Sn, U and Th drive their economic sources from pegmatites. The
study of pegmatites has therefore attracted a lot of attention in recent years because these elements they contain are critical towards
clean energy transition [1–7]. Pegmatites are texturally and compositionally complex igneous rocks exhibiting skeletal, radial, and
graphic crystal intergrowth patterns, as well as strong anisotropy of crystal orientations from the edges inward, mineralogical
zonation, and coarse but variable crystal size. However, most pegmatite deposits are made of simple mineralogy such as quartz,
feldspar and mica, and have a silicic composition identical to granite [8].

Pegmatites containing rare elements can be grouped into three types which are: Lithium–Cesium–Tantalum (LCT) pegmatites,
which have high concentrations of Li, Cs, Ta, B, Be, P, F, Ga, Mn, Sn, Nb, Rb, and Hf [9–11]. The Winneba-Mankoadze pegmatites in
Ghana [1,3,5–7] as well as the Greenbushes, Wodgina, and Pilgangoora pegmatites in Western Australia, the Tin Mountain pegmatite
in the United States, Altai Number 3 pegmatite in China, Tanco pegmatite in Canada, Kenticha pegmatite district in Ethiopia, and
Bikita pegmatite in Zimbabwe are all examples of major LCT pegmatite deposits [12,13]. The second kind is Niobium-Yttrium-Fluorine
(NYF) pegmatites, which exhibit enrichment in B, Be, Sn, Nb> Ta, Y, Ti, REEs, Th, Zr, Sc, U, and F but depleted in Rb, Cs, and Li [9–11].
As summarised by Ref. [9], the Stockholm granite, Ytterby pegmatite group, Grotingen granite, Abborselet, and other related peg-
matites in Sweden; the South Platte granite and pegmatite system in Colorado; and the Lac du Bonnet biotite granite and Shatford Lake
pegmatite group in Canada, are good examples of NYF pegmatite deposits. It is postulated that contamination of NYF pegmatites
during the magmatic or post-magmatic stage is what formed the third type, known as mixed or "hybrid" rare-element pegmatites,
which have blended rare-element signatures [9–11,14]. For instance, they might form when freshly crystallised NYF pegmatites are
re-melted by metasomatic fluids that are enriched in Li, B, Ca, and Mg. Mixed pegmatites can be found, for instance, in Kimito in
Finland, the O’Grady batholith in Canada [14], and the Tordal area of Norway [2].

[6,7] concluded that the pegmatites in the Bewadze-Mankoadze area of the Kibi-Winneba Belt are of the rare element class, LCT
(Li-Cs-Ta) subclass, hosted by columbite-tantalite. From this knowledge, this research seeks to focus on the metal tantalum (Ta) hosted
in the pegmatites due to its significant concentrations.

Pegmatites are the host of the majority of the economic deposits of tantalum globally. Tantalum has the chemical symbol Ta and the
atomic number 73. Tantalus, a figure from Greek mythology, inspired the term tantalium [15]. Tantalum is a shiny, ductile, blue-gray
transition metal with extraordinary hardness. It is a refractory metal that is frequently added to powerful alloys with high melting

Fig. 1. Geological map of the Gomoa West District showing the Bewadze-Mankoadze towns.

E.D. Sunkari et al. Heliyon 10 (2024) e38176 

2 



points. Tantalum is a group 5 element that is always found in geologic sources next to the chemically related element, niobium [15].
Tantalum is a critical metal in today’s advanced technological applications. The extraction process takes place from oxide minerals
that are present as trace amounts in granitic rare-element pegmatites and rare-metal granites (REP). Exploitation of Nb-Zr-rare earth
element (REE) from carbonatites and peralkaline igneous rocks is a promising by-product of mining spodumene from pegmatites. Most
of the current Ta production comes from columbite-group minerals (CGM), with minor contributions from ixiolite, rutile, tapiolite,
wodginite and pyrochlore-supergroup minerals [16]. Africa’s estimated Ta resources are greater than 50,000 tons of contained Ta2O5,
accounting for 16 % of global resources [16].

Previous studies have reported the presence of rare earth metals in the pegmatites in the Bewadze-Mankoadze area of the Kibi-
Winneba volcanic belt [1,3,6,7] but the geochemical characteristics and exploration implications for the mineral tantalum in the
pegmatites have not been well documented [6,7]. investigated the mineralogical and petrogenetic features of the pegmatites and
concluded that the pegmatites are highly fractionated resulting in the enrichment of light rare earth elements (LREE) than middle rare
earth elements (MREE) and heavy rare earth elements (HREE). The authors further indicated that the pegmatites formed in late to post
orogenic tectonic settings with calc-alkaline affinities. With that, this research seeks to investigate the exploration implications for
tantalum mineralization in pegmatites at Bewadze-Mankoadze area within the Kibi-Winneba metavolcanic belt in the southern part of
Ghana. It will help identify the geochemical vectors of the mineral tantalum as a new mineral deposit to mining companies seeking to

Fig. 2. Field photographs of pegmatites in Bewadze-Mankoadze area (Qz = Quartz, Kfs = Potassium Feldspar, Ab = Albite, Ms = Muscovite, Tur =
Tourmaline, Spd = Spodumene).

E.D. Sunkari et al. Heliyon 10 (2024) e38176 

3 



expand their operations and increase their output.

2. Geology of the area

The pegmatites at the Bewadze-Mankoadze area which is located along the coast of Ghana forms part of the Birimian Supergroup
(2195-2072 Ma) [17], which lies at the southernmost end of the West African Craton [1,3,6,7]. Ghana’s Paleoproterozoic Birimian
Supergroup is made up of six metavolcanic (“greenstone”) belts, five of which are parallel, regularly spaced and exhibit
northeast-trending. The easternmost of these belts, the Kibi-Winneba metavolcanic belt, is where the Bewadze-Mankoadze area is
located [17,18]. According to Ref. [17], the ’’greenstones’’ (mainly quartz-hornblende-actinolite schist), amphibolite, and
hornblende-biotite-quartz schist are the host rocks of the Bewadze-Mankoadze area (Fig. 1). These rocks have hornblende- and
biotite-bearing intrusions. Magnesite and ultramafic tectonic dykes are often intruded into the rocks along foliation planes that follow
the NE-SW Birimian trend. Granitoids with accompanying pegmatite and quartz veins run NE and abruptly dip between 70 and 84◦ to
the SE [17,18]. The dykes, which are typically 30 m wide, have undergone partial or complete amphibolite (hornblendite) trans-
formation. Meta-basalts along the shore west of Mankoadze provide evidence of late-stage volcanic activity [17]. The Kibi-Winneba
volcanic belt, on the other hand, stretches from Winneba in the Central Region along the coast and strikes northeast for about 100
km into Kibi in the Eastern Region and beyond. The Cape Coast granitoid divides it into northern and southern halves, and it has a basic
synclinal structure with the Ashanti and Bui belts [19,20]. The sites of the widespread quartz-veined pegmatites and aplites that are
found across the belt are frequently closer to the large extent contacts with the basin-type granitoids. Most of the sills and dykes in this
area of the belt appear to be north-south (N-S) trending and have experienced early-stage metamorphism. A few dolerite dykes do not
appear to have undergone metamorphism [21].

The Paleoproterozoic Birimian Supergroup (2195-2072Ma) of Ghana was intruded by syn-volcanic granitoids during the Eburnean
orogeny (2120-2115 Ma). The last stage of this Eburnean tectonism is pegmatitic veining [17]. However, the exact age of the Birimian
pegmatites is unknown but they are believed to be of post-Eburnean age.

3. Materials and methods

3.1. Field mapping and sample collection

Fresh pegmatite samples from the Bewadze and Mankoadze communities were collected for this study. Ten samples were taken
from each study area and labelled accordingly, B and M for samples taken from Bewadze and Mankoadze, respectively. The surface of
the in-situ pegmatites was broken off to obtain fresh samples since those on the surface were altered. This was done to collect fresh
samples for good petrographic studies. Field photos of the pegmatite exposures are shown in Fig. 2.

3.2. Petrographic and bulk-rock geochemical analyses

Eight pegmatite samples were selected for thin section preparation and analysis. To identify the minerals and their relationships,
thin sections of pegmatites from the Bewadze-Mankoadze area were examined with a transmitted-light microscope, LEICA DM2700P
housed at the Geological Engineering Department of the University of Mines and Technology, Ghana. Various optical characteristics,
including colour, texture, twining, pleochroism, extinction angle, cleavage, bireflectance, and isotropism/anisotropy were used to
discriminate between different minerals.

In this study, 50 samples were collected from Bewadze-Mankoadze area for whole-rock geochemical analysis. The samples were
prepared for the geochemical analysis at the Ghana Research Reactor-1 (GHARR-1) laboratories. The geochemical analysis was carried
out by Ref. [7] using standard methods including X-ray Fluorescence Spectroscopy (XRF) at the Ghana Geological Survey Authority,
Accra, and Instrumental Neutron Activation Analysis (INAA) at the Ghana Atomic Energy Commission, Kwabenya - Accra.

3.3. Data synthesis and statistical analysis

3.3.1. Q-Q plots
The quantiles of a dataset are compared to the quantiles of a selected theoretical distribution using a Q-Q plot. The standard normal

distribution is the theoretical distribution that is most frequently utilised. The link between the observed quantiles and the corre-
sponding quantiles anticipated under the distribution is shown in the Q-Q plot. Q-Q plots are widely used to evaluate a dataset’s
assumed normality [22,23]. If the observed data points closely resemble a normal distribution, the diagonal line may be a good
approximation for the data [24].

