calculating a normative mineral phase is generally not the same as using an observed mineral phase as a proxy for activity. If you don't see rutile in your crystal assemblage, then you have to assume that the rock was never saturated in TiO2 and the activity must be less than unity. The best approach would be to use MELTS to calculate the TiO2 and SiO2 activity of the melt based on the mineral assemblage observed and the composition of the rock.

Keep in mind that the zircon has to be in equilibrium with the melt when it crystallizes for this to work, even though we know that zircon can be recycled for 0.01 - 10's of Myr. Thus, if you think the zircon are anticrysts or xenocrysts, you will have to make an assumption of what the melt was at the time of crystallization.

I recomend Mathew A. Coble answer.

Thank you Matthew A. Coble . Yes I am aware it will be less than unity but since the values of aTiO2 in silicic melts at appropriate magmatic temperatures are rarely below ≈ 0.5 and typically 0.6 - 0.9 (Hayden et al. 2007). I will use MELTS as you suggested.

There are already some reseachers use the modified version pubilshed by Ferry and Watson (2007). For example, Shu et al (2019) and Shen et al (2015). I think the way they identify the aSiO2 and aTiO2 is suitable for most of the intermediate-felsic rocks. Because the intermediate-felsic rocks always have primary magmatic quartz, so the activity of SiO2=1. as for aTiO2 they use 0.75 the average of 0.6-0.9.

As Matthew A. Coble said, the premise is that you should make sure that the zircon you are dealling with is the primary crystal, not anticrysts or xenocrysts.

Shen, P., Hattori, K., Pan, H., Jackson, S., & Seitmuratova, E. (2015). Oxidation condition and metal fertility of granitic magmas: zircon trace-element data from porphyry Cu deposits in the Central Asian Orogenic Belt. Economic Geology, 110(7), 1861-1878.

Shu, Q., Chang, Z., Lai, Y., Hu, X., Wu, H., Zhang, Y., ... & Zhang, C. (2019). Zircon trace elements and magma fertility: insights from porphyry (-skarn) Mo deposits in NE China. Mineralium Deposita, 1-12.

Dear Kayode,

In oder to have "more" realistic zircon temperatures values, it is essential to well-constrain both aSiO2 and aTiO2. As Gaobin mentioned before, in the case that you have seen quartz in thin section, then you can apply a fixed SiO2=1. In the case that you're missing rutile (which is in most of the cases in volcanicand evolved plutonic rocks), then you can estimate model aTiO2 values with MELTS following Mathew's suggestion, or alternatively, by using the composition of a magnetite-ilmenite pair from your system as a proxy for TiO2 activity (if you don't have one, maybe there are some in the literature?). For the latter you can quickly obtained aTiO2 using this webpage from MELTS:

http://melts.ofm-research.org/CORBA_CTserver/OxideGeothrm/OxideGeothrm.php

Nevertheless I recommend you too also plot error bars that express different aTiO2 values, as the TiO2 activity calculated from your Ti-Fe oxide pairs could not be neccesarly in equilibrium with the crystallization of your entire zircon population.

Thanks a lot Gaobin and Alejandro. I will read up the articles and act appropriately for aTiO2 value(s). The aSiO2 is at unity as my rocks are quartz saturated, there are no normative Lc, Kp, Ne as such the FSSI > 0

There is new Zr-Hf zircon geothermometer:

Article New Zr–Hf Geothermometer for Magmatic Zircons

It has some advantages as it does not depend on the major-oxide melt composition and on the accuracy of the estimated SiO2 and TiO2 activities in the melts.

Thanks so much Viacheslav V. Akinin