TITANITE IN VOLCANIC ROCKS

Bart J. Kowallis, Eric H. Christiansen, Dana T. Griffen, and Nan Hu, Department of Geology, Brigham Young University, Provo, UT 84602

ABSTRACT

Introduction

Geologic Setting

Temporal Changes in Magmatism

Titanite Occurrence in Volcanic Rocks
Middle Jurassic
Late Cretaceous
Tertiary

Elemental Differences in Volcanic Titanite
Data Collection
General Trends
Jurassic Bentonites

Conclusions

References

Download Data Base in Quatro Pro

Download Data Base in Excel

Introduction.  Titanite from volcanic rocks is not well represented in published analyses in the literature and has not been extensively studied (Nakada 1991). Although titanite is a common accessory mineral in plutonic rocks, it is often described as uncommon or less common in volcanic rocks (Deer et al. 1997). Only a few papers have reported on titanite analyses in volcanic rocks, and many of these have been from studies of alkaline lavas (Fitton and Hughes 1977; Cundari, 1979; Keller 1981; Giannetti and Luhr 1983; Wörner et al. 1983; Weis et al. 1993; Luhr et al. 1984; Wolff 1984; Nono et al. 1994; Paslick et al. 1996; Ablay et al. 1998; Dawson and Hill 1998; Della Ventura et al. 1999). From calc-alkaline volcanic rock suites, titanite analyses have been reported from Quaternary dacites in the Andes (Nakada 1991; de Silva et al. 1994), the Tertiary Fish Canyon Tuff in Colorado (Whitney and Stormer 1985), and the Tertiary tuffs of the Lund Formation in western Utah and eastern Nevada (Maughan, 1996).

The assemblage quartz + magnetite + titanite is important in distinguishing relatively oxidized from relatively reduced conditions in granitic magmas (Wones 1989). The presence of titanite in rhyolitic and dacitic magmas may, therefore, assist in understanding the oxidation conditions present in the magma chamber prior to eruption. In addition, titanite has the ability to accept as substitutes a number of different elements into its basic formula. These minor elements also record varying conditions in the magma chamber (Fleischer and Altschuler 1969; Fleischer 1978; Kowallis et al. 1997) and may aid in trying to understand the evolution of a magma system. In this paper, we want to focus on the changes in composition and abundance of titanite from subduction-related volcanic rocks in samples ranging in age between Middle Jurassic to Tertiary in the western United States.

Geologic Setting.  Near the end of the Paleozoic Era, oceanic deposits were being thrust eastward onto the North American continental margin during the Sonoma orogeny (Burchfiel and Davis 1972; Silberling 1973). Subduction of an east-dipping oceanic slab was established in the late stages of this orogeny or immediately following it (Lawton 1994), beginning a long period of subduction related volcanism in western North America, starting in the Triassic and continuing to the present (van der Pluijm and Marshak 1997). During this extended period of subduction along the western margin of North America, several major tectonic changes affected the amount and type of volcanic rocks produced. The arc in the Late Triassic to Middle Jurassic was probably a low-standing, graben depression similar to the arc found today in Central America (Busby-Spera 1988). From Middle Jurassic to Late Jurassic, however, the arc changed from a low-standing depression to a higher standing region of modestly thickened crust (Elison 1995) of mixed origin that has been called the Mesocordilleran Geanticline (Kauffman and Caldwell 1993). The increased convergence along the arc during this time interval is thought to be related to westward motion of North America due to the opening of the North Atlantic (van der Pluijm and Marshak 1997). The rapid motion of North America was accompanied by a pulse of increased magmatic activity as demonstrated by histograms of dates on plutons and estimates of magmatic flux (Armstrong and Suppe 1973; Miller et al. 1987; Barton et al. 1988; Barton 1990; Elison et al. 1990; Christiansen et al. 1994).

The Middle to Late Jurassic period of active volcanism was followed by an extended Early Cretaceous period of about 20 Ma when volcanic activity was almost completely absent (Armstrong and Suppe 1973; Armstrong 1988, Armstrong and Ward 1993; Elison 1995), perhaps due to accretion of the Insular Composite Terrane with its accompanying tectonic adjustments (Armstrong and Ward 1993). About 125 Ma, volcanic activity resumed in what was to be the most voluminous episode of magmatism in the western North American subduction complex (Armstrong and Ward 1993; Monger 1993; Christiansen et al. 1994). Accompanying this pulse of volcanism was extensive crustal shortening and thickening (Armstrong and Ward 1993), building a high-standing, Andean-type arc (Kauffman and Caldwell 1993). Throughout the Cenozoic, arc-related volcanism along the western margin of North America has continued, although at a reduced level when compared to the Late Cretaceous.