3.3.2. Kolmogorov smirnov and Shapiro Wilk tests
It is frequently important in statistical analysis to compare two datasets or determine whether a dataset fits a specific distribution.

With the use of the robust statistical test known as the Kolmogorov-Smirnov, the ability of a single dataset to comply with a certain
theoretical distribution or whether two datasets exhibit the same distribution can be assessed [25]. The supremum class of EDF
(Empirical Distribution Function) statistics, which includes the Kolmogorov-Smirnov statistic (referred to as KS), is based on the
greatest vertical variation between the predicted and observed distribution [26]. For n ordered data points x1 < x2 < … <xn, the test
statistic according to Conover [26] can be computed from equation 1.

E.D. Sunkari et al. Heliyon 10 (2024) e38176 

4 



T= supx|F*(x) − Fn(x)| (1)

where;
sup = stands for supremum which means the greatest.
F*(x) = the hypothesized distribution function
Fn(x) = the EDF estimated based on the random sample
In KS test of normality, F*(x) is a normal distribution with known mean, μ, and a standard deviation, σ.
The KS test statistic is meant for testing,
H0: F(x) = F*(x) for all x from -∞ to ∞ (the data follows a specified distribution)
Ha: F(x) ∕= F*(x) for at least one value of x (the data do not follow the specified distribution).
If T is more than the 1-α quantile, H0 at the level of significance is neglected, α based on the quantile table for the Kolmogorov test

statistic. The dataset is said to be normal when it has a significant value greater than or equal to 0.05 (α ≥ 0.05)
The Shapiro-Wilk test, introduced in 1965 by Samuel Sanford Shapiro and Martin Bradbury Wilk, is a widely used statistical

technique for determining if a dataset has a normal distribution. If the p-value is greater than the chosen significant level, typically
0.05, the null hypothesis is accepted; however, if the p-value is less than the significant level, the null hypothesis is rejected, indicating
that the dataset deviates significantly from a normal distribution.

3.3.3. Spearman’s correlation
According to Ref. [23], the two-tailed Spearman’s Correlation approach may be used to identify the link between any two

geochemical variables while considering their multivariate and regionalised nature. The degree and direction of the monotonic link
between ranking variables are evaluated using the Spearman correlation test, a commonly used statistical technique. The null hy-
pothesis of the Spearman correlation test is the assumption that the variables do not have a monotonic relationship. The test statistic,
indicated as rho (ρ), reflects the strength and direction of the correlation, and has a range from − 1 to 1. In contrast, a negative rho
denotes a negative monotonic connection. A positive rho denotes a positive monotonic relationship [27].

3.3.4. Centred log-ratio (CLR) transformation
Given the deviation of geochemical data from normal distribution, the resulting multi-element dataset’s lowest and maximum

contents are lessened by the Centred Log Ratio (CLR) Transformation [22,23], which can be calculated from equation (2).

CLR(x)=
(

log
(
x1
g(x)

)

,…, log
(
xn
g(x)

))

(2)

where;
x = composition vector
x1, x2, …,xn = single elements in the samples
g(x) = geometric mean.

3.3.5. Factor and hierarchical cluster analyses
Factor analysis (FA), which uses the principal component approach, is a multivariate statistical method that may be used to identify

element relationships [23]. According to Ref. [28], the main objective of factor analysis is to uncover hidden multivariate data
structures and to reduce the number of "factors" necessary to adequately explain the variation in a multivariate data collection. The
Kaiser criterion was applied to determine the number of components to be extracted. Only key components with eigenvalues of 1 or
more are extracted using the Kaiser criteria. The covariance matrix and scree plots both display the number of components and the
similarity index, respectively [29].

Dendrograms were used to display results of the hierarchical cluster analysis (HCA). The dendrogram displays the steps taken by
the program to arrive at the answer, the multi-element clusters, and the distances (referred to as the squared Euclidean distance)
separating these multi-element clusters [29]. According to Ref. [29], the HCA technique is described by the computation of the
similarity of X number of components. Following that, two components are clustered under the constraint that a given agglomeration
criteria will be decreased because of the clustering. The two items are in a class that is created by this clustering. Furthermore, in the
aforementioned situation, two elemental classes are brought together by examining the resemblance or discrepancy between the
previously constructed class and the N-2 remaining components. This step is continued until all the components have been clustered.

3.3.6. Delineation of geochemical anomalies
Anomaly thresholds were determined to distinguish anomalies from background values; anomalies of interest in geochemistry are

those values that exceed the threshold. Calculating threshold anomalies for defining anomalies involves using the median absolute
deviation of [30]. The anomaly threshold can be expressed in equation (3) as:

T=med+ 2MAD (3)

where;
T = anomaly threshold
med = median value of the analysed elements.

E.D. Sunkari et al. Heliyon 10 (2024) e38176 

5 



MAD = median absolute deviation.
According to Ref. [30], the MAD (median absolute deviation), which measures data dispersion, is the most accurate statistical

indicator of variability in a univariate sample of quantitative data. MAD is determined using the formula (equation (4)):

MAD=median |xi − median (xi)| (4)

where xi is the element concentration.
An approach known as the "multi-element halos technique" can be used to outline geochemical halos surrounding mineral deposits

more effectively than using only one pathfinder element [31]. This approach allows one to use the formula (equation (5)) to get the
threshold values for the multi-element halos:

H(X+Y)=
(
X1
X0

+
Y1
Y0

)

;

(
X2
X0

+
Y2
Y0

)

;…;

(
XN
X0

+
YN
Y0

)

(5)

where;
X1, X2, …, XN = concentrations of X in samples 1, 2, …, N.
Y1, Y2, …, YN = concentrations of Y in samples 1, 2, …., N.
X0 and Y0 = threshold values of X and Y elements.
The single and multi-element threshold outcomes helped construct the geochemical anomaly maps of the research area using the

Inverse Distance Weighting (IDW)Method. Maps of single and multi-element anomaly dispersions were produced with the Geographic
Information System application software, ArcMap 10.7.1.

4. Results and discussion

4.1. Field relations and petrography of the Bewadze-Mankoadze pegmatites

Geological field mapping was first conducted to understand the trend of the pegmatites. It was observed that the pegmatites were

Fig. 3. Photomicrographs of pegmatites in the Bewadze area under transmitted light (Qz = Quartz, Spd = Spodumene, Plg = Plagioclase, Ms =
Muscovite, Kfs = Potassium Feldspar, Ab = Albite, CGM = Columbite Group Minerals).

E.D. Sunkari et al. Heliyon 10 (2024) e38176 

6 



trending in the NE/SW direction as shown in the geological map (Fig. 1). Rocks found included gneiss-hosted pegmatites with a few
quartz veins running through them, as well as mica schist towards the coast (Fig. 2a–f). They have an average dip reading of 43◦ and a
strike of 064◦. The pegmatite outcrops generally occur as smaller oval or tabular outcrops (Fig. 2a–c and e) except north of the
Egyasimaku hills where a huge, elongated outcrop occurs (30 m × 10 m) (Fig. 2b and d). They also occur as dykes intruding the mica
schists along the coast of Mankoadze (Fig. 2f). Pegmatites occur at Bewadze whiles both pegmatites and aplites occur at Mankoadze.
The aplites which contain green beryl and apatite largely occur at the northern and eastern parts of the Egyasimaku hills. The peg-
matites are coarse-to very coarse-grained whiles the aplites are fine-to medium-grained. The texture of the mineral grains is similar to
the texture of the pegmatites and aplites as reported by Refs. [3,6,7].

In terms of mineralogical composition, quartz, K-feldspar, albite, muscovite, spodumene, and black tourmaline were identified
under macroscopic view (Fig. 2a–f) during the field mapping. The tourmaline minerals are medium-grained in the aplites (Fig. 2a and
g) and medium-to very coarse-grained in the pegmatites (10 cm × 1 cm) (Fig. 2d). The spodumene are also coarse-grained (5 cm × 1
cm) (Fig. 2e). Other accessory minerals include pinkish garnet, columbite- and wodginite-group minerals [6,7].

Study of the Bewadze-Mankoadze pegmatites under the microscope shows anhedral to subhedral crystals. They are dominated by
minerals including quartz, K-feldspars, plagioclase, muscovite, spodumene, albite, tourmaline, columbite group minerals, and mon-
tebrasite (Fig. 3a–d; Fig. 4a–d) similar to the pegmatites of the Cape Coast sedimentary Basin of the Birimian Supergroup of Ghana
[32], the pegmatites of the Zenaga inlier in the Anti-Atlas, Morocco [33] and the Shigar Valley pegmatites, Pakistan [34]. Quartz in the
pegmatites was observed to have a first order of gray. Some other part of the quartz exhibits undulose extinction due the shattering and
deformation of the quartz mineral [35]. Potassium feldspar under the microscope had a second order dark-gray interference colour and
a distinctive twinning structure. Plagioclase was also observed to be displaying polysynthetic twining that is a zig-zag extinction
pattern due to its anisotropic nature. Muscovite and spodumene appeared to be the same but spodumene was more pleochroic than
muscovite. Muscovite had colours ranging from green to blue to purple and that of spodumene had differentiated colours due to the
presence of varying elements in different spodumene types. The colour variation in spodumene occurs because of different impurities
(e.g., Mn or Fe) in the mineral [36]. This resulted in the differentiation of the spodumene into spodumene I, II, III, IV, and V based on
the range of alteration colours (Fig. 3a–d; Fig. 4a–d). Also, muscovite had even cleavage plane whilst that of spodumene were uneven.

