Use this URL to cite or link to this record in EThOS:
Title: Geochemical aspects of the mesostasis of some fractionated igneous rocks
Author: Henderson, Paul
Awarding Body: University of Oxford
Current Institution: University of Oxford
Date of Award: 1966
Availability of Full Text:
Access from EThOS:
Full text unavailable from EThOS. Please try the link below.
Access from Institution:
Many layered igneous rocks consist of two main parts: the cumulus (plus adcumulus and heteradcumulus) material and the pore material (or 'mesostasis') produced by the crystallisation of the trapped liquid which necessarily had the composition of the contemporary magma. Part One of this thesis is primarily concerned with the geochemistry of the mesostasis of some layered igneous cumulates. In the past no workable method, has been developed for deducing the amount or composition of the mesostasis, for igneous cumulates in general. Wager (1963) established a method for the Skaergaard intrusion, Hast Greenland, by assuming that all the phosphorus in a rock of the Lower and Middle Zones (i.e. before the incoming of cumulus apatite) of the Layered Series existed in the mesostasis, so that the phosphorus content of any rock gave a relative measure of the amount of mesostasis. Furthermore, the phosphorus contents of the successive fractions of the Skaergaard magma were known from another approach and so it was possible to deduce the absolute amount of mesostasis. For most other layered intrusions the successive compositions of the fractionated magma are not known and the method cannot be applied to them. knowledge about the geochemistry of the mesostasis (and hence the contemporary magma) would be most useful in straight comparative geochemical studies; in deducing fractionation trends; and in the determination of solid/liquid trace element partition coefficients, which in turn could be useful in considering the petrogenesis of certain other fractionated igneous rocks. Thus, an attempt has been made to produce a geochemical method for deducing the amount of mesostasis in igneous cumulates. For this method, elements of contrasting geochemical behaviour are required: those that show a strong preference to remain in the residual liquid of a fractionating basic magma (the so called. 'low-k elements') and those which tend to preferentially enter one of the cumulus phases (the 'high-k elements'). Two methods are considered: the first uses the existence of rhythmic layering and is applicable, without adaptation, only to adcumulates, The concentration of a high-k element in a rock can be related to the amount and composition of the mesostasis in that rock by the equation: pN + (1 - p)C = T, where p is the proportion of the mesostasis; M, C, and T are the concentrations of the high-k element in the mesostasis, the overall cumulus assemblage, and the whole rock respectively. If this rock is taken from a band of rhythmic layering then another rock can be taken close to the first one but with very different proportions of the minerals in the cumulus assemblage. An equation of the same form can be drawn up for this rock. It is shown that it is possible to find C and T by direct measurement for each rock. The equations can then be solved simultaneously for p and M by using the low-k element concentrations in the two rocks to establish the relative amounts of the mesostasis. A second method is also discussed whereby the amount and composition of the mesostasis might be determined by analysing rocks from along the strike of the layering provided there were a variation in the amount of mesostasis in the rocks. Error analyses of both methods are presented. The main section of Part One is an investigation into the suitability of certain elements as indicator of the relative amounts of mesostasis in igneous cumulates. In a general discussion it is shown that As, Cs halogens, Hf , P, U, and Zr, are probably suitable low-k elements. The distribution of phosphorus in the minerals and rocks of the Skaergaard, Bushveld and Rhum igneous intrusions has been investigated using activation analysis. It is shown that in the Lower Zone of the Skaergaard intrusion most of the phosphorus resides in the mesostasis. The phosphorus content of the minerals from adcumulates is taken to be that of the cumulus phases and it is shown that the order of entry of phosphorus is: olivines (about 30 ppm.); felspars (about 50 ppm.); and clinopyroxenes (about 30 ppm.). Phosphorus concentrations of cumulus magnetite and ilmenite are low (about 9 and 20 ppm. respectively). However, the amounts of phosphorus in the minerals of mesocumulates or orthocumulates are generally more variable and much higher (e.g. in plagioclases up to 1,400 ppm. has been recorded) than those of adcumulates. These relatively high amounts are interpreted as arising from crystallisation of the trapped liquid to give zoning to the appropriate minerals which is enriched in phosphorus because of the ever increasing phosphorus concentrations of the remaining liquid as the trapped liquid fractionates. Compared to the Lower Zone rocks the amount of phosphorus in the cumulus minerals is small. Wager's method is critically