Temporal Changes in Magmatism.  The early subduction history of the Western Cordillera during the Late Triassic to Jurassic was one of fairly passive subduction with extension processes active along the arc. The oceanic lithosphere, which was being subducted in this early phase, was relatively old, cold, and covered with a thick blanket of oceanic sediment. It probably sank at a relatively steep angle with little resistance to subduction. As was pointed out earlier, the arc itself was a low-standing edifice within an extensional graben (Busby-Spera 1988). Isotopic compositions of plutonic rocks emplaced during the Jurassic magmatic pulse suggest mantle-derived magmas with little crustal assimilation (Miller et al. 1989; Wright and Wooden 1991; Elison 1995), another evidence that western North America was not significantly compressed during this interval.

As subduction proceeded into the Cretaceous and Tertiary, the character of the subducting oceanic slab became progressively younger and warmer, with less sediment cover creating more resistance to subduction. This caused the continental margin to undergo substantial compression, shortening, and crustal thickening (Armstrong and Ward 1993). Plutonic rocks emplaced during this period were strongly contaminated with crustal material and in many cases became essentially pure crustal melts (Best et al. 1974; Kistler et al. 1981; Lee et al. 1981; Miller and Barton 1987). Arc-related volcanism waned in the central part of western North American in the Miocene as the plate boundary changed from subduction to strike-slip motion. Miocene and younger volcanism became distinctly bimodal and extension once again dominated crustal tectonics (Christiansen and Lipman 1972; Best 1986).

Titanite Occurrence in Volcanic Rocks.  Titanite is a common accessory phase in plutonic and metamorphic rocks, but has been reported to be uncommon in volcanic rocks. This appears to be true for many volcanic suites, however, in some volcanic suites, titanite appears to be quite common. We have examined the mineralogy of numerous volcanic ash beds ranging in age from Middle Jurassic to Tertiary from the western U.S. and find some interesting patterns that may be related to changes in the character of the source region.

Middle Jurassic.  Altered ash beds (bentonites) from the Middle Jurassic Temple Cap and Carmel Formations and from the Late Jurassic Morrison Formation in Utah record the same Middle to Late Jurassic pulse in magmatism found in plutonic rocks (Christiansen et al 1994; Kowallis et al. in press). Interestingly, these Jurassic ash beds commonly have titanite (Everett 1989; Kowallis et al. 1991; Zhang, 1996; Kowallis et al. 1998). Of thirty ash beds examined from the Middle Jurassic Temple Cap and Carmel Formations, 50% had accessory titanite (Zhang, 1996), while of 99 ash beds examined from the Late Jurassic Morrison Formation 53% had accessory titanite (Kowallis unpublished data). Schermer and Busby (1994) describe the mineralogy of about ten different Middle Jurassic ash beds from the central Mojave Desert. In only two of their descriptions do they mention accessory mineral phases and both include titanite. Fackler-Adams et al. (1997) indicate that titanite occurs in 3 out of 5 tuffaceous units in Early to Middle Jurassic rocks of the Palen Mountains in southeastern California.

If titanite is an indicator of oxidizing conditions as suggested by Wones (1989), then the presence of titanite in half or more of the Jurassic ash beds may indicate relatively oxidizing conditions for these magmas. This, in turn, may be related to the fact that Jurassic plutons of this same age have Nd, Sr, and Pb isotopic ratios that cluster nearer to bulk earth ratios unlike younger Late Cretaceous and Tertiary plutons (Elison 1995). The isotopic signature of the Jurassic plutons has been interpreted to signify mantle-derived magmas with little crustal contamination. Organic material from crustal contamination would reduce the magmas, making titanite less likely to form.

Late Cretaceous.  During the Late Cretaceous pulse in magmatism, airfall ash from over 1300 eruptions found its way into the Western Interior Cretaceous Seaway and is now preserved as layers of bentonite (Kauffman and Caldwell 1993). We have examined 11 different ash beds from near the Cenomanian-Turonian boundary (Kowallis et al. 1995; Kowallis unpublished data) and another ash bed from near the Albian-Cenomanian boundary (Cifelli et al. 1997). Titanite is not found in any of these ash beds, although apatite and zircon are common, and monazite and garnet are present in one or two of these beds. The ash beds from near the Cenomanian-Turonian boundary were examined from several different localities in different sedimentary environments, and the same mineral assemblages for each ash bed were found at every locality (Kowallis et al 1989). It does not appear, therefore, that titanite is absent due to chemical instability, but simply was not formed in the magmas.