Fig. 4. Photomicrographs of pegmatites in the Mankoadze under transmitted light (Ab = Albite, Plg = plagioclase, Ms = Muscovite, Mb = Mon-
tebrasite, Tur = Tourmaline).

E.D. Sunkari et al. Heliyon 10 (2024) e38176 

7 



Albite was also seen to be exhibiting perthitic texture under the microscope. Globally, Ta has been reported to occur largely in
complex-type pegmatites and albite-rich aplites [4,9,37,38]. Hence, the high values of Ta at the Bewadze-Mankoadze area can be
concluded to be because of the occurrence of albite-rich pegmatites-aplites in the area as observed in the thin sections (Fig. 3a–d;
Fig. 4a–d). The columbite group of minerals appeared in black spots indicating the high presence of tantalite in them. Splintery
fractures were displayed by montebrasite having a first order colour of yellow.

4.2. Mineral assemblage paragenesis

The Ta mineralization and the formation of the host rock occurred in three stages: magmatic, magmatic-hydrothermal and the
hydrothermal stage (Fig. 5). During the magmatic stage the host rock began to form, and minerals related to the host rock like quartz
started to form in the magmatic stage and run into the magmatic hydrothermal stage and partly in the hydrothermal stage. K-feldspars
also started forming in the magmatic stage and ended up partly in the magmatic hydrothermal stage.

Plagioclase began forming in the magmatic-hydrothermal stage and ended up in the middle of the hydrothermal stage. Muscovite
also behaved similarly to the K-feldspar, forming in the magmatic stage, and ending up partly in the magmatic-hydrothermal stage.
Spodumene also started forming towards the end of the hydrothermal stage. Albite, tourmaline and the columbite group of minerals
developed towards the end of the magmatic stage moving into the magmatic-hydrothermal stage. Montebrasite started forming in mid
magmatic-hydrothermal stage and ended up in the mid hydrothermal stage.

4.3. Multi-element geochemistry

A summary of the multi-element geochemistry of the pegmatites in the Bewadze-Mankoadze area is presented in Table 1. The
following are the average concentrations and ranges of concentration (minimum and maximum) for the most significant elements:
0.69 ppm and 0.30–0.95 ppm for MgO; 260 ppm and 10.5–773 ppm for Ta; 38.5 ppm and 3.80–74.1 ppm for Nb; 851 ppm and
30.4–3465 ppm for Rb; 178 ppm and 10.0–551 ppm for Cs; 625 ppm and 22.0–1326 ppm for Cr; 417 ppm and 1.00–1574 ppm for Sm;
35.1 ppm and 1.80–92.3 ppm for Sc; 42.3 ppm and 23.5–60.4 ppm for Ga. This shows that the pegmatites are enriched in Cr, Cs, Rb,
Sm, and Ta and the rest having low concentrations as shown in the box and whisker plot (Fig. 6). Cs, Rb, and Ta have bulk continental
crust abundances of 2.00 ppm, 49.0 ppm, and 0.70 ppm, respectively [39]. Compared to these crustal abundances, the
Bewadze-Mankoadze pegmatites show remarkable enrichment in these elements as reported above, similar to the pegmatites of the

Fig. 5. Paragenetic sequence of mineral formation.

E.D. Sunkari et al. Heliyon 10 (2024) e38176 

8 



Cape Coast sedimentary Basin of the Birimian Supergroup of Ghana [32], the pegmatites of the Zenaga inlier in the Anti-Atlas, Morocco
[33] and the Shigar Valley pegmatites, Pakistan [34]. The elevated concentrations of these elements especially, Cs and Rb in the
pegmatites-aplites indicates the occurrence of pollucite in the area as reported by Ref. [1].

4.4. Implications of the petrographic analysis

The presence of quartz veinlets points to hydrothermal activities in the area. The columbite group of minerals, which formed during
the waning phase of the magmatic stage contribute to the high tantalum content due to the existence of tantalite in them. Other
minerals which contribute to the geochemistry in the area include albite, muscovite and tourmaline. They formed at the end of the
magmatic stage transitioning into the magmatic-hydrothermal stage of the pegmatite emplacement and these minerals contributed to

Table 1
Summary statistics of the multi-element geochemical data obtained (concentrations of the major elements are in wt% and that of the minor elements
are in ppm).

Element No. Min. Max. Mean STD Skewness Kurtosis

Na2O 12 0.72 3.80 2.80 0.88 − 1.28 1.75
MgO 12 0.30 0.95 0.69 0.16 − 0.97 2.34
Al2O3 12 10.8 18.5 12.8 2.27 1.74 2.6
SiO2 12 43.6 55.6 47.5 3.51 1.03 1.11
K2O 12 0.13 3.89 1.70 1.54 0.43 − 1.71
CaO 12 0.03 2.72 0.55 0.77 2.48 6.02
Fe2O3 12 0.16 31.2 3.38 8.79 3.43 11.8
V 12 10.0 94.0 62.4 26.0 − 1.45 1.40
Cr 12 22.0 1326 625 405 0.05 − 0.86
Co 12 4.50 17.0 9.13 4.26 0.88 − 0.53
Ni 12 15.3 68.6 31.4 14.0 1.83 4.23
Cu 12 1.50 4.10 2.42 0.70 1.19 2.31
Zn 12 3.30 344 68.5 95.6 2.52 7.03
Ga 12 23.5 60.4 42.3 8.66 − 0.19 2.63
As 12 0.40 12820 1836 4340 2.20 3.73
Rb 12 30.4 3465 851 1081 1.58 2.08
Y 12 0.50 3.00 1.30 0.82 0.83 − 0.07
Zr 12 6.10 291 60.0 99.6 2.03 2.73
Nb 12 3.80 74.1 38.5 25.5 0.18 − 1.29
Cs 12 10.0 551 178 145 1.50 3.34
Ba 12 13.0 501 104 155 2.00 3.43
La 12 17.0 20.0 18.8 0.87 − 0.44 0.23
Ce 12 24.0 61.0 32.0 11.6 1.99 3.28
Hf 12 2.80 14.3 5.75 3.76 1.49 1.33
Pb 12 0.90 13.2 3.06 3.88 2.22 4.21
Th 12 0.60 2.50 1.32 0.71 0.40 − 1.36
U 12 5.70 9.70 7.32 1.11 0.56 0.66
Ta 12 10.5 773 259 250 1.03 0.06
Sm 12 1.00 1574 417 573 1.08 − 0.24
Sc 12 1.80 92.3 35.1 31.4 0.70 − 0.92

Fig. 6. Box and Whisker plot showing the elemental concentration from the geochemical analysis.

E.D. Sunkari et al. Heliyon 10 (2024) e38176 

9 



Fig. 7. Q-Q Plots of Tantalum (Ta), Zircon (Zr) and Zinc (Zn) distributions before and after centred log-ratio transformation.

Table 2
Kolmogorov-Smirnov and Shapiro-Wilk tests for normality using untransformed geochemical data (values highlighted in red are <0.05).

Tests of Normality

Element Kolmogorov-Smirnova Shapiro-Wilk

Statistic df Sig. Statistic df Sig.

Na2O 0.163 12 0.200* 0.904 12 0.177
MgO 0.152 12 0.200* 0.935 12 0.435
Al2O3 0.27 12 0.016 0.796 12 0.008
SiO2 0.162 12 0.200* 0.91 12 0.211
K2O 0.209 12 0.154 0.835 12 0.024
CaO 0.422 12 0.00 0.604 12 0.00
Fe2O3 0.478 12 0.00 0.393 12 0.00
V 0.266 12 0.019 0.793 12 0.008
Cr 0.105 12 0.200* 0.97 12 0.914
Co 0.244 12 0.047 0.859 12 0.048
Ni 0.244 12 0.047 0.827 12 0.019
Cu 0.202 12 0.189 0.909 12 0.205
Zn 0.264 12 0.02 0.679 12 0.001
Ga 0.215 12 0.133 0.917 12 0.265
As 0.494 12 0.00 0.489 12 0.00
Rb 0.224 12 0.099 0.79 12 0.007
Y 0.186 12 0.200* 0.893 12 0.13
Zr 0.389 12 0.00 0.576 12 0.00
Nb 0.145 12 0.200* 0.924 12 0.321
Cs 0.202 12 0.189 0.869 12 0.064
Ba 0.34 12 0.00 0.664 12 0.00
La 0.28 12 0.01 0.884 12 0.099
Ce 0.299 12 0.004 0.695 12 0.001
Hf 0.277 12 0.012 0.775 12 0.005
Pb 0.401 12 0.00 0.605 12 0.00
Th 0.223 12 0.10 0.871 12 0.067
U 0.133 12 0.200* 0.963 12 0.821
Ta 0.209 12 0.155 0.878 12 0.082
Sm 0.327 12 0.001 0.761 12 0.004
Sc 0.21 12 0.149 0.891 12 0.123

E.D. Sunkari et al. Heliyon 10 (2024) e38176 

10 



the geochemistry of the pegmatites in the area as Rb, Cr and Sm were dominant. These elements form part of the chemistry of the
minerals that were dominant from the petrographic studies.