reviewed in the light of these results. The distribution of phosphorus in the Bushveld igneous intrusion (before the incoming of cumulus apatite) shows that many of the rock types are adcumulates as a large amount of the phosphorus in a rock can be accounted for by the amount in the minerals. The amount of phosphorus in the rocks does not show a steady increase with height as might have bean expected . The values fluctuate between 9 and 200 ppm. until the horizon where cumulus apatite enters and the value rises to 14,0O0 ppm. The minerals tend to have slightly less phosphorus than the Skaergaard cumulus minerals though the order of entry is the same: olivine (about 40 ppm.); felspar (about 35 pp.); pyroxenes (about 20 ppm.). The Rhum rocks generally have vary low phosphorus concentrations and of the units studied, those of 7, 8 and 9 appear to consist of adcumulates. The allivalite of unit 3 has considerably more phosphorus than the other units and this implies that it has more mesostasis. This confirms the work of Brown, (1956). The order entry of phosphorus into the Rhum minerals is the same as those of the Skaergaard and Bushveld cumulus phases. This order is not the same as that found by Koritnig (1965) who analysed similar minerals for phosphorus from a variety of environments and found that most was in the olivines and least in the felspars. Some olivines from olivine nodules have phosphorus concentrations very similar to those from the layered intrusion. Determinations of bromine in some Bushveld rocks show no coherence with the phosphorus values. An investigation into uranium, also considered to be a low-k: element, has been carried out. It is shown that in the Skaergaard Lower Zone rocks there is a strong coherence between phosphorus uranium irrespective of rock type. This implies that most of the uranium resides in the mesostasis and is confirmed by the fact that almost a negligible quantity of uranium is found in the cumulus minerals though a small amount is found in the clinopyroxenes. Like phosphorus, uranium shows a relative enrichment in those minerals which have partly crystallised from the trapped liquid. A similar coherence has been found between uranium and phosphorus in the Rhum rocks and similar amounts were found in the Rhum cumulus minerals as in the Skaergaard ones. However little coherence was noted between these elements in the Bushveld rocks and this is considered to be because of the complexities of the history of fractionation of the Bushveld magma. Amounts of uranium in the Bushveld minerals are much the same as for the cumulus mesostasis of the Skaergaard and Rhum intrusions. Data on strontium (considered to be a high-k element) is presented for some of the Skaergaard and Rhum rocks and minerals. The content of the Skaergaard Lower Zone felspars indicate that the contemporary magma contained less than 480 ppm. Sr. The Rhum felspars are interesting in that those from the allivalites have significantly lower strontium contents than those from the peridotites In the two units studied. The phosphorus, uranium and stronium data are used in applying the theoretical method to the Skaergaard Lower Zone and Rhum rocks. Determinations of specific gravities and, modes of these rocks are presented. It is shown that application of the method to mesocumulates and orthocumulates is possible if mineralogical evidence is used as an aid in determining the proportions of the different minerals that make up the cumulus assemblage. It is concluded that there is much variation in the amount of the mesostasis of the rocks of the Skaergaard Lower Zone; that the amount of strontium in the contemporary magma was of the order of 340 ppm. and of phosphorus, 2,000 ppm. The results for the Rhum rocks show that some of these are adcumulates whilst others may be regarded as mesocumulates. The application and failings of the second method, are briefly discussed. Tbe mineralogy of the rocks can be helpful in determining the relative proportions of the minerals in the cumulus assemblage. Zoning is particularly useful in this respect and in giving a lower limit to the amount of mesostasis in any rock. Some evidence of sintering of cumulus olivine crystals in the Skaergaard and Rhum melanocratic rocks is shown. The mineralogy of the mesostasis in the Skaergaard Lower Zone rocks is briefly discussed. In Part Two the determinative methods that have been developed and used are described. Phosphorus has been determined by activation analysis. the choice of final precipitate is discussed and magnesium ammonium phosphate decided upon. A previously published method using this precipitate has been tried but is shown to be inadequate for this work. Several adaptations to this method have been made and a full investigation into the suitability of the final precipitate is described in detail: this includes tests on self-absorption and stoichiometry. The method uses a colorimetric technique for determining the yield of the final precipitate. A method for the determination of bromine using activation analysis is described and criticised.
Supervisor: Not available Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID:  DOI: Not available