Petrological and geochemical characteristics and K-Ar ages for Cenozoic
tinguaite dykes of the Roztoky Intrusive Centre, České středohoří
Mountains, NW Bohemia
J. Ulrych, J.K. Novák, M. Lang, Praha, K. Balogh, Debrecen, E. Hegner, München, Z.
Řanda, Řež
With 11 Figures and 11 Tables
ULRYCH, J., NOVÁK, J.K., LANG, M., BALOGH, K., HEGNER, E. ŘANDA, Z. (XXXX):
Petrological and geochemical characteristics and K-Ar ages for Cenozoic tinguaite dykes of
the Roztoky Intrusive Centre, České středohoří Mountains, NW Bohemia. – N. Jb. Mineral.
Abh. (XXX): XXX–XXX; Stuttgart.
Abstract: Dykes of tinguaite-textured (acicular clinopyroxene arranged in radial or crisscross patterns interstitially in a mosaic of alkali feldspar and foids) phonolites, hereafter
referred
to
as
tinguaites,
represent
the
most
evolved
rocks
in
the
alkaline
(camptonite/monchiquite – phonolite/tinguaite) dyke suite of the Cenozoic Roztoky Intrusive
Centre (RIC), České středohoří Mts. Petrographic and field evidence reveal different textural
varieties of dykes ranging from tinguaites, porphyritic tinguaites to mixed-textured phonolite–
tinguaites. K-Ar ages of 31–28.5 Ma for five samples suggest that the tinguaitic dyke rocks
are coeval with the subvolcanic members of the RIC at 33 to 24 Ma. But, since field evidence
requires a younger crystallization age for the dykes than the RIC, we suggest that excess Ar
may be responsible for some of the older ages obtained in this study. The development of a
tinguaitic texture depends on the rate of magma cooling as can be concluded from the
observation that thin phonolitic dykes exhibit trachytic texture whereas thick dykes show a
tinguaitic texture. Thick dykes may also show glassy tinguaitic margins and trachytic-textured
phonolitic central parts. Rare perrierite and an as-yet unidentified Zr-silicate mineral are
among the accessory minerals. The tinguaite samples are rich in alkali and other incompatible
elements and have high concentrations of Cl, SO3, CO2, and H2O. The alkalinity index (A.I.)
is high and ranges from 0.79 to 0.96. Initial
87
Sr/86Sr ratios range from 0.70335 to 0.70419
and initial Nd values of +2.1 to +4.3 indicating mantle-derivation of their parental magmas.
1
The chemically evolved composition of the tinguaites was attained via melt differentiation at
shallow crustal levels with limited assimilation of crust.
Key words: České středohoří Mts., Ohře/Eger Rift, Roztoky Intrusive Center, tinguaite dykes,
Cenozoic volcanism.
Introduction
The Ohře/Eger Rift and associated Cenozoic alkaline volcanism are prominent geological
features of the NW part of the Bohemian Massif (BM). The volcanic rocks comprise an
olivine nephelinite/tephrite – phonolite and basanite – trachyte series that makes up the České
středohoří Mountains (CS). The structural position of the CS has been described by Hibsch
(1926) and Kopecký (1987). The geology of the Roztoky Intrusive Centre (RIC), representing
the main volcanic centre of the CS, was studied by Kopecký (1987), Ulrych et al. (1983),
Ulrych (1998, 2000), Ulrych and Balogh (2000), and Cajz (2003). Petrological and
geochemical characteristics of the rock associations of the RIC were published by Ulrych et
al. (1983), Jelínek et al. (1989), and Ulrych (1998). In this paper we present a petrographic,
mineralogical, and geochemical study of the youngest phonolitic – tinguaitic dykes of the RIC
(Hibsch 1936).
Geological setting of the Roztoky Intrusive Center
More than one thousand dykes of alkali lamprophyres, felsic differentiates and
undifferentiated rocks related to hypabyssal/subvolcanic intrusions of essexite/monzodiorite –
sodalite syenite series, are associated with the RIC (Hibsch, 1926; Jelínek et al., 1989; Ulrych,
1998; Cajz, 2003). Most of the dykes are radially arranged around the central part of the RIC
that consists of a trachytic vent breccia. There are also preferred strike directions for
compositionally distinct dyke rocks (Ulrych and Balogh, 2000; Cajz, 2003) and the felsic
dykes have prevailing strike directions of 90o, 0o, 40o and steep dips (80 to 90o). The average
dyke width is ~ 1m but widths range from 10 cm to 10 m. Structural analysis by Cajz (2003)
revealed even more complicated systems of dykes with preferred strike directions. Crosscutting relationships indicate that the felsic dykes belong to the youngest igneous rocks of the
RIC (Hibsch, 1936). This finding was only partly confirmed by the first K-Ar data reported by
Ulrych and Balogh (2000) probably due to disturbance of the K-Ar isotopic system.
2
The Roztoky Intrusive Centre including dykes within a distance of 17 km consists of
three major intrusive series (Ulrych and Balogh, 2000). In order of decreasing age these are:
(1) Hypabyssal, mildly alkaline rocks, 33 to 28 Ma, including essexite, monzodiorite,
sodalite syenite with diopside–kaersutite cumulates (?) that are also present in camptonites
and gauteites.
(2) Alkaline dykes, 31 to 26 Ma, including tephrite/basanite(?), monchiquite/camptonite,
tephriphonolite, and phonolite/tinguaite. The felsic dyke rocks only occur in close proximity
to the centre of the RIC.
(3) Mildly alkaline dykes, 28 to 24 Ma, including trachybasalt, gauteite (two generations),
camptonite/mondhaldeite, bostonite (two generations)/sodalite syenite porphyry, and trachyte.
The moderate degree of melt differentiation and enrichment in incompatible element, as well
as depletion in compatible elements, is an essential feature of the RIC. Primitive mantlenormalized concentrations of incompatible elements exhibit a progressive enrichment in the
series through essexite – monzodiorite – sodalite syenite – mafic dykes to the felsic dykes.
The felsic dyke rocks show a specific chemical differentiation trend and prevailing N–S and
NE–SW strikes in addition to E–W strikes, most typical for lamprophyric RIC dykes (Ulrych
and Balogh, 2000).
Methods of investigation
The major element concentrations were determined using wet chemical methods (Analyst J.