4.5. Multivariate statistical analysis

The multivariate statistical analysis considered every element whose concentration was measured as part of the geochemical data.

4.5.1. Q-Q plots
Q-Q plots of Ta, Zr and Zn in Fig. 7a–c, showed a deviation from the normal line. As geochemical data hardly follows normal

distribution, it indicates that the raw dataset is the result of several associations of elements, which might be connected to various
geochemical processes [40]. Extreme sample concentrations that seemed to be separate from the others were also noticed. Outliers are
seen in samples where certain element concentrations do not match the normal distribution [23]. Outliers in geochemistry are likely
indications of abnormalities connected to mineralization. As was already mentioned, these outliers are also typical of other
geochemical processes.

Centred log-ratio transformation was used to rectify them. By minimising the difference between the highest and lowest values in
the dataset, this transformation normalises the data and lessens the impact of outliers. Q-Q plots of the transformed data were shown in
Fig. 7d–f and compared with the untransformed Q-Q plots.

4.5.2. Kolmogorov-Smirnov and Shapiro Wilk tests
After the data was converted using the centred log-ratio transformation to test for normality, the Kolmogorov-Smirnov Test and

ShapiroWilk Test were carried out on the raw dataset. The test results indicate clearly that a few elements from the raw dataset had a p-
value greater than 0.05 in both the Kolmogorov-Smirnov and Shapiro Wilk tests (Table 2). Most of the p-values were greater than 0.05
after the data was subjected to centred log ratio transformation (Table 3).

4.5.3. Correlation analysis
Spearman correlation was used to assess the degree of correlation between the elements of interest. Since Spearman is a

nonparametric measure of rank correlation and may be used to assess a monotonic connection between variables, it was preferred to
alternatives like Pearson and Kendall. The Spearman correlation (Table 4) found a significant positive correlation between Ta and Cu

Table 3
Kolmogorov-Smirnov and Shapiro-Wilk tests for normality using transformed geochemical data (values highlighted in red are <0.05).

Tests of Normality

Element Kolmogorov-Smirnova Shapiro-Wilk

Statistic df Sig. Statistic df Sig.

Na2O 0.252 12 0.033 0.742 12 0.002
MgO 0.213 12 0.141 0.816 12 0.014
Al2O3 0.253 12 0.033 0.84 12 0.028
SiO2 0.163 12 0.200* 0.924 12 0.317
K2O 0.173 12 0.200* 0.873 12 0.071
CaO 0.253 12 0.033 0.909 12 0.204
Fe2O3 0.184 12 0.200* 0.876 12 0.079
V 0.387 12 0.00 0.629 12 0.00
Cr 0.245 12 0.045 0.824 12 0.018
Co 0.161 12 0.200* 0.921 12 0.292
Ni 0.184 12 0.200* 0.953 12 0.686
Cu 0.15 12 0.200* 0.965 12 0.858
Zn 0.099 12 0.200* 0.985 12 0.997
Ga 0.228 12 0.085 0.87 12 0.065
As 0.209 12 0.153 0.818 12 0.015
Rb 0.209 12 0.156 0.862 12 0.052
Y 0.165 12 0.200* 0.907 12 0.197
Zr 0.177 12 0.200* 0.857 12 0.044
Nb 0.187 12 0.200* 0.871 12 0.066
Cs 0.155 12 0.200* 0.912 12 0.227
Ba 0.222 12 0.107 0.848 12 0.035
La 0.286 12 0.007 0.88 12 0.089
Ce 0.237 12 0.061 0.77 12 0.004
Hf 0.244 12 0.047 0.865 12 0.056
Pb 0.235 12 0.066 0.802 12 0.01
Th 0.217 12 0.124 0.849 12 0.036
U 0.126 12 0.200* 0.976 12 0.962
Ta 0.18 12 0.200* 0.925 12 0.327
Sm 0.261 12 0.024 0.819 12 0.016
Sc 0.139 12 0.200* 0.921 12 0.293

E.D. Sunkari et al. Heliyon 10 (2024) e38176 

11 



Table 4
Spearman correlation analysis (r ≥ 0.50).

Na2O MgO Al2O3 SiO2 K2O CaO Fe2O3 V Cr Co Ni Cu Zn Ga As Rb Y Zr Nb Cs Ba La Ce Hf Pb Th U Ta Sm Sc

Na2O 1.00 ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​
MgO − 0.16 1.00 ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​
Al2O3 − 0.70 − 0.41 1.00 ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​
SiO2 − 0.32 0.46 0.22 1.00 ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​
K2O − 0.73 − 0.02 0.79 0.52 1.00 ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​
CaO 0.07 − 0.70 0.22 − 0.70 − 0.32 1.00 ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​
Fe2O3 − 0.22 0.07 0.01 0.07 − 0.15 0.19 1.00 ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​
V 0.07 0.48 − 0.21 0.63 0.23 − 0.82 − 0.21 1.00 ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​
Cr − 0.13 − 0.07 0.08 0.16 − 0.06 0.10 0.00 − 0.33 1.00 ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​
Co − 0.30 − 0.24 0.26 − 0.13 − 0.14 0.53 0.77 − 0.39 0.22 1.00 ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​
Ni − 0.51 − 0.14 0.64 0.54 0.56 − 0.09 0.41 0.15 − 0.22 0.31 1.00 ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​
Cu − 0.56 0.21 0.17 0.17 0.28 0.06 0.28 − 0.28 0.59 0.32 − 0.12 1.00 ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​
Zn − 0.04 0.55 − 0.45 0.17 − 0.17 − 0.32 0.40 0.33 − 0.06 0.31 − 0.15 0.35 1.00 ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​
Ga − 0.04 0.16 − 0.18 − 0.15 − 0.11 0.05 0.00 − 0.11 0.27 0.28 − 0.39 0.43 0.70 1.00 ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​
As − 0.06 0.00 0.08 − 0.18 − 0.23 0.22 0.60 − 0.24 − 0.17 0.50 0.27 − 0.24 − 0.09 − 0.32 1.00 ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​
Rb − 0.75 0.29 0.52 0.46 0.85 − 0.46 − 0.04 0.24 − 0.05 − 0.10 0.36 0.45 0.24 0.25 − 0.34 1.00 ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​
Y − 0.80 0.05 0.77 0.52 0.97 − 0.29 − 0.01 0.15 − 0.01 − 0.05 0.57 0.41 − 0.09 − 0.08 − 0.23 0.90 1.00 ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​
Zr − 0.22 0.12 0.30 0.06 − 0.11 0.19 0.32 − 0.30 0.23 0.45 0.27 − 0.08 − 0.22 − 0.15 0.70 − 0.18 − 0.08 1.00 ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​
Nb 0.15 0.72 − 0.70 0.11 − 0.43 − 0.26 − 0.02 0.20 0.11 − 0.15 − 0.49 0.32 0.61 0.40 − 0.32 − 0.08 − 0.35 − 0.18 1.00 ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​
Cs − 0.45 0.42 0.22 0.25 0.52 − 0.35 − 0.10 − 0.12 0.08 − 0.31 0.01 0.48 0.13 0.27 − 0.24 0.73 0.60 − 0.06 0.14 1.00 ​ ​ ​ ​ ​ ​ ​ ​ ​ ​
Ba − 0.27 − 0.48 0.54 − 0.25 0.25 0.28 − 0.02 − 0.27 0.26 0.29 0.05 0.03 − 0.45 − 0.18 0.44 − 0.01 0.18 0.35 − 0.71 − 0.18 1.00 ​ ​ ​ ​ ​ ​ ​ ​ ​
La − 0.39 − 0.47 0.49 − 0.25 0.17 0.58 0.63 − 0.38 − 0.08 0.78 0.38 0.25 0.01 − 0.05 0.51 0.01 0.20 0.18 − 0.48 − 0.25 0.52 1.00 ​ ​ ​ ​ ​ ​ ​ ​
Ce 0.23 − 0.69 0.19 − 0.55 − 0.20 0.74 0.12 − 0.48 0.00 0.40 − 0.13 − 0.08 − 0.30 − 0.07 0.39 − 0.50 − 0.28 0.08 − 0.45 − 0.45 0.55 0.67 1.00 ​ ​ ​ ​ ​ ​ ​
Hf − 0.63 − 0.18 0.67 0.13 0.40 0.28 0.38 − 0.40 0.50 0.64 0.32 0.54 − 0.06 0.15 0.41 0.27 0.43 0.54 − 0.38 0.17 0.66 0.64 0.37 1.00 ​ ​ ​ ​ ​ ​
Pb − 0.47 − 0.38 0.80 0.40 0.78 0.00 0.01 0.06 − 0.10 − 0.04 0.70 0.02 − 0.56 − 0.59 0.05 0.38 0.73 0.01 − 0.68 0.11 0.38 0.39 0.18 0.38 1.00 ​ ​ ​ ​ ​
Th − 0.77 − 0.30 0.87 0.37 0.80 0.11 0.12 − 0.22 0.25 0.19 0.64 0.44 − 0.39 − 0.27 − 0.04 0.56 0.83 0.11 − 0.51 0.33 0.38 0.43 0.05 0.64 0.85 1.00 ​ ​ ​ ​
U − 0.44 − 0.19 0.31 − 0.29 0.06 0.50 0.44 − 0.50 0.34 0.78 − 0.05 0.60 0.34 0.56 0.29 0.14 0.11 0.24 − 0.08 0.05 0.45 0.72 0.48 0.77 − 0.09 0.25 1.00 ​ ​ ​
Ta − 0.03 0.37 − 0.33 0.16 0.00 − 0.23 − 0.13 0.05 0.41 − 0.11 − 0.46 0.65 0.62 0.76 − 0.62 0.32 0.06 − 0.45 0.66 0.48 − 0.45 − 0.33 − 0.31 − 0.04 − 0.44 − 0.14 0.23 1.00 ​ ​
Sm − 0.62 0.48 0.11 0.29 0.22 − 0.33 0.50 0.18 0.08 0.28 0.27 0.36 0.21 − 0.23 0.45 0.31 0.31 0.32 0.08 0.11 0.25 0.29 − 0.29 0.38 0.13 0.28 0.23 − 0.16 1.00 ​
Sc − 0.01 0.40 − 0.38 0.10 − 0.06 − 0.18 − 0.07 0.03 0.40 − 0.05 − 0.52 0.69 0.66 0.76 − 0.57 0.27 0.01 − 0.43 0.70 0.43 − 0.42 − 0.27 − 0.25 − 0.03 − 0.48 − 0.19 0.29 0.99 − 0.11 1.00