A search of the literature on Cretaceous bentonites revealed that most often the detailed mineralogy of these beds has not been identified, but in cases where it was examined, no titanite was found or mentioned. For example, Dengo (1946) identified plagioclase, k-feldspar, biotite, quartz, magnetite, zircon, limonite, and tourmaline in Cretaceous bentonites from the Mowry Shale, Frontier Formation, and Steele Formations of Wyoming, but did not find titanite. Knechtel and Patterson (1962) mention quartz, biotite, feldspar, and cristobalite in a study of bentonites from the northern Black Hills District, but do not mention any minor accesory phases. Slaughter and Early (1965) found feldspars, biotite, quartz, magnetite, zircon, hornblende, and augite in a study of 800 samples from upper Mowry and lower Frontier Formation bentonites in Wyoming; but again they do not identify titanite in any of the samples. Interestingly, Forsman (1992) finds biotite, muscovite, quartz, plagioclase, apatite, cordierite, sphene (titanite), and zircon in the Sentinel Butte Tuff of Paleocene age in North Dakota, but in two Cretaceous tuff beds lower in the section, no mention is made of titanite.

The absence of titanite in these ashes is consistent with the idea that the Late Cretaceous magmatic pulse produced magmas that were largely crustal melts (Elison 1995). Assimilation of organic material from the crust would reduce the magmas and in these poorly oxygenated conditions titanite would be unlikely to form.

Tertiary.  Tertiary calderas and associated ash flow tuffs and lavas have been the focus of numerous studies. Best (personal comm.) indicates that of about 120 different cooling units studied in Tertiary volcanic deposits in the Great Basin, only 8 (< 7%) have been found to contain titanite phenocrysts. This is considerably lower than the 50%+ occurrence rate that we find in the Jurassic ash beds. Again, most Tertiary magmas are thought to be crustal melts or significantly contaminated with crustal material.

Elemental Differences in Volcanic Titanites.  It has been know for some time that the composition of titanite is sensitive to the environment of formation (Fleischer and Altschuler 1969; Fleischer 1978), but no study has examined the compositional variations in titanite from different volcanic rocks or suites of rocks. We have analyzed and report here compositional variations between titanite from intermediate to silicic volcanic rocks, and between suites of volcanic rocks from the Middle Jurassic, Late Jurassic, and Tertiary of the western U.S.

Data Collection.  Titanite was separated from Jurassic bentonites and Tertiary ash flows and lavas by standard mineral separation methods. The grains were then mounted and polished for electron microprobe analysis. Analyses were performed at the University of Utah using a CAMECA SX-50 electron microprobe. Acceleration voltage employed was 15 kv, the beam size was 30 microns, and both natural and synthetic standards were used. We have collected over 130 analyses on volcanic titanite; Table 1 gives typical examples of these analyses. The complete data set is available from the authors. Also included in our data set are 34 published analyses on alkalic volcanic rocks (Smith 1970; Cundari 1979; Giannetti and Luhr 1983; Luhr et al. 1984; Wolff 1984; Paslick et al. 1996; Dawson and Hill 1998; Ventura et al. 1999) and 8 published analyses on dacitic rocks (Nakada 1991; Coyle and Wagner 1998). Iron is reported as Fe³⁺ because it has been shown to be the most common oxidation state for iron in titanite, even though some iron is usually also present as Fe²⁺ (Muir et al. 1984; Holényi and Annersten 1987).

General Trends.  Volcanic titanite most commonly occurs in calc-alkaline dacitic to rhyolitic rocks, but also occurs in alkalic rocks over a wide range of SiO₂ concentrations (Fig 1). Using the titanite extracted from calc-alkaline Tertiary volcanic rocks of known composition, it can be seen that certain elements vary systematically from andesite to dacite to rhyolite. For example, Ca decreases from almost 1 atom per formula unit (apfu) in andesite to nearly 0.9 apfu in rhyolite, while Ti decreases from about 0.95 apfu in andesite to 0.85 apfu in rhyolite (Fig. 2). Light rare earth elements (LREEs) increase from almost zero apfu in andesite to 0.05 apfu in rhyolite (Fig. 3), while both Y and Mn are near zero in andesite, but show significant concentrations in dacite and rhyolite (Fig. 4). These changes in titanite mimic changes in the chemical composition of the rocks. Titanite from alkalic volcanic rocks is generally higher in Nb and Fe, but lower in LREEs, Y, and Al when compared to titanite from calc-alkaline rocks (Fig 5).

Interpreting Data from Jurassic Bentonites.  Titanites from Jurassic bentonites of the Morrison, Carmel, and Temple Cap Formations in Utah show some interesting chemical differences when compared to our Tertiary samples. Based on the residual phenocryst population in these altered ash beds, it appears that most of them were dacites and rhyolites (quartz, sanidine, plagioclase, biotite), with a few that were probably andesites or low silica dacites (pyroxene, hornblende, plagioclase) (Zhang 1996; Kowallis unpublished data). The titanite crystals extracted from these bentonites are euhedral and appear to be free of alteration, with little or no internal zoning. We believe, therefore, that the chemical differences observed between the titanite from Middle and Late Jurassic rocks and from Tertiary rocks are real, reflecting differences in the magma systems.

Conclusions.  This is a work in progress!