Janovská and J. Šikl, Czech Geological Survey, Prague). Analyses of the international
reference whole-rock standards (GM, TB, BN) and duplicate analyses of the samples suggest
total errors of  5% (1 s.d.).
The trace element concentrations were determined by instrumental neutron activation
analysis (INAA). Samples powders of 50–90 mg were sealed in polyethylene foils and
irradiated for 4 hours in a nuclear reactor LWR-15 of the Institute of Nuclear Research at Řež.
The neutron flux rate was 8.5 × 1013 cm-2s-1. The irradiated samples were measured after three
days and four months. The gamma-ray spectrometric system, equipped with a HPGe-detector,
was operated with the following parameters: efficiency 22 % and resolution FWHM 1.8 keV
for photons 1332.5 keV of 60Co. The counting time was 1.5 hours in the first run and 24 hours
in the second one. The precision (1 s.d.) is better than 10 % for the REE and 5–10 % for
other trace elements. The accuracy of the data, checked against international rock standards, is
better than 10% for the REE and 5% for the other trace elements.
3
Mineral analyses were carried out on a Cameca SX-100 electron microprobe using the
wavelength dispersive technique (Analyst: A. Langrová, Institute of Geology AS CR). The
beam diameter was 10 µm with an accelerating potential of 15 kV. A beam current of 20 nA
was measured on a Faraday cup. A counting time of 10s was used for all elements. The
standards employed were of synthetic (SiO2, TiO2, Al2O3, Fe2O3, and MgO) and natural
composition (jadeite, apatite, leucite, diopside, and spinel [all Kα], and barite [Lα]). Data
reduction was performed with X-PHI correction.
The K-Ar isotope measurements were carried out according to the procedure described in
Balogh (1985). For Sm-Nd isotopic work was carried out in the isotope laboratory at
Universität München according to the procedures outlined in Hegner et al. (1995). The
isotopic measurements were carried out on a MAT 261 in a dynamic quadruple mass
collection mode. The
143
external precision of the
Nd/144Nd ratios were normalized to
143
146
Nd/144Nd = 0.7219. The
Nd/144Nd ratios is 1.2 × 10-5 as has been with an Ames Nd
standard solution yielding 0.512142 ± 12 (N = 35), corresponding to 0.511854 in the La Jolla
Nd reference standard material. The Nd values were calculated with the parameters of
Jacobsen and Wasserburg (1980). Present-day values for the chondrite uniform reservoir
(CHUR):
147
Sm/144Nd = 0.1967,
143
Nd/144Nd = 0.512638 at
146
Nd/144Nd = 0.7219. 87Sr/86Sr
ratios were measured in a dynamic double cup mass collection mode and normalized to
86
Sr/88Sr = 0.1194. The NIST 987 reference material yielded 87Sr/86Sr = 0.71022 (N= 22).
Results of the study
Rock descriptions and sample localities
The phonolite/tinguaite dykes occur in small numbers in the central part of the RIC (Fig. 1).
In combination with the trachytic dykes they make up ca. one hundred dykes out of the total
of ca. one thousand dykes in the RIC (Ulrych and Balogh, 2000). The following tinguaitic
types were recognized in the RIC and a brief description of the dykes and their host rocks is
presented in Table 1. The location of the dykes is shown in Fig. 1.
1. Tinguaite (type locality Serra de Tinguá, Brazil). The type locality of tinguaite in the CS is
located west of the village of Skrytín near Roztoky. Here we collected sample PH-3/1 (details
in Table 1) from a dyke of aphyric tinguaite ca. 4 m thick. It exhibits a typical tinguaitic
texture with cellular arrangement sensu Voitsechovsky et al. (1997) (Fig. 2 A). An additional
phonolitic dyke has been described from an outcrop in the Královský potok Valley near the
village of Český Bukov (Souček et al., 1985 – No. 55). It is remarkable in that the phonolite is
4
the most evolved rock type with the highest enrichment of incompatible element so far
documented in the CS.
2. Porphyritic tinguaite (type locality Devil’s Tower, Wyoming). Sample PH-3A comes from
Skrytín village where (details in Table 1). The dyke is 4–5 m thick and exhibits a typical
porphyritic rock with up to 50 vol.% of phenocrysts set in a tinguaitic groundmass. This rock
was first described by Hibsch (1899) as “Tinguaitporphyr”. A similar dyke was described by
Hibsch (1902) as “Nephelinporphyr – Foyaitporphyr” which occurs in a railway cut near the
village of Roztoky. At this locality there are four subparallel dykes (thickness 0.5 to 15 m)
transecting the phonolite intrusion Bradlo/Pradelberg. Sample (PH-88 from the centre and
PH-89 from the margin of the thickest dyke (Table 1). The dyke exhibits no chilled margins
and/or sharp contacts with the host phonolite (cf. Hibsch, 1902). After the recognition of a
tinguaitic texture we interpret the rock here as an example of porphyritic tinguaite.
3. Tinguaites with mixed-textures (type locality “Flur Ratschin” near the village of Zubrnice)
This group comprises porphyritic phonolite transitional to porphyritic tinguaite in the centre
of dykes and glassy porphyritic tinguaite in dyke margins. It was first described by Hibsch
(1910) who referred to an outcrop standing 8 to 12 m high. Today the dyke outcrop is only an
about 2 m high with a 1–1.5 m thick rampart of small outcrops and boulders. Most of the dyke
is composed of a felsic rock with a trachytic texture showing transitions to a porphyritic
tinguaitic texture. The narrow margins of the dyke are formed by an altered glassy rock with
porphyritic tinguaitic texture. At this locality we collected sample PH-87 (Table 1, Fig. 2 B).
4. Phonolites and trachytes forming intrusions and rare dykes with common trachytic texture
and mineral paragenesis have been described from many localities of the RIC. Phonolite
sample PH-2 collected near the town of Roztoky represents the host rock of the tinguaite dyke
from which samples PH-88 and PH-89 were collected. Samples PH-3 and POL-235 were
collected in a trachytic breccia hosting the tinguaite dykes from which samples PH-3A and
PH-3/1 were collected at the village of Skrytín (Table 1).
The felsic dyke rocks associated with the RIC are mostly of trachytic to phonolitic
composition. In detail, the majority of trachytes and “pseudotrachytes” reported in Kopecký
(1987) plot in the phonolite, tephriphonolite and also trachyandesite fields in the TAS
classification diagram of Le Maitre et al. (2002).