E.D.Sunkarietal.
Heliyon 10 (2024) e38176 

12 



(r = 0.65), Ta and Zn (r = 0.62), Ta and Nb (r = 0.66), Sc and Cu (r = 0.69), Sc and Zn (r = 0.66) as well as Sc and Ga (r = 0.76).
A high to strong positive correlation between the constituent components indicates their paragenetic existence [41]. The distri-

bution of the elements in the region was caused by intricate geochemical processes [23]. Thus, similar mechanisms are responsible for
their enrichment. The negative correlation between some elements is caused by a variety of processes, including the distribution and
concentration of elements caused by deformational activity during metamorphic events and hydrothermal enhancement or deposition
of elements. The element of interest, Ta correlates moderately to strongly with Sc, Ga, Nb, and Cs. This may be an indication that Sc,
Ga, Nb, and Cs could have occurred paragenetically with Ta.

4.5.4. Factor analysis (FA)
Seven (7) factors, which defined 95.355 % of the total variance having eigenvalues more than 1 were extracted in the FA (Table 5).

Factor 1 accounts for 34.891 % of the total variance (Table 6) and is composed of Ta-Cs-Zn-Nb-MgO-Sc-Ga-Vmulti-element association
(Fig. 8). This factor’s elements are generated by wall rock and host rocks. Ta-Nb-MgO-Sc-Ga-V are usually associated with the host
rocks and that of Cs-Zn are associated with the wall rocks [42]. Factor 2 explains 21.807 % of the total variance (Table 6) and is
composed of Hf-Zr-Ba-CaO-As-Comulti-element association (Fig. 8). This factor is mainly associated with host rocks and arsenic (As) in

Table 5
Rotated component matrix showing the distribution of elements into factors.

Rotated Component Matrixa

Component

Element 1 2 3 4 5 6 7

Pb − 0.898 − 0.059 0.378 − 0.103 0.035 0.037 0.098
Ce − 0.885 0.209 − 0.125 − 0.025 − 0.069 − 0.137 0.286
Ga 0.843 − 0.262 0.243 0.088 − 0.136 − 0.266 0.185
Nb 0.832 − 0.239 − 0.263 0.269 0.065 0.207 − 0.132
MgO 0.822 − 0.290 0.103 − 0.224 − 0.098 0.391 − 0.088
Sc 0.787 − 0.328 − 0.020 0.434 − 0.070 0.083 0.235
Ta 0.785 − 0.380 − 0.016 0.459 − 0.081 − 0.011 0.099
Zn 0.764 − 0.222 − 0.149 − 0.026 0.408 0.095 0.376
Cs 0.677 − 0.078 0.559 − 0.040 − 0.237 − 0.075 − 0.188
Zr − 0.214 0.922 − 0.043 0.091 0.180 0.089 − 0.025
As − 0.314 0.880 − 0.095 − 0.076 0.228 0.067 0.194
V 0.360 − 0.860 − 0.060 − 0.200 0.013 0.123 − 0.176
Hf − 0.279 0.674 0.352 0.488 0.058 0.116 0.270
Ba − 0.541 0.650 0.102 − 0.122 − 0.323 0.143 0.316
CaO − 0.426 0.627 − 0.183 − 0.004 0.123 − 0.546 0.221
Co − 0.074 0.604 − 0.074 0.134 0.579 − 0.072 0.455
Y 0.058 − 0.043 0.970 0.023 0.105 0.164 0.075
K2O − 0.100 − 0.079 0.943 − 0.178 0.080 0.230 0.034
Rb 0.203 − 0.121 0.930 − 0.078 0.112 0.183 0.009
Th − 0.417 0.187 0.780 0.336 0.066 0.084 0.039
Al2O3 − 0.167 0.323 0.768 0.221 − 0.005 − 0.396 0.207
Cr 0.269 0.298 − 0.166 0.854 − 0.118 − 0.015 − 0.071
Cu 0.294 0.037 0.255 0.783 − 0.011 0.379 0.287
Ni − 0.157 0.108 0.441 − 0.179 0.783 0.081 − 0.173
Fe2O3 0.087 0.181 0.014 − 0.033 0.760 0.087 0.265
Sm 0.027 0.450 0.174 0.028 0.414 0.719 0.192
Na2O − 0.141 − 0.186 − 0.540 − 0.340 0.190 − 0.676 − 0.049
SiO2 − 0.023 − 0.285 0.408 0.320 0.330 0.633 − 0.209
U 0.179 0.529 0.126 0.223 0.080 0.023 0.771
La − 0.351 0.383 0.191 0.007 0.330 − 0.061 0.754

Table 6
Total variance explained for transformed dataset.

Total Variance Explained

Factor Initial Eigenvalues Extraction Sums of Squared Loadings

Total % of Variance Cumulative % Total % of Variance Cumulative %

1 10.467 34.891 34.891 10.467 34.891 34.891
2 6.542 21.807 56.698 6.542 21.807 56.698
3 4.607 15.358 72.056 4.607 15.358 72.056
4 2.554 8.513 80.569 2.554 8.513 80.569
5 2.053 6.843 87.412 2.053 6.843 87.412
6 1.309 4.363 91.776 1.309 4.363 91.776
7 1.074 3.579 95.355 1.074 3.579 95.355

E.D. Sunkari et al. Heliyon 10 (2024) e38176 

13 



this factor may be derived from hydrothermal sources.
Factor 3 also accounts for 15.358 % of the total variance (Table 6) and is made up of Y-Pb-K2O-Al2O3-Th-Rb association (Fig. 8),

which are mainly from the wall rocks with Pb from hydrothermal sources. Factor 4 makes up 8.513 % of the total variance (Table 6)
and is composed of Cr-Cu mainly sourced from mafic lithologies and hydrothermal fluids. Factor 5 with Na2O-Fe2O3-Ni element as-
sociation (Fig. 8) explains 6.843 % of the total variance (Table 6), which is associated with a mixture of the wall rocks and the host
rocks.

Factor 6 explains 4.363 % of the total variance (Table 6) and is made up of SiO2-Sm association (Fig. 8) and is associated with wall
rocks of the pegmatites, which are the gneisses from the Birimian Supergroup. Factor 7 with La-Ce-U association (Fig. 8) accounts for
3.579 % of the total variance of the data (Table 6). La-Ce is of the host rocks and U is from the hydrothermal sources [43].

4.5.5. Hierarchical cluster analysis for transformed dataset
The dendrogram generated from the transformed dataset shows three groups of multi-element relationships (Fig. 9). Cluster 1 has

multi-element associations of Zr, As, La, Hf, CaO, U, Co, Pb, Ce, Ba and Na2O; cluster 2 consists of V, Nb, Ta, Ga, Sc, Cr, Cu, Nb andMgO
multi-element association; and cluster 3 is made up of multi-element association of Ni, SiO2, Cs, Zn, Sm, Rb, K2O, Y, Th, Al2O3 and
Fe2O3 (Fig. 9). This confirms the sources inferred from the FA, mainly from the host rocks, wall rocks, and hydrothermal sources.

4.5.6. Comparison of multivariate statistical techniques
Using the dataset from the centred log-ratio, cluster 2 with element associations of Ta-V-Nb-Ga-Sc-Cr-Cu-MgO and factor 1 with

element association of Ta-Cs-Zn-Nb-MgO-Sc-Ga-V from factor analysis show similar elemental association of Ta-Nb-Ga-Sc. Cesium (Cs)
was also considered due to its similarity with tantalum and was also found to correlate good with Ta and in the same factor with Ta as
well. Considering all these, using cluster 2 from the HCA, factor 1 from the FA, and also the correlation analysis, Sc-Ga-Nb-Cs could be
considered as the elemental cluster associated with tantalum deposits in the Bewadze-Mankoadze area.

4.6. Single and multi-element halo mapping

4.6.1. Mapping single element geochemical anomalies
Single element halo maps were plotted for the elements as shown in Figs. 10 and 11a-d. Green to light green colouration indicates

low elemental concentrations, yellowish green to yellow indicates moderate elemental concentrations while the orange to red shows
high anomalous concentrations of the elements. The Ta anomaly map (Fig. 10) showed notable Ta anomalous concentration in the
south-western part of the study with moderate concentrations towards the northeastern part of the study area.

