Full Length Article
Fractionation in the graphitization reaction for 14C-AMS analysis: The role of Zn × the role of TiH2

https://doi.org/10.1016/j.ijms.2017.10.005Get rights and content

Highlights

  • Isotopic fractionation in the graphitization reaction for 14C AMS was evaluated.

  • Processes with different amounts of Zn and TiH2 were compared.

  • Optimum amounts of reagents for the TiH2/Zn reaction achieved less than 2‰ depletion.

  • In the Zn free process fractionation increases for the highest yields (80%).

  • In the TiH2 free process, for Zn:C > 50, fractionation tends to 0 and yield to 100%.

Abstract

Isotopic fractionation in the graphitization reaction for Radiocarbon Accelerator Mass Spectrometry measurements was evaluated. At LAC-UFF we use the TiH2/Zn reduction method, where H2 and Zn enable the reduction of CO2 into graphite. In previous works, we analyzed C stable isotopes ratio for different amounts of Zn in the range of 0–50 mg and the results indicated that the amount of Zn is strongly linked to both the degree and variability of isotopic fractionation during the reaction when the quantity of TiH2 ranges from 10 to 15 mg for samples containing between 0.7 and 1.0 mg of C. For no changes in fractionation have occurred by increasing the graphitization temperature from 520 °C to 550 °C, in the present paper we evaluate the effects of different temperatures for the H2 release step. Previous works have also led us to question the role of Zn in the TiH2-based process and how it affects fractionation. Therefore, graphitization reactions without TiH2 or without Zn, using different amounts of TiH2, were tested. The former reaction has required values of Zn:C > 50 while the latter is not recommended. When Zn:C = 30, minimum amounts of TiH2 (TiH2:C < 20) have led to reasonable results of isotopic fractionation (<2‰).

Introduction

Sample preparation for Radiocarbon Accelerator Mass Spectrometry (14C-AMS) requires specific physical and chemical pre-treatments followed by conversion to a suitable matrix, allowing the extraction of intense and stable ion beams. 14C AMS targets can be either CO2 or graphite, with both advantages and drawbacks for the two types [1], [2], [3], [4]. Sample graphitization is usually performed by reduction of CO2 to solid carbon on a metal catalyst, using Zn [5], [6], H2 [7] or TiH2/Zn [8], [9]. The latter has been successfully applied to both regular [9] and ultra-small samples, i.e. <0.1 mg [10], [11], providing a less expensive alternative to H2 reduction method. Moreover, the TiH2/Zn reduction method may have higher throughput, allowing more samples to be prepared in a work day. Recently, Xu et al. (presented in the 14th Accelerator Mass Spectrometry conference in Ottawa, Canada) have demonstrated another great advantage of this method, which is that the samples can be stored under vacuum in the sealed tubes since graphitization and for over a decade with no implications to the isotopic ratios. However, mass-dependent fractionation in such reaction can lead to results a few parts per mil lighter [8], [9] than the original CO2 δ13C values or even discrepancies as large as −30‰ and great scattering of results, what can lead to inaccurate results in the AMS system [12].

The degree of fractionation depends on the sample size, the amounts of reagents used, temperature and time of the process. Indeed, temperature plays a major role considering that discrepancies can be largely reduced when temperature is increased from 460 °C to 520 °C [12]. McNichol et al. [13] claim that Fe seems to be a better catalyst between 575 °C and 650 °C, and Tschekalinskij et al. [14] achieved very good results in the reaction with Zn and TiH2 at 600 °C. However, such high temperatures prevent the use of Pyrex® tubes [9]. Xu et al. [9] proposed a two-steps reaction in which a first period at a lower temperature (3 h at 500 °C) would favor the reduction of CO2 to CO while in the second period (4 h at 550 °C), CO would be reduced to graphite. For ultra-small samples, Xu et al. [11] lowered the reaction temperature to 450 °C for 7 h for the Zn reduction without TiH2 after Santos et al. [15] (based on the H2 reduction method), who had shown that lower temperatures help to increase graphitization yields. Marzaioli et al. [16], using a two-steps temperature protocol for regular-sized samples, reported a reduction in both the offset and the scattering in δ13C values when the reaction temperatures were increased from 530–550 °C to 550–560 °C. Macario et al. [17], [18] observed similar results with temperatures of 520 °C and 550 °C [12], in which δ13C mean values were depleted by approximately 5‰.