K-Ar geochronology
5
New K-Ar ages for tinguaitic whole-rock samples (PH-3/1a, -3/A and -87), were determined
on feldspar and analcime separates from PH-88, and a nepheline concentrate from PH-3/1b
(Table 2). The ages range from 30.9 ± 1.2 (1σ) Ma to 28.6 ± 1.1(1σ) Ma. Considering the
analytical error of ages, this time interval suggests a brief period for the emplacement of the
tinguaite dykes. The young age of 25.6 ± 1.0 Ma measured on sample PH-3/1a will be
discussed below.
Nepheline is traditionally interpreted as having good Ar retentivity and therefore useful
for K-Ar dating (McDougal and Harrison, 1988, and references therein, Faure, 1986).
However, Lippolt et al. (1990) demonstrated the presence of excess Ar in Quaternary
phonolitic rocks from the eastern Eifel, and also found rocks with initial
40
Ar/36Ar ratios as
low as 280, that resulted in erroneously young K-Ar ages. The authors explained this
phenomenon as being due to uptake of fractionated Ar from the atmospheric origin by the
feldspathoids. This explanation is plausible for the Tertiary samples analyzed by Lippolt et al.
(1990), but excess argon would not significantly change the age of Palaeogene rocks.
Significant amounts of excess Ar have been detected in altered nepheline by Zhirov et al.
(1968) and by Balogh et al. (1999) who could show a correlation of excess Ar with the degree
of alteration. They interpreted this as evidence for supply of excess Ar by the fluids
responsible for the alteration of the samples.
Taking these findings into account, K-Ar ages for nepheline may indicate the true age, or
alternatively, in the presence of excess Ar, too great an age. However, Balogh et al. (2005)
reported that leucite-nepheline-bearing alkali basalts may lose significant amounts of
radiogenic Ar during the baking of the sample at 250° C in the Ar extraction line. They
showed that whereas nepheline and leucite do lose Ar, this loss can be avoided if the
extraction line is baked only at 150 °C. They assumed that syngenetic and an unspecified
submicroscopic alteration may be responsible for the lowered Ar retentivity of nepheline and
leucite.
In order to study this possible effect in the samples from the Roztoky Intrusive Centre,
nepheline sample separated from sample PH-3/1b was dated after regular and reduced baking
temperatures. The measured Ar concentrations indicate some loss of radiogenic Ar at the
regular baking temperature of 250 °C, but this effect is insufficient to explain the age
difference between PH-3/1a (25.6 ± 1.0 Ma) and -3/1b of ca. 28 Ma. We suggest instead that
the young age is indeed due to resetting of the K-Ar system during alteration. Accepting the
values measured after moderate baking of the Ar extraction line for the real geological age of
6
PH-3/1b, we note that out of the 5 determinations (Table 2) the four younger ages are in close
agreement, ranging between 31 and 28.5 Ma (aver. 28.9 ± 0.4 Ma).
This age interval overlaps the time-span suggested for the formation of the subvolcanic
members of the Roztoky Intrusive Centre by Ulrych & Balogh (2000). This, however, does
not contradict the opinion of Hibsch (1936), who – on the basis of geological criteria –
recognized that the intrusion of the tinguaite dykes marked the final stage of magmatic
activity. However, a fully convincing interpretation of the young K-Ar age is still missing.
Ulrych & Balogh (2000) published a set of K-Ar ages from different laboratories, on
differentiated dyke rock series from the Roztoky Centre. These samples yielded ages of
28.2 Ma to as young as 23.6 Ma and were interpreted as evidence for prolonged magmatic
activity. This may be true, but a systematic chronological investigation including the Ar-Ar
dating technique needs to be carried out so that effects of overprinting by younger thermal
events and/or fluids can be identified.
Petrological characteristics of the tinguaite dykes
The pertinent petrographical characteristics of the tinguaite dykes and their host rocks are
presented in Table 1. Despite the fact that the tinguaites are uniform with respect their modal
(cf. Table 2) and chemical (cf. Table 3) compositions, samples can have contrasting textural
characteristics. Although they underwent crystallization at similar PT conditions in a similar
geological environment, these cogenetic rocks substantially differ in their textures, grain-size,
phenocryst/groundmass ratios, and abundances of recrystallized volcanic glass.
Chemical composition of minerals
Feldspar
Feldspar is the most abundant mineral in the tinguaites, occurring both as phenocrysts and as
a major proportion of the groundmass (cf. Table 5). Their normative composition is plotted in
the Ab – Or – An diagrams (Fig. 3a, b, c, d).
K-oligoclase, mantled by anorthoclase – Na-sanidine of limited composition, forms the
cores of the phenocrysts in tinguaite sample PH-3/1. Rims around these consist of sanidine to
Na-sanidine. Alkali feldspar of the groundmass is rich in an Or-component. In addition there
is albite in rare quantities. In the coarse-porphyritic sample PH-3A, anorthoclase – Koligoclase forms the cores and anorthoclase – sanidine the rims. The alkali feldspar of the
groundmass is also anorthoclase (Or33-39) but it does not have so high an Or-content (Or69-76)
7
as the sanidine rims of the phenocrysts. The celsian component in the alkali feldspars ranges
from 2–6 mol.% with the high contents confined to the rims of the feldspars. Finely
porphyritic tinguaite samples PH-88 and PH-89 reveal a distinct chemical composition of
their alkali feldspars. The chemical zoning of small phenocrysts of Na-sanidine is minor,
except for an enrichment in the celsian component (Cn = 5.7) in the core (Fig. 2C). This
contrasts with the situation in PH-3/A where there is enrichment of the clesian component in
both rims and groundmass feldspar (Cn = 5.3 and 6.1, respectively). The alkali feldspars of
the matrix are both sanidine and K-oligoclases-andesine. The porphyritic tinguaite/phonolite
sample PH-87 is characterized by poikilitic ternary K-andesine to K-oligoclase cores in the
microphenocrysts mantled by K-oligoclase – anorthoclase rims. This reflects successively
increasing potassium contents from core to rim as is also found in the porphyritic tinguaite
PH–3A. An atoll texture of microphenocrysts is characterized by the presence of glass (Fig.
2D). The alkali feldspar of the groundmass is anorthoclase; microlites formed by groundmass
recrystallization have a composition of oligoclase.
