The Use of Thermobarometry to Determine the Extent of Metamorphism Has Been Used in Various

1Abstract

2Introduction

The use of thermobarometry to determine the extent of metamorphism has been used in various manners for the past hundred years. The original studies of Barrow and Buchanan based approximate temperatures and pressures of metamorphism upon the presence or absence of certain index minerals. The recognition that mineral composition could be used to further determine the extent of metamorphism was made clear through the varying types of a given mineral (e.g. garnets, pyroxenes, amphiboles) depending on the tectonic history of a province.

3Geologic History

The crystalline rocks of New England have been deformed during three major orogenies: the Taconian (mid-Ordovician), the Acadian (early Devonian) and the Alleghanian (Pennsylvanian-Permian) (Rankin, 1994). The polymetamorphic history of the various terranes has been the object of study for decades and the overprinting relationships among the various orogenies are still not well understood. Sutter et al. (1985) utilized 40Ar/39Ar and K-Ar data from the Berkshires of western Massachusetts (Fig. 1) to define the boundary between rocks metamorphosed primarily in the Taconian orogeny and rocks overprinted by the Acadian orogeny. Similar research for the Bronson Hill terrane was conducted by Boyd et al. (1993) to discriminate between rocks metamorphosed principally in the Acadian orogeny and those later remetamorphosed by the Alleghanian orogeny in the Permian.

3.1Taconian Orogeny

The Taconian orogeny occurred during the Ordovician from ~470 to 440 Ma (Rankin, 1994). The Champlain thrust emplaces the lower Cambrian Dunham Dolostone over the deformed mid-Ordovician Iberville Shale (Stanley, 1987). The thrust is unconformably overlain by lower Devonian rocks in the Hudson Valley (Bosworth et al., 1988; Thompson et al., 1993), which defines the Taconic orogeny as Ordovician in age. Tucker and Robinson (1990) have dated a series of subduction-related volcanics and plutons in the Bronson Hill terrane as having ages between ~470 and 435 Ma. Ordovician and Silurian K-Ar and 40Ar/39Ar cooling ages of biotite and hornblende in the Berkshires are due to heating during the Taconian orogeny (Sutter et al., 1985).

3.2Acadian Orogeny

The Acadian orogeny is Devonian and ranges from ~400 to 380 Ma in age (Rankin, 1994). The Catskill Delta is a clastic wedge of Devonian sediments that was deposited as a result of the Acadian orogeny in New England (Faill, 1985). Plutons across New England crystalized and were deformed during the Acadian (Bradley et al., 1998). Late Devonian and Mississippian K-Ar and 40Ar/39Ar cooling ages of hornblende and micas reflect cooling from the Acadian orogeny (Clark and Kulp, 1968; Sutter et al., 1985; Harrison et al., 1989; Spear et al., 1989; Tucker and Robinson, 1991; Moecher et al., 1997).

3.3Alleghanian Orogeny

The Alleghanin orogeny transpired between ~320 and 270 Ma during the Pennsylvanian and Permian (Rankin, 1994). In Pennsylvania, folded Carboniferous sediments are unconformably overlain by undeformed Triassic sediments, defining the Alleghanian orogeny. Similar constraints are also found in New England, where the Pennsylvanian sediments of the Narragansett Basin have been metamorphosed to staurolite grade, but have late Permian mica cooling ages (Mahler and Mosher, 1994). The increasing number of isotopic ages for crystalline rocks in New England are indicative of a wide-spread Permian cooling event (Clark and Kulp, 1968; Zartman, 1988; Harrison et al., 1989; Spear et al., 1989; Wintsch et al., 1993; Boyd et al., 1995).

3.4Extent of the Orogenies

K-Ar and Ar40/Ar39 research into muscovite, biotite, and amphibole cooling ages has been used to delineate the extent of the three orogenies within New England. This process is based on the fact that the argon closure temperature for muscovite is ~350E C, biotite ~275E C, and amphibole ~500E C. For muscovite, Zartman et al. (1970) define ages >350 Ma as being Taconian, 350-260 Ma as Acadian, and 260-200 Ma as Alleghanian in age.

3.4.1Acadian versus Taconian

The line between rocks solely metamorphosed in the Taconian Orogeny versus rocks metamorphosed in the Acadian Orogeny was drawn by Zartman et al. (1970) using K-Ar ages for muscovite. The boundary of the Acadian overprint in western New England was further refined by Sutter et al. (1985) based on both K-Ar and 40Ar/39Ar cooling ages for muscovite, biotite, and amphibole (Fig. 1).