Anomaly maps for Sc, Ga, Nb and Cs are shown in Fig. 11a–d. Scandium (Sc) anomaly of the study shows high concentrations in the
south-western domain of the study area and reduced toward the north-eastern part of the study area. Gallium (Ga) geochemical
anomaly showed high concentrations in the middle of the southwestern domain of the study area. Niobium (Nb) anomaly map also
shows high concentrations in the southwestern side of the study area. Cs showed high concentrations in the southwestern and a few
moderate to high concentration in the northern part of the study area. These patterns clearly overlap with the pattern displayed by Ta
(Fig. 10). Hence, they can indeed be regarded as pathfinder elements of Ta.

4.6.2. Multi-element anomaly mapping and exploration implications
Using the thresholds as computed and geochemical halos, characterised by the presence of high concentration of pathfinder

Fig. 8. Rotated component matrix showing the factors and total variances in percentages.

E.D. Sunkari et al. Heliyon 10 (2024) e38176 

14 



elements in relation to Ta mineralization, multi-element halo maps were generated. These multi-elemental anomaly maps were
produced based on the elemental relationships. The Ta-Sc, Ta-Ga, Ta-Nb (Fig. 12a–c) maps show moderate to high concentrations in
the south-western part of the study area and low concentrations towards the north-eastern part of the study area. The Ta-Cs anomaly
map (Fig. 12d) shows high concentrations in the southwestern corridor of the study area and moderate concentrations in the north-
eastern part of the study area. Ta-Sc-Ga-Nb-Cs (Fig. 13) map shows moderate to very high concentrations in the southwestern corridor
and low to very low concentrations towards the northeastern corridor.

From this study, the multi-element patterns reveal similar geochemical anomalies that are concentrated in the south-western part of
the study area. Prospecting for tantalum should therefore focus on this area and the pegmatites serve as host rocks of the tantalum in
the Bewadze-Mankoadze area. This suggests that using these elements as pathfinder elements can significantly enhance tantalum
mineral discovery in the area.

4.7. Implications and guide to tantalum exploration

Tantalum is only found in significant amounts in granitic pegmatites that belong to the rare-element category of orogenic pegmatite
associations [44]. These pegmatites are abundant in southern Ghana, especially in Ewoyaa, Biriwa, Winneba, Bewadze, and

Fig. 9. Dendrogram from hierarchical cluster analysis showing three (3) elemental clusters.

E.D. Sunkari et al. Heliyon 10 (2024) e38176 

15 



Mankoadze areas. In the Bewadze-Mankoadze pegmatite field, post-tectonic intrusions of productive granites and their pegmatite
aureoles are connected to fault systems or lithologic boundaries within schists and gneisses containing quartz, K-feldspar, plagioclase,
albite, muscovite, spodumene, tourmaline, columbite group minerals, and montebrasite as confirmed by the petrographic analysis in
this study. Specific granite and pegmatite clusters might have tantalum mineralization in the pegmatitic-granite domes of the original
granites, typically located in pegmatites at the outer edges of the aureoles. Decreasing metamorphic grade in host rocks, increasing
complexity in texture and paragenesis of pegmatites, evolving typomorphic minerals, and higher fractionation [7] suggest areas with
greater tantalum potential.

The geochemical data reveals that the pegmatites are enriched in Cr, Cs, Rb, Sm, and Ta. These elements are reported in literature as
pathfinders to LCT pegmatites [9–11], hence confirming that the Ta pegmatites in this study are indeed LCT-type pegmatites.
Multivariate statistics identified Sc, Ga, Nb, and Cs as the primary elements associated with Ta in the Bewadze-Mankoadze area.
Geochemical anomalies related to these elements are present in the southern part of the study area, as indicated by single and
multi-element halo mapping. Exploration for Ta mineralization should concentrate on the southern-western region of the study area,
where anomalies of the pathfinder elements are situated. To evaluate the overall level of fractionation for Ta enrichment, different
geochemical indicators are present in individual pegmatitic granites and pegmatite bodies in the study area. In the initial phase of
exploration for Ta-pegmatites in the Bewadze-Mankoadze area, the relationship between Ta and Cs in coarse-grained early white mica
is found to be the most informative indicator. Albitization of pegmatites commonly conceals the original zonal pattern yet enhances the
recoverable quantity of Ta oxides [45]. Hence, albitized pegmatites undergoing metasomatic changes with high levels of both Ta and

Fig. 10. Single element halo maps showing the anomaly distribution of Ta.

E.D. Sunkari et al. Heliyon 10 (2024) e38176 

16 



Cs are the most promising targets for exploration. Geochemical and geophysical techniques can be used to locate and delineate the
Ta-pegmatite deposits.

5. Conclusion

This study focused on hard rock geochemistry of the Bewadze-Mankoadze area. Petrographic studies reveal the presence of
plagioclase, quartz, spodumene, columbite groupminerals, albite, tourmaline, muscovite, potassium feldspar andmontebrasite. Cr, Cs,

Fig. 11. Single element halo maps showing the anomaly distribution of (a) Sc (b) Ga (c) Nb and (d) Cs.

E.D. Sunkari et al. Heliyon 10 (2024) e38176 

17 



Rb, Sm, and Ta are the most dominating elements in the Bewadze-Mankoadze area, according to box and whisker plots. Q-Q plots
involving the untransformed geochemical data showed sporadic behaviour revealing that different geological processes contributed to
the mineralization in the area. Spearman’s Correlation, factor analysis and hierarchical cluster analysis indicated Sc, Ga, Nb and Cs as
pathfinders of Ta and can be associated directly to the pegmatites in the Bewadze-Mankoadze area. Geochemical anomaly maps reveal
high anomaly values of Ta and its identified pathfinder elements in the south-western corridor. Hence, exploration activities for Ta
mineralization should focus on the south-western corridor, where the anomalies of the pathfinder elements are located. During the
preliminary investigation of Ta-pegmatites in the Bewadze-Mankoadze area, the correlation between Ta and Cs in coarse-grained early
white mica proves to be the most significant marker. The original zonal pattern in pegmatites is often obscured by albitization, but it

Fig. 12. Multi-element halo maps showing the anomaly distribution of (a) Ta-Sc (b)Ta-Ga (c) Ta-Nb (d) Ta-Cs.

E.D. Sunkari et al. Heliyon 10 (2024) e38176 

18 



increases the amount of recoverable Ta oxides. Therefore, pegmatites that have been altered by albitization and are experiencing
metasomatic changes with significant concentrations of both Ta and Cs present the most potential for exploration.

This study is a contribution to the global transition to clean energy since Ta is a critical mineral that can be used in the production of
capacitors. Tantalum capacitors play a crucial role in energy-saving technologies, such as renewable energy systems. They facilitate
effective storage and transportation of energy in solar panels, wind turbines, and electric vehicles, aiding in decreasing reliance on
fossil fuels and decreasing greenhouse gas emissions. Therefore, if the Ta deposit is economically harnessed, it will boost the foreign
exchange earnings of Ghana, and the by-products can be used for clean energy production in the country. Moreover, the Ta-bearing
pegmatites in the Bewadze-Mankoadze area contain a wide variety of precious minerals, including spodumene, beryl, and tourmaline.
These minerals frequently appear as massive, properly formed crystals, and their presence raises the market value of pegmatites in
Ghana. In all, the discovery of the Ta-bearing pegmatites in the Bewadze-Mankoadze area adds to the list of granitic rare-element
pegmatites and rare-metal granites found in Africa. This study did not investigate the age of the Ta-bearing pegmatites and the
timing of the ore settlement as well as the fluid evolution history. Therefore, it is recommended that detailed geochronological, stable
isotopes, and fluid inclusion studies be carried out on the Ta-bearing pegmatites in the area to constrain the age of the pegmatites and
the mineralization and the genesis of the Ta ore.

Fig. 13. Multi-element halo maps showing the anomaly distribution of Ta-Sc-Ga-Nb-Cs.

E.D. Sunkari et al. Heliyon 10 (2024) e38176 

19 



Data and code availability statement

Data included in article.

CRediT authorship contribution statement

Emmanuel Daanoba Sunkari: Writing – review & editing, Writing – original draft, Visualization, Validation, Supervision, Soft-
ware, Resources, Project administration, Methodology, Investigation, Funding acquisition, Formal analysis, Data curation, Concep-
tualization. Joshua Nkansah:Writing – review & editing, Writing – original draft, Visualization, Validation, Software, Methodology,
Investigation, Formal analysis, Data curation, Conceptualization. Salaam Jansbaka Adams: Writing – review & editing, Writing –
original draft, Visualization, Validation, Supervision, Resources, Project administration, Methodology, Funding acquisition.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to
influence the work reported in this paper.

Acknowledgement

This research is part of the second author’s BSc project work under the supervision of the first author at the University of Mines and
Technology, Ghana. The Government of Ghana is duly appreciated for the financial support through the Book and Research Allowance
to Faculty Members.

References

[1] A.K. Agomor, Geological indications and prospecting guide to rare metalmineralisation of birimian granite pegmatite of Ghana, in: Proceeding of the
International Conference on the Geology of Ghana with Emphasis on Gold, vol. 1, Geological Survey Department, Accra, Ghana, 1987, pp. 2–24.

[2] P. Černý, Rare-element granitic pegmatites Part I: anatomy and internal evolution of pegmatite deposits, Geosci. Can. 18 (2) (1991) 49–67. https://id.erudit.
org/iderudit/geocan18_2art01.