Clinopyroxene
Clinopyroxene occurs as phenocrysts and in minor amounts in the groundmass of the
tinguaites. Their composition plots in the QUAD-field attributing them to the diopside–
hedenbergite series (Fig. 6) of Morimoto ed. (1988) classification. An exception is the
clinopyroxene of the groundmass of sample PH-3/1 that corresponds to aegirine. Low-Ti
aluminian hedenbergite with little chemical zoning forms phenocrysts in tinguaite and
porphyritic tinguaite. It displays a small variation in the composition of the end-member
molecules (Table 5, Fig. 4). Euhedral phenocrysts showing characteristically narrow rims and
reaction zones document changing conditions during the crystallization of the phenocrysts and
groundmass. Diopside phenocrysts with typical atoll texture with a glass rim and outer rim
formed by K-oligoclase (Fig. 2 E) occur only in the porphyritic tinguaite/phonolite sample
PH-87. Diopside is also present as resorbed relict cores of some phenocrysts in sample PH3A. The amount of the Na is lower than that of Fe3+ (in atoms p.f.u.) implying the presence of
an Ae-component (7–17 mol.%) and suggesting Ca + (Mg, Fe) – Na + Fe3+ substitution. Rare
subhedral prisms of aegirine (Ae-component up to 90 mol.% in sample PH-3/1) and Na-rich
diopside (Ae-component up to 18 mol.% in sample PH-88) in the groundmass shows that
these rocks have an agpaitic character.
8
Amphibole
Amphibole occurs predominantly as phenocrysts and more rarely in the groundmass of the
tinguaites and porphyritic tinguaites. Compositionally zoned phenocrysts vary according to
the IMA recommendations from ferroan pargasite to ferro-pargasite (Leake ed., 1998; see Fig.
5). The TiO2 contents range from 2.5 to 4.1 wt% (Table 6, Fig. 5). Kaersutite and ferrokaersutite forming the rims of clinopyroxene phenocrysts in tinguaite sample PH-3/1 and
trachyte to porphyritic tinguaite sample PH-87, respectively, are more magnesian than the
pargasite of the phenocrysts. The samples also show ubiquitous reaction rims of amphibole
phenocrysts in contact with matrix.
Feldspathoids and analcime
Feldspathoids occur both as phenocrysts and groundmass in various proportions. The
compositions of nepheline, nosean, and sodalite are presented in Table 8.
Nepheline phenocrysts were mostly resorbed in the tinguaites and porphyritic tinguaites. The
relicts were confirmed in porphyritic tinguaite samples PH-3A and PH-3/1, whereas in other
samples they were either totally analcimized or altered to a mixture of clays and zeolites (Fig.
2F). The chemical composition of the nepheline plots near the composition of Si-poor
nepheline and does not exceed the limit defining an excess of SiO2 in solid solution in the Ne
– Ks – Qz – H2O system at 700 oC and 1 kbar pH2O (Fig. 6) defined by Hamilton (1961). The
compositional variation of the Ks-component ranging from 11.5–16.0 mol.% is small.
Noseane occurs only as microphenocrysts in tinguaite sample PH-3/1. Sodalite occurs as
pristine grains in interstices of tinguaite sample PH-3A. Analcime not only replaces nepheline
but is also present in the interstices.
Glass
The variation in the chemical composition of glass, frequently occurring in the groundmass,
can be attributed to an evolving composition of the residual liquid (Table 9). Three types of
glassy inclusions of 100 to 250 μm were distinguished in the tinguaite samples PH-87 and
PH-89.
Type I glass reveals a low A/CNK index of 1.2 to 2 and a high Na/(Na+K) ratio of 0.96 to
0.99 due to low K and high Na contents. The alkalinity index (A.I.) of 0.7 to 1.4 is highly
variable corresponding to rocks of alkaline to peralkaline compositions. Due to the generally
high SiO2 content of 52 to 62 wt.% and relatively low Na2O content, the composition of the
glass is rather albitic than nephelinic.
9
Type II glass is most common and characterized by an extremely high A/CNK index of 7.1 to
12.5. It markedly differs from type I glass in its low total alkali contents of 0.5 to 0.7 wt.% vs.
7.4–10.0 wt.%. The high A/CNK ratios are due to the very high Al2O3 contents of 34 to 36
wt.% in the glass. The A.I. of 0.02 is very low.
Type III glass has a medium A/CNK index of 3.5 to 7.2 and high contents of K2O (0.6–4.1
wt.%) when compared to other varieties of glass. It significantly differs from the other glass
categories by its substantially higher FeO, MgO, and TiO2 contents (Table 8). The A.I. index
of 0.02 to 0.2 is low in this material.
Iron-titanium oxides
Titanian magnetite has moderate Ti, Al, and Mg contents (Table 10). The high content of the
ulvöspinel molecule (Usp = 43 mol.%) is typical for the microphenocrysts, lower contents are
present in the dusty dissemination of magnetite (Usp = 7 mol.%, Table 9). A high content of
MnO is typical for the titanian magnetite in the rim and the groundmass. The titanian
magnetite clusters along the magnetite – ulvöspinel solid solution series (Fig. 7) unlike more
oxidized magnetite of the hematite – ilmenite series in other phonolites, e.g., Col de Guéry
area, Massif Central (Bernth et al., 2002).
Titanite
Microphenocrysts and phenocrysts of titanite are the characteristic and most ubiquitous
accessory phases of the tinguaites. Notable are the low Al and Fe contents. The contents of Zr
and Nb are usually explained by a coupled substitution of Ti4+ + Al3+ – Zr4+ + Fe3+. Titanite in
the porphyritic tinguaite/phonolite PH-87 shows in some cases lower Zr concentrations (0.1–
0.6 wt.% ZrO2) than those in the porphyritic samples PH-88 and -89 (0.5–1.7 wt.% ZrO2).
Apatite
Patchily-zoned apatite is a subordinate component in the tinguaites. It is a fluorapatite
characterized by low Sr and REE contents.
Perrierite and Zr-silicate
Perrierite
with
the
empirical
formula
(Th0.03La1.47Ce1.92Y0.09Ca0.63Na0.05K0.02)4.21
(Fe2+1.72Mg0.28)2.00(Ti2.57Al0.57)3.14 [Si4.02O22] was first identified in the tinguaite sample PH3/1 (Ulrych and Pivec, 2001). It forms tiny, partly elongated grains about 0.05 mm in
diameter or short columns with indications of prismatic terminations. It has a strong yellow10
brown and red-brown pleochroism. Perrierite (Table 10) mostly forms inclusions in sodalite
or penetrates the feldspathic groundmass.