3.4.2Alleghanian versus Acadian

Zartman et al. (1970) differentiated between rocks with solely Acadian cooling ages and rocks with an Alleghanian overprint in the muscovite ages. The western age of the Alleghanian overprint approximately coincident with the boundary between the Bronson Hill Terrane and the Connecticut Valley Synclinorium. In southern New England, Alleghanian muscovite cooling ages extend from the Bronson Hill to the western edge of the Avalon terrane. The bulk of the Avalon terrane, however, has ages >260 Ma. Rocks within the Putnam-Nashoba Terrane recorded Acadian cooling ages in the northern portion, but several Alleghanian muscovite ages are located within the southern section of the terrane. Several muscovite ages in the northern portion of New Hamphshire and the mid-section of Maine are Alleghanian in age, but a large area of both states had not been analyzed as of 1970.

Boyd (1995) analyzed hornblende and muscovite for 40Ar/39Ar cooling ages along a traverse of the Bronson Hill Terrane in conjunction with data from previous studies (Brookins, 1970; Brookins and Armstrong, 1980; Wintsch and Sutter, 1986; Harrison et al.,1989; Spear and Harrison, 1989; and Tucker and Robinson, 1990). Boyd concluded that the muscovite ages are all approximately equal at 250 Ma for the length of the terrane (fig. 1 and 2), but the amphibole cooling ages are latitude dependent. An age gradient from north to south existed with the amphibole ages, with the older cooling ages in the northern section of the terrane (fig. 2). Boyd proposed that the loading of the Bronson Hill Terrane was unequal, with a greater amount of burial in the southern portion of the terrane (fig. 3). Wintsch et al. (2001) used cooling path models to demonstrate that the rocks in the southern portion of the Bronson Hill Terrane were exhumed at a greater rate than those in the northern section of the terrane, in order for all of the muscovite cooling ages to be equivalent (fig. 4).

4Geologic Setting

The Bronson Hill is a north-south trending terrane that is continuous from Long Island Sound in southern Connecticut to the Maine-Quebec border. Traditionally considered the island arc that formed over an east-dipping subduction zone during the Taconian Orogeny, the terrane was accreted during the Ordovician to Laurentia and later metamorphosed in the Acadian and Alleghanian orogenies (Bradley, 1983; Harrison et al., 1989).

The rocks collected for this study contain amphibolite from the Bronson Hill terrane along a traverse from southern Connecticut through Massachusetts. The amphibolite was sampled from the Partridge Formation and the Ammonoosuc Volcanics, both of which are Ordovician in age (Zartman and Leo, 1985; Tucker and Robinson, 1990). The two formations represent volcanics extruded onto the island arc. Specimens were located a maximum of 18 km from each other on a north-south traverse of the terrane, with at least one sample per 15' quadrangle (fig. 5 and appendix 1). These amphibolites were chosen for the assemblage amphibole + plagioclase " garnet to be analyzed for petrologic and thermobarometric calculations using Holland and Bludny=s (1994) amphibole-plagioclase thermometer, the Kohn and Spear (1989) garnet-hornblende-plagioclase-quartz thermometer, and the Dale et al. (2000) hornblende-garnet-plagioclase thermobarometer.

5Methods

Specimens used for this study were located a maximum of 18 km from each other on a north-south traverse of the Bronson Hill Terrane, with at least one sample per 15' quadrangle (fig. 5 and appendix 1). These amphibolites were chosen for the assemblage amphibole + plagioclase " garnet to be analyzed for petrologic and thermobarometric calculations using Holland and Bludny=s (1994) amphibole-plagioclase thermometer, the Kohn and Spear (1989) garnet-hornblende-plagioclase-quartz thermometer, and the Dale et al. (2000) hornblende-garnet-plagioclase thermobarometer.

Rock samples were cut perpendicular to the main foliation whenever identifiable and billets were also aligned parallel to mineral lineations where visible. The thin sections were singly polished for analysis on the Cameca 500 SX Microprobe at Indiana University, as well as microstructural analysis with the optical microscope.

6Plagioclase

6.1Petrography

Within the thin sections studied, plagioclase is anhedral and round to oval in shape. When oval, the long direction of the plagioclase is aligned with the dominant foliation of the sample. Plagioclase and quartz are commonly intergrown and embay one another.