[3] C.I. Chalokwu, M.A. Ghazi, E.E. Foord, Geochemical characteristics and K Ar ages of rare-metal bearing pegmatites from the Birimian of southeastern Ghana,
J. Afr. Earth Sci. 24 (1–2) (1997) 1–9, https://doi.org/10.1016/S0899-5362(97)00022-5.

[4] R.L. Linnen, M. Van Lichtervelde, P. Černý, Granitic pegmatites as sources of strategic metals, Elements 8 (4) (2012) 275–280, https://doi.org/10.2113/
gselements.8.4.275.

[5] S.J. Adams, Rare-metal mineralization of the winneba-mankoadze pegmatites, southwestern Ghana; MPhil thesis. Graduate School of Nuclear and Allied
Sciences, University of Ghana, 2013, pp. 1–101pp.

[6] S.J. Adams, M. Van Lichtervelde, P.O. Amponsah, P.M. Nude, D.K. Asiedu, S.B. Dampare, Characterisation and rare-metal potential of the Winneba-Mankoadze
pegmatites, Southern Ghana: evidence of two pegmatite fields, J. Afr. Earth Sci. 207 (2023) 105049, https://doi.org/10.1016/j.jafrearsci.2023.105049.

[7] S.J. Adams, E.D. Sunkari, J.E.K. Hanson, K.K. Tandoh, S.E.K. Tetteh, P.M. Nude, D.K. Asiedu, S.B. Dampare, Geochemistry of the Winneba-Mankoadze
pegmatites in southern Ghana: a clue to the petrogenesis of the pegmatites, Ore and Energy Resource Geology 17 (2024) 100051, https://doi.org/10.1016/j.
oreoa.2024.100051.

[8] D. London, Ore-forming processes within granitic pegmatites, Ore Geol. Rev. 101 (2018) 349–383, https://doi.org/10.1016/j.oregeorev.2018.04.020.
[9] P. Černý, T.S. Ercit, The classification of granitic pegmatites revisited, Can. Mineral. 43 (6) (2005) 2005–2026.
[10] H.G. Dill, Pegmatites and aplites: their genetic and applied ore geology, Ore Geol. Rev. 69 (2015) 417–561, https://doi.org/10.1016/j.oregeorev.2015.02.022.
[11] H.G. Dill, The CMS classification scheme (chemical composition – mineral assemblage – structural geology) – linking geology to mineralogy of pegmatitic and

aplitic rocks, Neues Jahrbuch Mineral. Abhand. 193 (3) (2016) 231–263, https://doi.org/10.1127/njma/2016/0304.
[12] A. Guastoni, LCT (lithium, cesium, tantalum) and NYF (niobium, yttrium, fluorine) pegmatites in the central alps. Proxies of exhumation history of the alpine

nappe stack in the lepontine dome. PhD Thesis in Earth Science. Italy, University of Padua, 2012, p. 161.
[13] B. Bekele, A.K. Sen, The mineral chemistry of gahnite, garnet and columbite-group minerals (CGM): implications for genesis and evolution of the Kenticha Rare-

element granite-pegmatite, Adola, Ethiopia, J. Afr. Earth Sci. 162 (2020) 103691, https://doi.org/10.1016/j.jafrearsci.2019.103691.
[14] T.S. Ercit, L.A. Groat, R.A. Gault, Granitic pegmatites of the O’Grady batholith, NWT, Canada: a case study of the evolution of the elbaite subtype of rare-element

granitic pegmatite, Can. Mineral. 41 (1) (2003) 117–137.
[15] S.L. Lee, M. Doxbeck, J. Mueller, M. Cipollo, P. Cote, Texture, structure and phase transformation in sputter beta tantalum coating, Surf. Coating. Technol.

177–178 (2004) 44–51, https://doi.org/10.1016/j.surfcoat.2003.06.008.
[16] F. Melcher, T. Graupner, T. Oberthür, P. Schütte, Tantalum-(niobium-tin) mineralisation in pegmatites and rare-metal granites of Africa, S. Afr. J. Geol. 120 (1)

(2017) 77–100, https://doi.org/10.25131/gssajg.120.1.77.
[17] J. Agyei-Duodu, G.K. Loh, K.O. Boamah, M. Baba, W. Hirdes, M. Toloczyki, D.W. Davis, Geological Survey Department of Ghana (GSD), Geological Map of

Ghana, Accra, Ghana, 2009, pp. 1–1000000. Report.
[18] K.K. Tandoh, S.F. Gbedemah, M.R.S. Doku, S.J. Adams, B.A. Brako, Petrographic and structural analysis of Paleoproterozoic Birimian granitoids and associated

rocks of Boankra areas in the Kumasi Basin of Ghana, Arabian J. Geosci. 16 (1) (2023) 8, https://doi.org/10.1007/s12517-022-11010-8.
[19] B. Eisenlohr, The Structural Geology of Birimian and Tarkwaian Rocks of Southwest Ghana: Republic of Ghana. Federal Republic of Germany, Technical

Cooperation Project No. 80.2040. 6, BGR, 1989.
[20] B.N. Eisenlohr, W. Hirdes, The structural development of the early proterozoic birimian and tarkwaian rocks of southwest Ghana, west Africa, J. Afr. Earth Sci.

14 (3) (1992) 313–325, https://doi.org/10.1016/0899-5362(92)90035-B.
[21] R.J. Griffis, K. Barning, F.L. Agezo, F.K. Akosah, Gold deposits of Ghana, Minerals Commission Report (2002) 19–37.
[22] E.D. Sunkari, M. Appiah-Twum, A. Lermi, Using laterite geochemistry for exploration of orogenic gold deposits in the Wa-Lawra belt, NW Ghana: kunche in

perspective. Ömer Halisdemir Üniversitesi Mühendislik Bilimleri Dergisi, Niğde Ömer Halisdemir Üniversitesi 7 (3) (2018) 1137–1141, https://doi.org/
10.28948/ngumuh.502334.

[23] E.D. Sunkari, M. Appiah-Twum, A. Lermi, Spatial distribution and trace element geochemistry of laterites in Kunche area: implication for gold exploration
targets in NW, Ghana, J. Afr. Earth Sci. 158 (2019) 103519, https://doi.org/10.1016/j.jafrearsci.2019.103519.

[24] R.R. Wilcox, Understanding and Applying Basic Statistical Methods Using R, John Wiley & Sons, 2016, p. 478.
[25] M.A. Stephens, EDF statistics for goodness of fit and some comparisons, J. Am. Stat. Assoc. 69 (347) (1974) 730–737, https://doi.org/10.1080/

01621459.1974.10480196.

E.D. Sunkari et al. Heliyon 10 (2024) e38176 

20 

http://refhub.elsevier.com/S2405-8440(24)14207-7/sref1
http://refhub.elsevier.com/S2405-8440(24)14207-7/sref1
https://id.erudit.org/iderudit/geocan18_2art01
https://id.erudit.org/iderudit/geocan18_2art01
https://doi.org/10.1016/S0899-5362(97)00022-5
https://doi.org/10.2113/gselements.8.4.275
https://doi.org/10.2113/gselements.8.4.275
http://refhub.elsevier.com/S2405-8440(24)14207-7/sref5
http://refhub.elsevier.com/S2405-8440(24)14207-7/sref5
https://doi.org/10.1016/j.jafrearsci.2023.105049
https://doi.org/10.1016/j.oreoa.2024.100051
https://doi.org/10.1016/j.oreoa.2024.100051
https://doi.org/10.1016/j.oregeorev.2018.04.020
http://refhub.elsevier.com/S2405-8440(24)14207-7/sref9
https://doi.org/10.1016/j.oregeorev.2015.02.022
https://doi.org/10.1127/njma/2016/0304
http://refhub.elsevier.com/S2405-8440(24)14207-7/sref12
http://refhub.elsevier.com/S2405-8440(24)14207-7/sref12
https://doi.org/10.1016/j.jafrearsci.2019.103691
http://refhub.elsevier.com/S2405-8440(24)14207-7/sref14
http://refhub.elsevier.com/S2405-8440(24)14207-7/sref14
https://doi.org/10.1016/j.surfcoat.2003.06.008
https://doi.org/10.25131/gssajg.120.1.77
http://refhub.elsevier.com/S2405-8440(24)14207-7/sref17
http://refhub.elsevier.com/S2405-8440(24)14207-7/sref17
https://doi.org/10.1007/s12517-022-11010-8
http://refhub.elsevier.com/S2405-8440(24)14207-7/sref19
http://refhub.elsevier.com/S2405-8440(24)14207-7/sref19
https://doi.org/10.1016/0899-5362(92)90035-B
http://refhub.elsevier.com/S2405-8440(24)14207-7/sref21
https://doi.org/10.28948/ngumuh.502334
https://doi.org/10.28948/ngumuh.502334
https://doi.org/10.1016/j.jafrearsci.2019.103519
http://refhub.elsevier.com/S2405-8440(24)14207-7/sref24
https://doi.org/10.1080/01621459.1974.10480196
https://doi.org/10.1080/01621459.1974.10480196


[26] W.J. Conover, Practical Nonparametric Statistics, third ed., John Wiley & Sons, Inc., New York, 1999, pp. 428–433.
[27] C. Spearman, The proof and measurement of association between two things, Am. J. Psychol. 100 (3/4) (1987) 441–471, https://doi.org/10.2307/1422689.
[28] J. Zhao, S. Chen, R. Zuo, Identifying geochemical anomalies associated with Au–Cu mineralization using multifractal and artificial neural network models in the