A Zr-silicate, as yet not clearly identified, occurs in porphyritic sample PH-3A (Ulrych
and Pivec, 2001). It forms small brown-red grains (50 to 100 m) and is associated with
sodalite commonly occupying the wedges between the laths of feldspars.
Chemical composition of the tinguaite dykes and host-rock samples
Bulk and trace element chemistry
Major element and trace element data of five tinguaite dyke and four host-rock samples are
presented in Table 4. The latter include a sample from the vent breccia that hosts dyke sample
PH-3A and a phonolite that hosts dyke sample PH-3/1. Additional chemical analyses of
phonolitic and trachytic dyke rocks were published by Hibsch (1926), Kopecký (1987), and
Souček et al. (1985).
In Fig. 8 the concentrations of Na- and K-oxides and SiO2 are plotted in the TAS diagram
of Le Maitre (2002). It can be seen that the dykes, as well as their host rocks, have phonolitic
compositions or straddle the trachyte and tephriphonolite field boundaries.
The tinguaite as well as the host rock samples have low Mg-numbers 32 to 8 with the
lowest Mg-numbers of ca. 8 confined to the host phonolites. The chemically evolved
composition is supported by low Ni and Cr concentrations and very high Al2O3 concentrations
of about 20 wt.%. These features are consistent with substantial fractionation of olivine and
pyroxene but with little or no feldspar fractionation (Table 4). The subordinate role of feldspar
fractionation in the dyke samples is confirmed by the high Sr concentrations and the absence
of Eu-anomalies in the samples (Fig. 10). However, in the host-rock samples PH-2 and S-55,
the low Sr and Ba concentrations provide some evidence for feldspar fractionation. In
particular Sample PH-2 underwent feldspar fractionation as can be inferred from the distinctly
negative Eu- and Ba-anomalies (Fig. 10c).
The catanorms calculated for the set of tinguaitic dyke rocks plot around the lower
boundary of the low-temperature trough of the petrogeny’s residua system (Bowen, 1937),
lying either below (tinguaites and porphyritic tinguaites) or above (phonolites, trachytes and
recrystallized porphyritic tinguaite) this boundary, see Fig. 9. Shand (1922) denoted the Na2O
+ K2O/Al2O3 ratio (in molar proportions) as the alkalinity index (A.I.) and distinguished (i)
agpaitic (peralkaline), (ii) miaskitic (alkaline) and (iii) plumasitic (subalkaline) magmatic
rocks. The terms sensu Le Maitre et al. (2002) are in parentheses. The values of the alkalinity
11
index A.I. (0.85–0.96) relate the tinguaitic dykes to the alkaline types whereas only the host
phonolite PH-2 and dyke S-55 fall into the peralkaline, and trachyte breccia POL-235 to
subalkaline categories. The alkaline type is the most frequent (59%) among the trachytic –
phonolitic rocks in the CS, whereas peralkaline (24 %) and subalkaline (17 %) types are
minor (Pazdernik, 1994; 1998). The textures and incompatible trace element contents relate
tinguaite dyke rocks to Le Bas’ (1987) Group II, representing the evolved, low-pressure
fractional crystallisation products of an olivine-poor nephelinite magma. These phonolites are
porphyritic with high Sr, Ba and low Rb, Nb contents.
In comparison with the average compositions of the Cenozoic phonolites from the BM
(Shrbený, 1995), the tinguaite dykes have substantially higher Sr, Ba and Cr contents but
lower Rb, Y, U, Th, Zr, Hf, Nb and Ta contents. The Th/U ratio (4–8) is higher than the
average ratio of phonolites of the BM (Th/U = 3.8 – Shrbený, 1995). The higher Th/U ratio in
the tinguaite dykes may be due to partial loss of the U as a result of post-magmatic activity of
hydrothermal fluids. The contents of REE in the tinguaite dyke rocks are moderate (230–
330), comparable, e.g., with values of REE (290) for the average phonolite of the BM
(Shrbený, 1995).
Normalized REE and other incompatible trace element patterns
The chondrite-normalized REE and trace element patterns of the dyke and host rock samples
are very similar (Fig. 10): all are characterized by remarkably steep LREE patterns and, in
most samples, by unfractionated HREE patterns. These traits are best developed in the dyke
samples whereas the host rocks show some irregularities in the middle and the heavy REE,
probably due to greater degrees of crystal fractionation, as exhibited by sample PH-2. The
REE patterns of the dykes overlap and thus indicate similar partial melting and fractional
crystallization histories. It is notable that the unfractionated HREE patterns suggest a garnetfree mantle source in contrast to that of the OIB-like basalts of the Cenozoic Volcanic
Province in Europe (e.g., Wilson and Downes, 1991). The normalized incompatible trace
element patterns exhibit a number of interesting characteristics such as high and OIB-like Nbconcentrations suggesting a genetic relationship with the Cenozoic European intra-plate
volcanism. In addition negative P- and Ti-anomalies underscore the important role of apatite
and titanian magnetite during melt fractionation. A subordinate role of feldspars is indicated
by a small and negative Ba-anomaly.
An unusual positive Zr anomaly may be due to the high solubility of alkali zirconium
silicate complexes in strongly alkaline melts (Watson, 1979; Ulrych et al., 1992) although
12
post-magmatic addition of zirconium from hydrothermal solutions can not be precluded
(Ulrych and Pivec, 1997). In general, crystallization of rare Zr, REE, Ti minerals is a
ubiquitous feature of the melt evolution of the tinguaite dyke rocks of the RIC. The
normalized trace element patterns of the host rocks of the dykes exhibit often lower
concentrations of the middle REE than in the dyke samples and in some samples there is an
irregular behaviour of the HREE with depletion and enrichment of Lu. The extended trace
element patterns of the host-rock samples are generally similar to those of the dykes,
consistent with their having experienced similar melt evolutions. The negative Nb-anomalies
in phonolite samples PH-2 and PH-3 may be explained with involvement of old subductionmodified upper mantle sources or melt contamination with felsic crust. The Zr/Hf and Nb/Ta
ratios of tinguaite dyke rocks are slightly higher than that published for average phonolite of
the BM (Shrbený, 1995).