6.1.1Optical Zonation

Visibly zoned metamorphic plagioclase was observed in a number of the thin sections. Optical zonation of metamorphic plagioclase has been discussed by Nord et al. (1978), Crawford (1966), Stoddard (1985) and Passchier and Trouw (1998). Plagioclase zoning is visible due to the difference in extinction angles between various portions of the grain (Nord et al., 1978; Stoddard, 1985). The boundary between different zones may be either gradual or sharp depending on the amount of equilibration the grain has undergone (Passchier and Trouw, 1998). The samples examined ranged from several distinct zones to having only indistinct cores and rims (fig. 6).

Of the samples analyzed, the presence of optically zoned plagioclase ranged from no optical zonation visible to 80% of the plagioclase crystals within a given thin section visibly zoned (fig. 7). In the Massachusetts section of the traverse, plagioclase zoning is rare. When elongate, the rims tend to align with the dominant foliation and rim-core or rim-mantle boundaries are diffuse, though core-mantle boundaries are sharp in mid-Massachusetts. In Connecticut, zoned plagioclase accounts for 50 to 80% of the total plagioclase grains. The elongate rims continue to be commonly aligned with the dominant foliation. Though in central Connecticut, the boundaries between the cores and mantles are sharp, the mantle-rim boundaries are diffuse. In the remaining portions of the state, the boundaries between cores and rims are diffuse, with no distinguishable mantles.

6.1.2Twinning

Within the thin sections studied, several types of feldspar twinning were observed. Passchier and Trouw (1998) differentiate between growth twinning (e.g. Carlsbad and Albite) and deformation twins (also called mechanical twins), which form to accommodate strain. The latter twins tend to be concentrated along the rim of individual grains and taper inwards towards the center of the feldspar. Deformation twins are most likely to form at metamorphic temperatures under 400E C (Passchier and Trouw, 1998).

One sample contained very rare Carlsbad growth twins (99ERG07c), but deformation twins were the most common found in the thin sections analyzed. The percentage of deformation twins for each sample was estimated and forms a pattern along strike of the traverse. Deformation twins are common in the Massachusetts samples and in the southernmost Connecticut rocks, but are rare throughout most of Connecticut (Fig. 8).

6.1.3Petrographic Interpretation

Passchier and Trouw (1998) describe the deformational behavior of plagioclase as being dependent on the temperature at which the deformation occurred and if the grains were heated to a higher temperature after the deformation.

At low temperatures (<400E C), plagioclase will be likely to deform via deformation twinning and undulose extinction. Deformation twinning can occur at higher temperatures, but is less abundant. Added heat to the system may provide sufficient energy for the plagioclase to Aheal@ due to dislocation climb and recrystallization. These processes can begin in feldspars at ~400-500E C. Deformation twins that are present within a thin section normally indicate that the rock was deformed at temperatures less than 400E C and the sample was not consequently reheated to greater than 400E C.

The zoning of plagioclase indicates that though the rock was at a temperature great enough for plagioclase to grow, the rock did not reach a sufficient temperature for the grains to equilibrate and become homogenous. The growth zonation can occur due to several different circumstances as outlined by Spear (1995): a change in P-T conditions, a change in mineral assemblage, the fractionation of material into the core of the mineral (like in igneous fractional crystallization), or the change of the bulk composition of the rock due to infiltration and metasomatism. The circumstances that cause the variation in conditions can occur during continuous or discontinuous crystal growth. The cation diffusion in plagioclase has been documented by Grove et al. (1984) as being very slow, so that plagioclase cannot easily reequilibrate except by dissolution and reprecipitation.

From north to south along the Bronson Hill traverse zoning and deformation twinning are dominant at different periods. Of the samples analyzed, those in Massachusetts and the southern-most section of Connecticut had a large percentage of plagioclase grains with deformation twins. Zoning of the plagioclase was prevailing feature in most of Connecticut. The rocks in Massachusetts were deformed at low temperatures but were not consequently heated to temperatures greater than 400E C. In Connecticut, the low percentages of twins indicate that the rocks were also deformed. The deformation twins may have formed at higher temperatures or the samples were deformed and then heated to >400E C, which healed most of the twins. The high percentage of twins in southern Connecticut may be due to a high amount of differential stress at temperatures >400E C that occurred during the rapid uplift of the southern-most section of the terrane as suggested by Wintsch et al. (2001).

6.2Microprobe Analysis

Of the samples collected and cut for thin sections, ?? thin sections were analyzed with the electron microprobe at Indiana University for plagioclase (fig. 9). Plagioclase was analyzed using a beam size of 10 kV and a current of 10 nA on the Cameca 500 SX microprobe.

