The effects of temperature and f O2 on the Al-in-hornblende barometer

The AI-in-hornblende barometer potentially offers a basis for estimating crystallization pressure for granitic batholiths. However, owing to the simplicity of its formulation, misuse of the barometer can occur. Many granitic intrusions are emplaced at conditions inconsistent with those of the existing experimental calibrations, including f02 < NNO and/or variable to high temperature. The barometer is sensitive to variations in both f02 and temperature: low f02 can cause calculated pressures to be high by a factor of two or more, and the effect of temperature is up to 2 kbar per 100 DC, depending on total Al abundance. Batholiths emplaced at elevated temperature and portions of plutons that crystallized below H20 saturation yield artificially high pressures relative to near-solidus experiments conducted at H20 saturation. Al-in-hornblende pressures within the tonalitic Mount Stuart batholith of Washington, for example, define a 1-4 kbar domal structure that is exaggerated by emplacement crystallization temperatures that range from 620 to 760 DC.Pressure correlates with temperature (r2 = 0.86), indicating that temperature-sensitive edenite exchange has greatly influenced observed pressure variations. Data for other plutons also exhibit a marked positive correlation between temperature and hornblende pressure. If corrections are not made for temperature, such large apparent pressure variations can lead to overly high estimates of pluton thickness and tilt in addition to invalid tectonic interpretations. Using experimental data at -675 DC (Schmidt, 1992) and at -760 DC (Johnson and Rutherford, 1989), a revised expression fur the barometer incorporating the effect oftemperatureisP(:tO.6kbar) = 4.76AI3.01 ([TeC) 675]/85} x {0.530AI + 0.005294[T eC) 675]}. For a pluton emplaced at 100 DCabove wet-solidus temperatures, this reformulation ofthe barometer lowers derived pressures by 1.3 to > 2 kbar at typical crustal pressures. For the Mount Stuart batholith, consideration of temperature yields revised pressures that are in agreement with pressures obtained from wall rocks and eliminates much of the apparent domal structure. Low-f02 granites have amphibole Fe/(Fe + Mg) ratios that exceed the typical 0.40-0.65 range used in most experimental and empirical calibrations. Examples from anorogenic granitic batholiths of mid-Proterozoic age yield pressures that are too high by a factor of two to three in comparison with pressures obtained from adjacent metamorphic assemblages. Hornblende in these granites not only has high Fe/(Fe + Mg) but also low ratios of FeH to Fe2+. The anomalously high Al in Fe-rich, FeH -poor hornblende is inferred to be the result of increased [6JAIoccupancy of the M2 site not buffered by the Mg and FeH abundances typical of amphiboles in calc-alkaline and other highf02 plutonic rocks. INTRODUCTION empirical calibrations (Hammarstrom and Zen, 1986; Hollister et aI., 1987) and the experimental calibration conducted at near H20-saturated solidus temperatures (Schmidt, 1992) are similar, whereas the experimental calibrations conducted at elevated temperatures (Johnson and Rutherford, 1989; Thomas and Ernst, 1990) record higher AItot in hornblende at the same pressure largely because of increased [4JAl.The Schmidt (1992) calibration was determined over a temperature range of 655-700 DC at 2.5-13 kbar. The two high-temperature calibrations employed different approaches. The Johnson and RuthThe initial empirical formulation of the AI-in-hornblende barometer (Hammarstrom and Zen, 1986; Hollister et aI., 1987) was based on pressures determined in adjacent aureole rocks. The barometer has subsequently been experimentally calibrated for the assemblage quartz + alkali feldspar + plagioclase + hornblende + biotite + iron titanium oxide + titanite + melt + fluid (Johnson md Rutherford, 1989; Thomas and Ernst, 1990; Schmidt, 1992). Ail calibrations are summarized in Figure 1. The )003-004 X/9 5/0506-0549$02.00 549