The chondrite-normalized REE patterns show a typical U-shaped form (Fig. 10a). The
low abundance of MREE may account for the fractionation of amphibole and titanite (Wilson
et al., 1995). No expressive Eu anomaly is present in the tinguaite dyke differentiates (Eu/Eu*
= 0.8–1.3).
Sr and Nd isotope characteristics
The initial
87
Sr/86Sr and
143
Nd/144Nd isotopic ratios of four tinguaite samples and one host
rock sample (PH-3) are presented in Table 11 and plotted in Fig. 11. The initial
87
Sr/86Sr
ratios range from ~0.7033 to 0.7042 and the initial Nd values from +2.1 to +4.3. Two dyke
samples yielded Nd values of ~4.3 indistinguishable from the value for the host rock of the
dykes. Two additional dyke samples have lower Nd values of +2.1 and +3.5 indicating some
isotopic heterogeneity among the sources of the dykes. It is noteworthy that the initial
Sr/86Sr ratios of the samples with similar Nd values show a large variation in their 87Sr/86Sr
87
ratios of 0.7033 to 0.7036 (Fig. 11; Table 11). We suggest that the Sr isotopic heterogeneity
may be due to resetting of the Rb-Sr system during rock alteration, and consequently
calculation of unrealistic initial
87
Sr/86Sr ratios rather than invoking a heterogeneous mantle
source with respect to Sr. Considering that the sample with the highest Sr concentration is
least affected by secondary processes we accept sample PH-88A with an initial 87Sr/86Sr ratio
of 0.70362 as best reflecting the Sr isotopic composition of the three samples with Nd values
of ~+4.2. On the other hand the initial 87Sr/86Sr ratio for sample PH-3 with the highest Rb/Sr
ratio and low Sr concentration is easiest affected by modification of the Rb-Sr system.
13
The initial Nd values for the tinguaite dykes and their host rocks suggest a cogenetic
origin ultimately from moderately depleted mantle sources, similar to those of Cenozoic
basalts from Western Europe (Wörner et al., 1986; Wilson and Downes, 1991; Hegner et al.,
1995; Wilson and Paterson, 2002; Blusztajn and Hegner, 2002). When comparing the
143
Nd/144Nd and 87Sr/86Sr isotopic ratios of the dyke and host-rock samples with data for mafic
Cenozoic volcanic rocks from the Bohemian Massif (84Sr/86Sr = 0.7032–0.7037 and
143
Nd/144Nd = 0.51278–0.51288; Wilson et al. 1994 and Bendl et al. 1993) and basanite
formation from the České středohoří Mts. (Ulrych et al. 2002), see Fig. 11, our data plot
somewhat to the left of the mantle array (Fig. 11). This position of data points has previously
been interpreted as evidence for the involvement of HIMU mantle plume sources (e.g.,
Wilson and Downes 1991). However, as pointed out above, for sample PH-3 plotting to the
left of the mantle array, we need to invoke resetting of the Rb-Sr system. Thus all reliable data
points plot in the mantle array suggesting that typical HIMU-like mantle sources were
unimportant in the origin of the rocks.
Isotopic data for the Roztoky Intrusive Centre subvolcanic rocks show affinity to the
trachybasalt volcanism of the České středohoří Mts. (Fig. 11) complicated by a substantial
scatter associated with a secondary alteration process (Ulrych et al. 2000). The tinguaitic
rocks have higher
143
Nd/144Nd ratios and much lower
87
Sr/86Sr isotopic ratios those of the
České středohoří Mts trachybasalt formation, (Ulrych et al., of 2002). This might suggest an
independent source. It rules out an origin by crustal anatexis for the parental magma of the
RIC tinguaites but allows for minor crustal contamination of mantle-derived parental magma.
A similar source for tinguaitic dykes (87Sr/86Sr = 0.70314,
143
Nd/144Nd = 0.51203) in the Fen
Complex was suggested by Andersen and Sundvoll (1987).
Some of the phonolite intrusions in the České středohoří Mts. are characterized by less
radiogenic Nd and more radiogenic Sr isotopic ratios compared to those of mafic samples
from the same volcanic area. This relationship rules out extensive crustal contamination in the
genesis of the felsic magmas. In the Sr-Nd isotopic diagram, the tinguaite samples plot within
the mantle ~ 2 to 4 epsilon units above the bulk Earth composition (cf. data on felsic rocks
and alkali lamprophyres from the České středohoří Mts. – Fediuk, 1995; Ulrych et al., 2000,
2002). In terms of their Sr-Nd isotopic characteristics, the phonolite magmas in the CS
probably represent low-pressure differentiation products of mantle-derived magma, with some
crustal modification whereas the alkaline lamprophyres of the RIC are uncontaminated nearprimary products from the same magma source.
14
Discussion
The origin of the phonolitic and trachytic magmas is attributed to tapping of different levels
within a compositionally stratified magma chamber and/or to melting of upper mantle or deep
crust (e.g., Lippard, 1973; Price et al., 1985; Wilson et al., 1995). The rock series of the RIC
may be derived from a primitive olivine-poor nephelinite magma by means of progressive
tapping of a stratified magma chamber within the crust that was replenished on several
occasions (Ulrych, 1998). The phonolitic dykes from the RIC plot in the low-temperature
trough of Bowen´s (1937) residua system, and as the H2O content is relatively high (inferred
from the presence of amphibole), an origin from a magma chamber in the deep crust is most
plausible (Bailey, 1974; Harris, 1974). As suggested by the high Sr and Nb contents in the
phonolite/tinguaite dykes, their origin may well be associated with a fractionating mantlederived
basanitic
parental
magma
(Borodin,
1989).
Crystallization
of
a
(clinopyroxene)/amphibole  (olivine, plagioclase), titanite, apatite assemblage can account
for the composition of both dyke series of the RIC (Ulrych, 1998; Ulrych and Balogh, 2000).
From a geological point of view, the tinguaites represent the youngest dykes of the RIC
(sensu
Hibsch,
1936).