Ningqiang district, Shaanxi, China, J. Geochem. Explor. 164 (2016) 54–64, https://doi.org/10.1016/j.gexplo.2015.06.018.
[29] P.M. Nude, J.M. Asigri, S.M. Yidana, E. Arhin, G. Foli, J.M. Kutu, Identifying pathfinder elements for gold in multi-element soil geochemical data from the Wa-

Lawra belt, Northwest Ghana: a multivariate statistical approach, Int. J. Geosci. 3 (1) (2012) 62–70, https://doi.org/10.4236/ijg.2012.31008.
[30] F.N. David, J.W. Tukey, Exploratory data analysis, Biometrics 33 (4) (1977), https://doi.org/10.2307/2529486, 768-768.
[31] A.P. Reis, A.J. Sousa, E. Cardoso Fonseca, Soil geochemical prospecting for gold at marrancos (northern portugual), J. Geochem. Explor. 73 (1) (2001) 1–10,

https://doi.org/10.1016/S0375-6742(01)00169-8.
[32] P.M. Nude, J.E.K. Hanson, S.B. Dampare, T.T. Akiti, S. Osae, E.S. Nyarko, N. Zkaria, S. Enti-Brown, Geochemistry of Pegmatites associated with the cape coast

granite complex in the Egyaa and Akim Oda areas of southern Ghana, Ghana J. Sci. 51 (2011) 89–100. https://www.ajol.info/index.php/gjs/article/view/
116697.

[33] Y. Morsli, Y. Zerhouni, A. Wafik, A. Elouazzani, S. Alikouss, A. Saidi, Z. Baroudi, Eburneen pegmatites of the Zenaga inlier (Anti-Atlas, Morocco): petrography,
geochemistry, and classification, J. Afr. Earth Sci. 186 (2022) 104438, https://doi.org/10.1016/j.jafrearsci.2021.104438.

[34] M. Hassan, Mineralogy and Geochemistry of the Gemstones and the Gemstone- Bearing Pegmatites in the Shigar Valley of Skardu Northern Areas of Pakistan,
PhD Thesis, National Centre of Excellence in Geology, University of Peshawar, 2007, pp. 1–361.

[35] E.D. Sunkari, B.M. Kore, S.E.K. Tetteh, Petrography and structural features of the Precambrian basement rocks in the Benin-Nigerian Shield, NW Nigeria:
implications for their correlation with South Atlantic Precambrian terranes, AIMS Geosciences 8 (4) (2022) 503–524, https://doi.org/10.3934/geosci.2022028.

[36] T. Roy, B. Plante, M. Benzaazoua, I. Demers, Geochemistry and mineralogy of a spodumene-pegmatite lithium ore at various mineral beneficiation stages, Miner.
Eng. 202 (2023) 108312, https://doi.org/10.1016/j.mineng.2023.108312.

[37] P. Černý, The Tanco rare-element pegmatite deposit, Manitoba; regional context, internal anatomy, and global comparisons, in: R.L. Linnen, I.M. Samson (Eds.),
Rare Element Geochemistry and Mineral Deposits, vol. 17, Geological Association of Canada Short Course Notes, 2005, pp. 127–158.

[38] M. Van Lichtervelde, S. Salvi, D. Beziat, R.L. Linnen, Textural features and chemical evolution in tantalum oxides: magmatic versus hydrothermal origins for Ta
mineralization in the Tanco Lower pegmatite, Manitoba, Canada, Econ. Geol. 102 (2) (2007) 257–276, https://doi.org/10.2113/gsecongeo.102.2.257.

[39] R.L. Rudnick, Composition of the continental crust. The Crust, Treatise on Geochemistry 3 (2005) 17–18.
[40] R. Zuo, Identifying geochemical anomalies associated with Cu and Pb–Zn skarn mineralization using principal component analysis and spectrum–area fractal

modeling in the Gangdese Belt, Tibet (China), J. Geochem. Explor. 111 (1–2) (2011) 13–22, https://doi.org/10.1016/j.gexplo.2011.06.012.
[41] D.J. Lapworth, K.V. Knights, R.M. Key, C.C. Johnson, E. Ayoade, M.A. Adekanmi, T.M. Arisekola, O.A. Okunlola, B. Backman, M. Eklund, P.A. Everett, R.

T. Lister, J. Ridgway, M.J. Watts, S.J. Kemp, P.E.J. Pitfield, Geochemical mapping using stream sediments in west-central Nigeria: implications for
environmental studies and mineral exploration in West Africa, Appl. Geochem. 27 (6) (2012) 1035–1052, https://doi.org/10.1016/j.apgeochem.2012.02.023.

[42] R. Morgenstern, R.E. Turnbull, M.P. Hill, P.M.J. Durance, M.S. Rattenbury, Rare earth element mineral potential in New Zealand, GNS sci consult Rep 23 (2018)
243–250.

[43] S. Makvandi, G. Beaudoin, M.B. Beth McClenaghan, D. Quirt, P. Ledru, PCA of Fe-oxides MLA data as an advanced tool in provenance discrimination and
indicator mineral exploration: case study from bedrock and till from the Kiggavik U deposits area (Nunavut, Canada), J. Geochem. Explor. 197 (2019) 199–211,
https://doi.org/10.1016/j.gexplo.2018.11.013.

[44] P. Černý, Exploration strategy and methods for pegmatite Depositsof tantalum, in: P. Moller, P. Černý, F. Saupe, Tantalum Lanthanides, Niobium (Eds.), Special
Publication No. 7 of the Society for Geology Applied to Mineral Deposits, Springer, Berlin, 1989, pp. 274–302, https://doi.org/10.1007/978-3-642-87262-4_13.

[45] P. Möller, G. Morteani, Geochemical exploration guide for tantalum pegmatites, Econ. Geol. 82 (7) (1987) 1888–1897, https://doi.org/10.2113/
gsecongeo.82.7.1888.

E.D. Sunkari et al. Heliyon 10 (2024) e38176 

21 

http://refhub.elsevier.com/S2405-8440(24)14207-7/sref26
https://doi.org/10.2307/1422689
https://doi.org/10.1016/j.gexplo.2015.06.018
https://doi.org/10.4236/ijg.2012.31008
https://doi.org/10.2307/2529486
https://doi.org/10.1016/S0375-6742(01)00169-8
https://www.ajol.info/index.php/gjs/article/view/116697
https://www.ajol.info/index.php/gjs/article/view/116697
https://doi.org/10.1016/j.jafrearsci.2021.104438
http://refhub.elsevier.com/S2405-8440(24)14207-7/sref34
http://refhub.elsevier.com/S2405-8440(24)14207-7/sref34
https://doi.org/10.3934/geosci.2022028
https://doi.org/10.1016/j.mineng.2023.108312
http://refhub.elsevier.com/S2405-8440(24)14207-7/sref37
http://refhub.elsevier.com/S2405-8440(24)14207-7/sref37
https://doi.org/10.2113/gsecongeo.102.2.257
http://refhub.elsevier.com/S2405-8440(24)14207-7/sref39
https://doi.org/10.1016/j.gexplo.2011.06.012
https://doi.org/10.1016/j.apgeochem.2012.02.023
http://refhub.elsevier.com/S2405-8440(24)14207-7/sref42
http://refhub.elsevier.com/S2405-8440(24)14207-7/sref42
https://doi.org/10.1016/j.gexplo.2018.11.013
https://doi.org/10.1007/978-3-642-87262-4_13
https://doi.org/10.2113/gsecongeo.82.7.1888
https://doi.org/10.2113/gsecongeo.82.7.1888

	Geochemical exploration for tantalum in coltan-rich pegmatites at Bewadze-Mankoadze area of the Kibi-Winneba Belt, southern ...
	1 Introduction
	2 Geology of the area
	3 Materials and methods
	3.1 Field mapping and sample collection
	3.2 Petrographic and bulk-rock geochemical analyses
	3.3 Data synthesis and statistical analysis
	3.3.1 Q-Q plots
	3.3.2 Kolmogorov smirnov and Shapiro Wilk tests
	3.3.3 Spearman’s correlation
	3.3.4 Centred log-ratio (CLR) transformation
	3.3.5 Factor and hierarchical cluster analyses
	3.3.6 Delineation of geochemical anomalies


	4 Results and discussion
	4.1 Field relations and petrography of the Bewadze-Mankoadze pegmatites
	4.2 Mineral assemblage paragenesis
	4.3 Multi-element geochemistry
	4.4 Implications of the petrographic analysis
	4.5 Multivariate statistical analysis
	4.5.1 Q-Q plots
	4.5.2 Kolmogorov-Smirnov and Shapiro Wilk tests
	4.5.3 Correlation analysis
	4.5.4 Factor analysis (FA)
	4.5.5 Hierarchical cluster analysis for transformed dataset
	4.5.6 Comparison of multivariate statistical techniques

	4.6 Single and multi-element halo mapping
	4.6.1 Mapping single element geochemical anomalies
	4.6.2 Multi-element anomaly mapping and exploration implications

	4.7 Implications and guide to tantalum exploration

	5 Conclusion
	Data and code availability statement
	CRediT authorship contribution statement
	Declaration of competing interest
	Acknowledgement
	References