Pergamon Geochimica et Cosmochimica Acta, Vol. 59, No. 7, pp. 1339-1352, 1995 Copyright© 1995Elsevier Science Ltd Printedin the USA.All rightsreserved 0016-7037/95 $9.50 + .00 0016-7037(95)00048-8 Variations of oxygen and carbon isotopes in carbonatites: A study of Brazilian alkaline complexes ROBERTO V. SANTOS l,. and ROBERTN. CLAYTON 2 The University of Chicago, Department of the Geophysical Sciences, Chicago, IL 60637, USA 2The University of Chicago, Department of the Geophysical Sciences, Department of Chemistry, and Enrico Fermi Institute, Chicago, IL 60637, USA (Received February 28, 1994; accepted in revised form December 12, 1994) Abstract--We have studied the oxygen and carbon isotopic compositions of carbonatite complexes from South America (Jacupiranga, Araxfi, Catal~o, Tapira, and Mato Preto) and interpreted the results in terms of magmatism, contamination by country rocks, and hydrothermal processes. These complexes range in age from 130 to 65 Ma and were emplaced during the reactivation of the South American Platform during the Mesozoic Era. Except for the samples from Mato Preto (613C = -6.9 to +0.8%o), which have been contaminated by limestone from the country rock, there are no large carbon isotope differences among the samples of Jacupiranga ( -7.3 to -6.6%o), Araxfi ( -7.5 to -4.8%o), Catal~o ( -7.1 to -5.3%o), and Tapira (-6.8 to -4.8%0). In contrast, the carbonatites have a wide range in oxygen isotopic composition, which seems to be related to their degree of hydrothermal alteration and their emplacement level. For instance, while the samples from Jacupiranga have a narrow range of 6180 (6.6 to 7.3%0) and have not been exten- sively affected by fenitization, the carbonatites from the complexes of Araxfi ( 8.7 to 16.3%o), Catal~o (7.3 to 19.3%o), and Tapira (9.7 to 15.4%o) have a wide range in 6180 and are accompanied by pervasive potash-fenitization of their host rock. The potash-fenitization is marked by the replacement of carbonatite host rock (mainly pyroxenite) by carbonate-pblogopite-magnetite-rich rocks. We conclude that fractional crystallization and liquid immiscibility may not significantly affect the oxygen and carbon isotopic composition of carbonatites. Large isotopic variations may be produced when these rocks intrude limestone country rocks (i.e., Mato Preto) and/or have been subjected to postcrystal- lization alteration processes. Variations of 6180 and 613C in the carbonatites may be explained by isotopic exchange between these rocks and H20-CO2-fluids at different temperatures and with different H20/CO2 ratios. The isotope exchange model implies that the isotopic variations in carbonatites take place under low-temperature conditions (below 300°C) and involve fluids with high H20/CO2 ratios. INTRODUCTION Initial studies of the oxygen and carbon isotopic compositions of carbonatites suggested that they were isotopically homo- geneous, leading Taylor et al. ( 1967 ) to define a range in 6180 (6.0 to 8.5%0) and 6~3C (-5.1 to -7.3%o) characteristic of these rocks. Later studies (Pineau et al., 1973; Deines and Gold, 1973; Nelson et al., 1988; Censi et al., 1989; Deines, 1989; Santos et al., 1990) confirmed that many carbonatites have isotopic compositions within the above range, but also showed that a significant number of samples have high 6180 and 6 ~3C values. For instance, Deines and Gold ( 1973 ) stud- ied the oxygen and carbon isotopic composition of fourteen carbonatites and concluded that the variability in 6180 and 613C could be related to their emplacement level. They showed that compared to deep-seated complexes, shallower complexes have a wider range in isotopic composition, which they attributed to near-surface processes. In another study, Pineau et al. (1973) also reported 6180 and 6~3C from car- bonatites and divided them into three groups: ( 1 ) those which have 6 J80 and 613C values typical of mantle rocks, (2) those in which variations in 6180 are correlated with variations in 613C, and (3) those in which the variations in 6~80 are not correlated with variations in 6 ~3C. They argued that the vari- * Present address: Universidade Federal de Ouro Preto, Dept. de Geologia, 35400-000, Ouro Preto-MG, Brazil. ations in 6 tSo and 613C observed in groups 2 and 3 could be explained by "late magmatic and deuteric processes." Nelson et al. (1988) reported data on trace elements and Sr, Nd, O, and C isotopes from carbonatites worldwide and pointed out the difficulty in distinguishing between primary and second- ary variations of 6180 in carbonatites. Nevertheless, they ar- gued that oxygen and carbon isotope variations in carbonatites may be, in part, related to isotopic heterogeneity in the mantle. In a review of the oxygen and carbon isotopic composition of carbonatites, Deines (1989) suggested that some isotopic variations could be attributed either to magmatic processes related to the formation of carbonatite magma or the nature of the primary reservoir within the mantle. Most of the variations in 6180 from carbonatites have been explained by interaction involving hydrothermal fuids, whereas the variations in 6 ~3C have been explained both by primary (i.e., isotopic composition of carbonatite parental magma) and secondary (i.e., hydrothermal alteration) pro- cesses (Taylor et al., 1967; Pineau et al., 1973; Deines, 1989). Determining to what extent the 6180 and 6 ~3C variations in carbonatites are primary or related to crustal processes has important implications for evaluating the isotopic composi- tion of their mantle source region. Furthermore, it provides insight into questions concerning mantle isotopic heteroge- neity, recycling of crustal material into the mantle, and the relationship between the parental magma of carbonatite and other sorts of magmas derived from the mantle (e.g., OIB and 1339