They
are
geochemically
associated
with
the
alkaline
monchiquite/camptonite dyke series of the RIC (sensu Ulrych et al., 1998). The ages of the
dykes are close to those of the other subvolcanic rocks of the RIC (Ulrych et al. 1999, 2000);
the detailed chronology of the dyke rocks cannot be further refined by geochronological
methods. The roles of excess Ar and its retention in nepheline/sodalite are substantial and may
account for the spread in K-Ar ages (cf. Balogh et al., 1999).
The sources of the tinguaites are characterized by less radiogenic Nd and partly by more
radiogenic Sr than those for mafic rocks from the BM. This relationship suggests that
extensive crustal contamination was not involved in formation of the felsic magmas. A low
Nd value of +2.1 in sample PH–87 (SiO2 ~58 wt.%) when compared to values of +4.2 in the
more primitive rocks (sample 88, SiO2 ~53 wt.%) may be taken as evidence for the effects of
melt contamination during formation of the felsic rocks in a deep crustal magma chamber.
However, involvement of isotopically different mantle-derived melt batches could also
explain the isotopic heterogeneity among the samples. If a single fractionating magma
chamber is hypothesized, however, the isotopic variation among the samples could be
explained with assimilation-fractional crystallization (AFC) processes. In the Sr-Nd isotope
diagram, the samples plot within the mantle array together with data for the camptonite dyke
15
from the RIC. Therefore, in terms of their Sr-Nd isotopic characteristics, the phonolite
magmas in the CS appear to represent products of a mantle-derived magma that was only
partly influenced by crustal contamination, whereas the comagmatic alkaline lamprophyres
may have been unaffected by crustal assimilation.
Tinguaite rocks associated with carbonatites are known from the Kaiserstuhl Volcanic
Centre and the Katzenbuckel Complex in the Cenozoic Central European Volcanic Province
(Stahle and Koch, 2003). Based on the homogenization temperatures of inclusions in
clinopyroxenes from the tinguaites, Panina et al. (2000) inferred a wide temperature range for
their crystallisation of 1140 to 900o C. Tinguaite dykes (550 ± 7 Ma) are also associated with
the Fen Complex in Norway (BergstØl, 1979) although at this locality they occur exclusively
outside the main complex. This relationship led to the interpretation that the tinguaite
intrusion represented an early event in the evolution of carbonatite and peralkaline complexes.
However, a Sr-Nd geochemical study of Andersen and Sundvoll (1986) showed that the
parent magma of tinguaites, originating from a depleted mantle source with a minor crustal
contamination, evolved from a common ijolitic precursor by fractional crystallization in the
shallow crust. In contrast to these evolutionary scenarios for tinguaites, alkaline rocks of the
olivine-free nephelinite – phonolite/tinguaite association from the Gujarat Province, India
were interpreted as products of liquid immiscibility of carbonated nephelinitic mantle magma
(Viladkar and Avasia, 1994; Shrivastava, 1994).
As suggested by the Sr-Nd isotopes of the samples of this study, the parental magma of
the tinguaites clearly originated from a depleted mantle source and an origin by anatexis of an
old mafic lower crust is not supported by the isotopic data.
Conclusions
The RIC tinguaite dykes are the most evolved members of a strongly alkaline
(camptonite/monchiquite – phonolite/tinguaite) dyke series. The K-Ar ages of these dykes
range from 31 to 28.5 Ma and overlap the ages of 33 to 24 Ma determined for the other
subvolcanic members of the RIC. We note that ubiquitous excess argon, preferentially located
in nepheline and/or sodalite, may explain some of the K-Ar results for the tinguaite dykes that
are older than those for the country rock that they intrude.
Specific textural varieties of the tinguaites, ranging from aphyric tinguaite to porphyritic
tinguaite, sometimes recrystallized with relict tinguaite-textured margins, occur in the central
part of the RIC. Some of the disequilibrium textures observed, e.g., the peculiar “cell-shaped”
16
arrangement of alkali feldspar, foids and clinopyroxene microlites and the felty-to-prismatic
groundmass textures, are interpreted as due to prolonged metastable crystallization in the
presence of volatiles. Microlithic fillings of the equidimensional “cells” are composed of
relict glass and recrystallized ternary feldspars, analcime, and amphiboles, in the order of an
agpaitic succession. The presence of either a tinguaitic or trachytic texture in the phonolitic
dykes probably depended on the rate of magma cooling. This may be inferred from the
observation that thin phonolitic dykes show a trachytic texture whereas thick dykes show a
tinguaitic texture, or at least glassy tinguaitic margins, and a trachytic texture in central parts
of the dyke disequilibrium textures, e.g., peculiar “cell-shaped” arrangement and felty-toprismatic groundmass. Microlithic fillings of the equidimensional “cells” are composed of
relict glass and recrystalised ternary feldspars, analcime and amphiboles, in the order of an
agpaitic succession. Whether a tinguaitic or a trachytic texture resulted in the dyke rocks was
probably dependent on the rate of cooling. This can be inferred from the observation that
thick phonolitic dykes have a trachytic texture – at least in their central parts – and sometimes
with glassy tinguaitic margins, whereas thin dykes show a tinguaitic texture.
Crystallization of the tinguaites proceeded primarily in three stages: a) an early
magmatic stage involving crystallisation of (micro)phenocrysts of (Na-K)-feldspar, Koligoclase to K-andesine, nepheline >> diopside, kaersutite/pargasite, sodalite >> titanian
magnetite, titanite, apatite, b) a main magmatic stage that saw the crystallization and
recrystallization of groundmass consisting of relict glass, (Na-K)-feldspar, K-oligoclase,
diopside to aegirine augite > titanian magnetite, and c) a late magmatic stage characterized by
hydrous minerals such as analcime, clay minerals, and natrolite formed by the decomposition
of earlier minerals due to residual fluids rich in Zr-REE-Ti (perrierite and an unknown Zrsilicate).
We conclude from the geochemical evidence, including the Sr-Nd isotopic data, that the
magmas parental to the tinguaite dykes were ultimately derived from depleted (possibly
lithospheric) mantle. Formation of the evolved rock types took place in shallow-level magma
chamber, with limited crustal assimilation.
Acknowledegements
This research was supported by the Grant Agency of the Academy of Sciences of the Czech
Republic A3048201 within the Research Programme of the Institute of Geology, CEZ:
AV0Z30130516. K-Ar dating was supported by OTKA projects No. T043344 and M41434 to
17
K. Balogh. The manuscript benefits from the comments and criticism by F. Fediuk, Praha, and
B.G.J. Upton, Edinburgh. In addition, our special thanks go to B.G.J. Upton for the revision
of the English text. We are indebted to A. Langrová for microprobe analyses and J. Pavková
for technical assistance, both of the Institute of Geology AS CR, Praha.