Orsovszki and Rinyu [19] reported results for the reactions using 60 mg Zn/10 mg TiH2 and 15 mg Zn/no TiH2. They concluded that eliminating TiH2 suppresses the formation of CH4 during graphitization. For the process without TiH2 there was an improvement in yield and from measurements within the accelerator they observed an increase in current intensity, while background levels remained similar for both methods.

Macario et al. [17], [18] have reported results showing that lowering the amount of Zn reduces scattering without increasing the degree of isotopic fractionation. Moreover, Macario et al. [18] call attention to the possibility of eliminating Zn from the process. However, the degree of isotopic fractionation in such process was still too high, with a mean value of (−6.1 ± 1.1) ‰. Furthermore, if on the one hand Macario et al. [18] observed that the amount of TiH2 plays no major role in the TiH2/Zn reduction, on the other hand, they suggest that in the absence of Zn, the availability of TiH2 should be increased to allow the completeness of graphitization reaction. Considering stoichiometry, in order to reduce CO2 to C, at least two moles of TiH2 should be added for each mol of CO2. This means that for each 1 mg of C at least 8 mg of TiH2 should be available. In fact, Tschekalinskij et al. [14] concluded that the best amounts of reagents for 1 mg samples and the reaction at 600 °C were 20 mg Zn and 8 mg TiH2, reaching −0.2‰ from the expected values.

In the present paper, we report several tests performed to evaluate the processes without Zn and without TiH2 (for 7 h and 10 h long processes), plus the TiH2/Zn reduction with variable amounts of TiH2. Moreover, we compare the results for two temperature steps with our regular 550 °C for 7 h process.

Section snippets

Methods

Using an IRMS system we have measured the C stable isotope ratios for graphitized NBS oxalic acid (OXII) standard 4990c (δ13C = −17.8 ± 0.1‰) [20]. Sample combustion took place in sealed quartz tubes, containing previously heated CuO (Fisher Scientific, carbon compounds 0.0004%) and Ag wire (Aldrich ≥99.99% 0.5 mm diameter), at 900 °C for 3 h. The gas was then purified in a stainless steel vacuum line [12] using cryogenic traps and transferred into graphitization tubes: borosilicate glass tubes with 9 

Results and discussion

Results of test A are shown in Fig. 1. Fig. 1a and b shows the discrepancy of measured δ13C from consensus value versus the TiH2:C and Zn:C ratios, respectively. Mean values of discrepancy, mostly between −7 and −6, show no significant variation either for two steps processes or for the use of Zn or not.

Fig. 1c shows that no variation in yield is observed for the two steps processes. On the other hand, there is an increase in yield, from 40% to 60%, when Zn is used. It is desirable to have the

Conclusions

From the obtained results, it is possible to see that the graphitization without Zn may lead to a low fractionation pathway, though the yield is much lower, demonstrating the importance of Zn in the process. Two other pathways have led to reasonable results (<2‰): optimum amounts of TiH2 plus Zn and the reaction free of TiH2. Therefore, we do not recommend the reaction without Zn. Indeed, minimum amounts of TiH2 or even none of it should be used, provided enough Zn (Zn:C > 50) is available in the

Acknowledgements

The authors would like to thank Brazilian financial agencies CNPq (Conselho Nacional de Desenvolvimento Científico e Tecnológico, 305079/2014-0 and INCT-FNA464898/2014-5) and FAPERJ (Fundação Carlos Chagas Filho de Amparo à Pesquisa do Estado do Rio de Janeiro, E-26/110.138/2014) for their support. We are also grateful to the UCDavis personnel. We thank the editor and reviewers for comments and suggestions that helped to improve the quality of the manuscript.

References (22)

  • C.P. McIntyre et al.

    Online 13C and 14C Gas Measurements by EA-IRMS-AMS at ETH Zürich

    Radiocarbon

    (2017)
  • View full text