I. Introduction

The effect of contact between mantle or cumulate xenoliths and basaltic melts are polycrystalline selvages around the xenolith and diffusion profiles in the xenolith minerals.
The Mg-Fe interdiffusion in olivine allowed calculations of residence time of xenoliths in the host magma thereby also the time scales of petrogenetic events such as magma mixing. In this study we have measured compositional profiles to calculate diffusion time in two olivine grains.
The investigated sample E-41 is an ultramafic cumulate xenolith from a basanite dyke of Rothenberg (Fig. 1). The host magma of this xenolith is a basanite which contains phenocrysts of green pyroxene, olivine, phlogopite, and microphenocrysts of magnetite and leucite.
In one olivine we have detected no diffusion profile. The determined diffusion zone in the second olivine in this sample is characterised by increasing forsterite content in 0,06 mm wide diffusion profile.


II. Sample description

The lherzolite type cumulate xenolith is up to 20 mm in size, is rounded and characterized by occurrence of a polycrystalline selvage.



Fig. 1. Xenolith and host magma in thin section.

It consists of ~65 vol.% olivine, up to 20 vol.% clinopyroxene, 15 vol.% orthopyroxene and rare little grains of opaque Ti-magnetite (1 vol.%).
Crystal size is < 1 mm. (Fig. 2)



Fig. 2: ol, opx, magnetite inclusions (black grains) in plane polarized light in thin section.

The paragenesis of magnetite and Mg-poor olivine (Fo81), absence of Cr-Spinel and its texture indicate that the xenolith is possibly a fragment of cumulate crystals already formed in a magma chamber and ripped up during magma movement.

The anhedral olivine grains often show intracrystalline deformation bands; they are rich in fluid inclusions and sometimes magnetite inclusions. In the centre and along the rim of the xenolith, rounded little olivine grains are poikilitically included in large pyroxenes. They do not show any euhedral crystal shape. (Fig. 3)


A) B)

Fig. 3: Thin section images of E-41 xenotlith show poikilitic structure in (A) the centre and (B) the rim.

The xenolith is surrounded by an irregular, about 1 mm wide, polycrystalline selvage. It consists mostly of clinopyroxene and in addition rhoenite, phlogopite, magnetite, apatite, plagioclase. This very fine grained mineral assemblage is visible as black and discontinuous rim in thin section (Fig.1+4).


A) B)

Fig. 4: Microscopic pictures of (A) selvage cpx and (B) clinopyrxene+rhoenite+phlogopite+magnetite +apatite+plagioclase in thin section in plane polarized light.

The selvage clinopyroxenes are anhedral to subhedral. Large cpx crystals, independently grown in melt, show well developed crystal faces and chemical zonation (Fig. 5). The selvage cpx is Ca-rich with low concentrations of Al, Ti, Mg, Fe (compare Fig. 6).



Fig. 5: Selvage clinopyroxene in thin section, polarized light.

Fig. 6 shows the element distribution maps of selvage. The contact between xenolith olivine and selvage is either sharp or it shows poikilitic crystal growth: fine grained olivine in cpx or cpx-rhoenite matrix or irregular cpx accumulation (Fig. 6).

Fig. 6: Element distribution maps of polycrystalline selvage and olivine of the xenolith.


III. Diffusion in olivines

To calculate contact time between xenolith and host magma before eruption we have measured the compositional gradient in two olivine grains at the reaction selvage. Element concentrations were measured with JEOL JXA 8900 R electron microprobe. Analysing step size was 10 µm.
Variation of Mg and Fe concentration in both olivine crystals is 1 wt%. One cannot recognize definite diffusion profiles (compare Fig. 7 + 9).

1. E-41 Olivin 1

The compositional gradient shows the expected trend: magnesium content decreases towards the rim. Average difference between core and rim is only about 0,5 wt. %.

Fig.7: Compositional profile of Mg and Fe in olivine 1.

Fig. 7 shows the element profile of olivine 1 for Mg and Fe. The irregular trend does not define a diffusion profile. Thus, there is no indication to define the distance of a compositional gradient or a time the diffusion. Fig. 8 shows BSE-image of the analysed olivine. The picture indicates, that the profile did not reached the crystal rim.



Fig.8: BSE-image of olivine 1.

2. E-41 Olivin 2

The change in element content in the second olivine is demonstrated in Fig.9.

The profile does not show any obvious diffusive element concentration gradient, except, possibly for the outermost rim.

Such constant concentration profiles without diffusion also were detected in xenoliths from different localities in the Eifel, e.g. Baarley, by Shaw, 2004.



Fig.9: Compositional profile of Mg and Fe in olivine 2.

The Mg concentration in olivine 2 decreases slightly (1,5 wt%) from core to rim. Fe concentration increases about 0,8 wt.

Within a distance of 0,06 mm at the rim, the Mg content increases about 1,08 wt.% and Fe content decreases about 1,01 wt.%. The values lie within the error of 0,5 wt.%. The points near to the selvage may additionally be influenced by secondary fluorescence during measurement.

Shaw (2004) interpreted the almost constant forsterite profile with low Fo-concentrations compared to peridotite-xenoliths as caused by a complete re-equilibration to lower Fo contents.

The measured forsterite content of Fo81 in olivine 2 is typical for cumulates. Thus, the constant Fo-profile of Fo80-Fo82 is not the result of equilibration.

Fig. 10 shows that the Fo-content slightly but continuously increases outwards in the rim. If the small gradient may be caused by diffusion, the calculated residence time is the minimal time the xenolith resided in the basaltic magma with most mafic composition. We cannot say if the gradient was completely preserved or if the rim was partly resorbed by mineral reaction.



Fig.10: Forsterite profile in olivine 2.

The analysed olivine crystal is surrounded by a reaction selvage (Fig. 11). The shape of the reaction selvage suggests that the selvage describes former crystals (green line in Fig. 11).



Fig.11: BSE-image of olivine 2 with measurement line (red) and selvage contour (green).

If reaction with the host magma was faster than diffusion one cannot detect visible diffusion. If diffusion occurred in this area, the width of the reaction selvage would correlate with the largest compositional gradient and thus with the maximal contact time between xenolith and host magma. A problem is the determination of the width of the reaction selvage width because it is irregular.


III. Results

For the calculation of the diffusion time for profile 2 we used DMg-Fe from Chakraborty (1997).

Table 1: Calculation of contact time for increasing Fo content in olivine 2.

The contact time between xenolith and host magma from the Rothenberg dyke is assumed to be 232 days (minimum).

The small width of the diffusion zone may be caused by only short residence time between entrainment of the xenolith and eruption in the magma chamber prior the eruption or the zone is only a fragment of a larger compositional gradient being lost by dissolution and mineral reaction processes.