Explanations to the figures
Fig. 1. A – Position of the Ohře/Eger Rift and the České středohoří Mts. in the Central
European Volcanic Province. Modified from Wimmenauer (1974). B – Geological sketch
map of the Cenozoic Roztoky Intrusive Centre in NW Bohemia. Information compiled
from Hibsch (1902) and Ulrych (1998). C – Central parts of the Roztoky Intrusive Centre
with location of sampled phonolitic/tinguaitic dykes.
Fig. 2 A. Back scattered electron image of:
A. Tinguaitic texture with typical “cells” in aphyritic fine-grained tinguaite sample (PH-3/1
from road cut at Skrytín). The cells are filled with alkali feldspar, analcime, and glass.
Clinopyroxene and sodalite occur as phenocrysts.
B. Porphyritic tinguaite/phonolite sample PH 87 from “Flur Ratschin” at Stará Homole
showing relics of a trachytic texture.
C. Porphyritic tinguaite sample PH-89 from Roztoky, in the Bradlo/Pradelberg railwaycutting. The image shows a Na-sanidine phenocryst rich in celsian, set in the pale
marginal zone containing up to 3.4 wt.% BaO.
D. Corroded phenocryst of anorthoclase with atoll texture filled with glass in mixed-textured
tinguaite/phonolite sample PH-87 from “Flur Ratschin” at Stará Homole
E. Corroded diopside phenocryst with atoll texture filled with glass and K-oligoclase in
marginal ring. Mixed-textured tinguaite/phonolite sample PH-87 from “Flur Ratschin” at
Stará Homole.
F. Pseudomorphed euhedral nepheline completely transformed to a mixture of analcime and
clay minerals. Porphyritic tinguaite sample PH-89, Roztoky, from the Bradlo/Pradelberg
railway-cutting.
Fig. 3. Ab – Or – An diagram after Smith (1974) showing compositions of feldspars.
Fig. 4. Quadrilateral diagram of Morimoto ed. (1988) showing the compositions of
clinopyroxenes. Symbols as in Fig. 3.
18
Fig. 5. Classification diagram of Leake ed. (1998) showing the composition of amphiboles.
Symbols as in Fig. 3.
Fig. 6. Ne – Ks – Qz diagram showing the composition of nepheline. The dashed line (Barth
join) corresponds to the solution of feldspar in nepheline at 1068 oC, 1 kbar, the dot-anddash line marks the limit of solid solutions at 700 oC, 1 kbar pH2O (Hamilton, 1961).
Symbols as in Fig. 3.
Fig. 7. FeO – Fe2O3 – TiO2 diagram showing the composition of Fe-Ti oxides. Symbols as in
Fig. 3.
Fig. 8. Total alkali-silica diagram (Le Maitre ed., 2002) showing the composition of Cenozoic
tinguaite dyke samples from the Roztoky Intrusive Centre, NW Bohemia. Small solid
circles represent data for phonolitic and trachytic rocks s.l. from the RIC (data sources
Souček et al., 1985 and Kopecký, 1987). Symbols as in Fig. 3.
Fig. 9. Diagram of the Q – Ne – Kp residual system showing the composition of the tinguaitic
samples. The low-temperature trough is indicated after Bowen (1937). Symbols as in Fig.
3.
Fig.10. (a) Chondrite-normalized REE patterns, (b) Primitive mantle-normalized incompatible
trace element patterns for the tinguaite dykes samples of the Roztoky Instrusive Centre.
Symbols as in Fig. 3. Normalizing values from Sun and McDonough (1989).
Fig. 11. Initial
87
Sr/86Sr and
143
Nd/144Nd isotopic ratios for tinguaitic dyke samples of the
Roztoky Intrusive Centre. The lines encompass the mantle array. Symbols as in Fig. 3.
For comparison, isotopic data for Cenozoic volcanic rocks of the Bohemian Massif are
plotted: Mafic Cenozoic volcanic rocks from the Bohemian Massif (grey field “Mafic
rocks BM”; Wilson et al., 1994), basanites (black dots) and trachybasalts (crosses) from
the České středohoří Mts. (Ulrych et al., 2002), “MD” monzodiorite, “C” camptonite,
“G” gauteite (maenaite) from the Roztoky Volcanic Centre, České středohoří Mts., “P”
phonolites from the České středohoří Mts. (Ulrych et al., 2000).
Tables
Table 1 Geological, petrographical, and mineralogical characteristics of tinguaite dykes and
their host rocks from the Roztoky Intrusive Centre, NW Bohemia.
Table 2 K-Ar ages of tinguaite dyke samples from the Roztoky Intrusive Centre.
Table 3 Modal analyses of tinguaite dyke samples from the Roztoky Intrusive Centre.
Table 4 Table 4 Chemical analyses of Cenozoic tinguaite dyke samples and their host rocks
from the Roztoky Intrusive Centre, NW Bohemia
19
Table 5 Representative chemical analyses of feldspar in tinguaite dykes from the Roztoky
Intrusive Centre. The element concentrations are given in wt.% and the number of ions per
formula unit.
Table 6 Representative chemical analyses of clinopyroxene in tinguaite dyke samples from
the Roztoky Intrusive Centre. The element concentrations are given in wt.% and the number
of ions per formula unit.
Table 7 Representative chemical analyses of amphibole in tinguaite dyke samples from the
Roztoky Intrusive Centre. The element concentrations are given in wt.% and the number of
ions per formula unit.
Table 8 Representative chemical analyses of feldspathoid in tinguaite dyke samples from the
Roztoky Intrusive Centre. The element concentrations are given in wt.% and the number of
ions per formula unit.
Table 9 Representative chemical analyses of glass in tinguaite dyke samples from the Roztoky
Intrusive Centre. The element concentrations are given in wt.%.
Table 10 Representative chemical analyses of Fe-Ti oxide in tinguaite dyke samples from the
Roztoky Intrusive Centre. The element concentrations are given in wt.% and the number of
ions per formula unit.
Table 11 Rb-Sr and Sm-Nd isotopic data for tinguaite dyke samples from the Roztoky
Intrusive Centre.
20