6.2.1Anorthite value vs Latitude

The anorthite content of plagioclase ranges systematically from wide range in the northern sections of the travers to a low variability in northern Connecticut before rising once again in southern Connecticut (Fig. 10). The plagioclase composition varies along the length of the Bronson Hill Terrane from ~An15 to An90. Anorthite content has a mean value of An46. The An values within individual plagioclase have a difference between 0.1 and 25.1, with a mean value of 5.7. Within a thin section, the An value differs from between a change of only 1% to just over 50%.

6.2.2Plagioclase Stability

Carpenter (1994) defined regions of stability for specific anorthite values that includes several gaps on a binary phase diagram (Fig. 11). At temperatures lower than 400E C, plagioclase solid solution is very limited and except for extremely albitic or anorthitic compositions, two crystals are stable with one another. Greater than 400E C, solid solution in plagioclase becomes more prevalent, increasing in range with composition.

6.2.3Implications of Microprobe Plagioclase Studies

The systematic variance of plagioclase compositions from north to south in the Bronson Hill terrane may be directly linked to the stable areas of plagioclase solid solution as defined by Carpenter (1994). Rocks in the northern section of Massachusetts were heated to lower temperatures than those in southern Connecticut and Massachusetts. The possible anorthite solid solution would have been limited at low temperatures and extend over a greater range at higher temperatures. This temperature variation is also evidenced by the presence of migmatites in the southernmost section of Connecticut, pegmatites in the southern to central portion of Connecticut and in the Massahusetts section of the Bronson Hill terrane, and quartz veins in mid- to northern Connecticut (fig. 12).

7Amphibole

7.1Petrography

Two types of amphibole are located within the thin sections examined: the dominant form is an aluminous amphibole ranging from edenite to ferro-pargasite hornblende and ferro-tschermak hornblende to tschermak in composition and the second is tremolite. The aluminous amphibole dominates within the samples studied, but coexists with tremolite is within the thin sections 99ERG17 and 00ERG12 (fig. 5).

7.1.1Aluminous Amphibole

The aluminum-rich amphiboles are aligned with the dominant foliation in all of the thin sections studied. The aluminous amphibole defines the foliation, except in samples where the mode of aluminum is low. Deflection of the amphibole foliation occurs in some of the garnet bearing samples around the porphyblasts.

Aluminous amphibole grains are anhedral and elongate. Opaque inclusions are common within the amphiboles, but equant quartz inclusions are rare. In a few samples, the amphibole contain inclusions of sphene (numbers?).

7.1.2Tremolie

When tremolite is present within the thin section, it is aligned parallel to the dominant aluminous amphibole foliation. The tremolite can occur both as isolated crystals or as beards elongate in the direction of foliation on aluminous amphibole grains. Inclusions of aluminous amphibole are common within the tremolite and the aluminous amphibole is often truncated by the tremolite. The tremolite can increase the aspect ratio of the amphiboles by from 3:1 to 9:1.

7.1.3Implications of Tremolite and Aluminous Amphibole Relationships

Tremolite forms at low pressures and temperatures during greenschist facies metamorphism. The aluminous amphibole present, in contrast, formed during amphibolite facies metamorphism documenting a change from high-grade to low-grade metamorphism (fig. 13). The alignment of the tremolite with the dominant aluminous amphibole foliation is evidence that the differential stress in both cases must have been in the same direction.

7.2Microprobe Analyses of Amphibole

The amphibole was analyzed using a beam size of 10 kV and a current of 10 nA on the Cameca 500 SX microprobe. Only aluminous amphiboles were analyzed and their compositions are shown in figure 9.

7.2.1Amphibole Compositions

The analyzed amphiboles plot on an approximately straight line on both the Mg / Mg + Fe vs Si diagram for Na + K > 0.5 cations and Na + K < 0.5 cations (fig. 14 and 15). The amphiboles on the diagram with less that 0.5 cations of sodium and potassium plot in the edenite, edenitic hornblende, and ferro-pargasitic hornblende fields. What about the amphs in that field that has no name? What are they called? Are they even amphs? On the > 0.5 cation diagram, amphiboles are located within the ferro-tschermak hornblende, ferro-tschermak, and tschermak fields. The majority of the amphiboles have less than 0.5 cations of sodium and potassium. A few of the analyses produced amphiboles with less than 6 cations of silica and are probably due to experimental error.

7.2.2Igneous versus Metamorphic stability of aluminous amphiboles