榴辉岩常用温压计在应用中应注意的问题

Reviews on applying common-used geothermobarometers for eclogites.

本文通过再现相平衡实验数据和检查热力学活度模型两种手段,对榴辉岩中几种常用的温压计进行了检验,发现榴辉岩中某些常用温压计存在以下问题:(1)在 Eliis and Green(1979)、Powell(1985)、Krogh(1988)和 Ravna(2000)四种石榴石-单斜辉石温度计中只有 Ravna(2000)的版本能较好的再现相平衡实验数据。(2)将石榴石-单斜辉石温度计应用于含 X_(Jd)>0.55绿辉石的榴辉岩中会出现很大的误差。(3)Green and Hellman(1982)的石榴石-多硅白云母温度计计算的高压含多硅白云母榴辉岩变质温度普遍偏高,但是计算超高压榴辉岩的结果能较好的与 Ravna(2000)的石榴石-单斜辉石温度计计算结果保持一致。(4)Waters and Martin(1993)的石榴石-单斜辉石-多硅白云母压力计、Ravna and Terry(2004)的石榴石-单斜辉石-多硅白云母-蓝晶石-柯石英/石英温压计的精度都受到了 Holland(1990)的单斜辉石活度模型的限制,它们不能适用于绿辉石 X_(Jd)>0.55的榴辉岩,而 Waters and Manin(1996)对 Waters and Martin(1993)的版本做了一个经验校正,弥补了单斜辉石活度模型的缺陷,因此可以适用于绿辉石 X_(Jd)>0.55的榴辉岩。(5)Nakamura and Banno(1997)的石榴石-绿辉石-蓝晶石-柯石英温压计因运用了不恰当的石榴石和铁钙辉石的活度模型,从而使得计算结果与岩相学观察结果不一致。因此,我们建议:对于绿辉石 X_(Jd)<0.55的多硅白云母榴辉岩,可以运用 Waters and Martin(1993)压力计和 Ravna(2000)的温度计联合求解温压;对于含高硬玉组分(X_(Jd)>0.55)绿辉石的超高压多硅白云母榴辉岩,可选用 Waters and Martin(1996)压力计和 Green andHellman(1982)的温度计联合求解温压;对于含有石榴石+绿辉石+蓝晶石+柯石英矿物组合的榴辉岩,在 X_(Jd)<0.55的情况下。可选用 Ravna and Terry(2004)的温压计求解温压。在应用这些温压计时,应注意各温压计的适用温压范围和成分范围,尤其是石榴石 X_(Jd)、Mg~#和绿辉石 X_(Jd)的范围。另外,由于矿物中 Fe^(3+)的含量对温度计算结果影响很大,所以还必须合理地校正所选矿物的 Fe^(3+)。

Several geothermobarometers, common-used for P-T estimations of eclogites, have been evaluated through reproducing phase equilibrium experimental data and checking their reliability of thermodynamic activity model. We found that there have five problems in these thermobarometers. （ 1 ） In the four versions of common-used garnet-clinopyroxene Fe-Mg exchange thermometer （Ellis and Green, 1979; Powell, 1985; Krogh, 1988; Ravna, 2000）, the version of Ravna （2000） yields the smallest absolute error in reproducing the experimental temperatures collected from literatures and thus is the first choice for temperature estimation; （2） These four versions of garnet-clinopyroxene thermometer cannot be used for eclogites with Xjd 〉 0. 55 in omphacite; （3） The temperatures for ultrahigh-pressure （UHP） phengite-bearing eclogites calculated by the garnet-phengite thermometer of Green and Hellman （ 1982 ） are consistent with those obtained by the thermometer of Ravna （2000）, suggesting that the thermometer of Green and Hellman （1982） is reliable for UHP eclogites. However, Green and Hellman （1982） garnet-phengite thermometer usually overestimats temperature of high-pressure （HP） phengite-bearing eclogites; （4） The validity of the garnet-clinopyroxene-phengite barometer of Waters and Martin （ 1993 ） and the garnet-clinopyroxene-kyanite-phengite-coesite/quartz thermobarometer of Ravna and Terry （2004） is controlled by sodic clinopyroxene activity model as ascribed by Holland （1990）. They cannot be applied to eclogites with Xjd 〉0.55 in omphacite. However, Waters and Martin （1996） made an empirical correction for the barometer of Waters and Martin （1993） to remedy the limitation of clinopyroxene activity model, thus it is suitable for eclogites with high jadeite content （ Xjd 〉 0. 55 ） in omphacite; （5） Nakamura and Banno （ 1997 ） used unreliable garnet and hedenbergite activity models in calibrating a thermobarometer for the UHP assemblage garnet-omphacite-kyanite-coesite, this thermobarometer usually yields too much lower pressures. We suggest （1） to combine the barometer of Waters and Martin （1993） and the thermometer of Ravna （2000） for P-T estimation of phengitebearing eclogites with Xjd 〈 0. 55 in omphacite ; （2） to combine the thermometer of Green and Hellman （1982） and the barometer of Waters and Martin （1996） for P-T estimation of UHP phengite-bearing eclogites with high jadeite content （ Xjd 〉 O. 55 ） in omphacite; （3） to use the thermobarometer of Ravna and Terry （2004） for UHP kynaite eclogites with Xjd 〈 0. 55 in omphacite. Application range, especially the range of Xc, and Mg^# in garnet and Xjd in omphacite, must be fully taken into account in applying the thermobarometers. Temperature estimates are strongly affected by Fe^3＋ contents, thus the Fe^3 ＋ contents in garnet and omphacite must be correctly normalized.