INTRODUCTION
Accelerator mass spectrometry (AMS) is a type of high-energy mass spectrometry based on accelerator technology and ion detection, which was developed in the late 1970s (Bennett et al. Reference Bennett, Beukens, Clover, Gove, Libert, Litherland, Purser and Sondheim1977; Nelson et al. Reference Nelson, Korteling and Stott1977). Due to its extremely high isotope abundance sensitivity (10–15), minimal sample requirements (μg level), and short measurement time, AMS has attracted widespread interest, and the technique has made great progress with respect to miniaturization and specialist applications. The 14C dating technique based on AMS can provide accurate chronological results for nearly 50,000 years, and this plays an underpinning role in much earth system science (Kang et al. Reference Kang, Park and Kim2001).
The radiocarbon laboratory of Lanzhou University was established in 1987, and, at that time, the samples were prepared in benzene and analyzed using a liquid scintillation counter. A graphite synthesis system was set up in 2001, using the H2-Fe method for graphite target synthesis, and this afforded the preparation of one graphite target. To further increase the graphite preparation capability, a new synthesis system was built in 2011 and the prepared graphite samples were then sent to established AMS laboratories in China to complete the dating process. To further improve the working capacity of the 14C laboratory at Lanzhou University, a new compact 200kV AMS system, the mini carbon dating system (MICADAS), was introduced from Ionplus AG corporation in 2018 (Figure 1a), together with an automatic graphitization device (AGE III, Figure 1b). The installation, commissioning, and acceptance of the MICADAS were completed in the autumn of 2018. The MICADAS system is installed in many institutes worldwide, and good performance has been reported (e.g., Synal et al. Reference Synal, Stocker and Suter2007; Christl et al. Reference Christl, Vockenhuber, Kubik, Wacker, Lachner, Alfimov and Synal2013; Molnár et al. Reference Molnár, Janovics, Major, Orsovszki, Gönczi, Veres, Leonard, Castle, Lange, Wacker, Hajdas and Jull2013; Salehpour et al. Reference Salehpour, Håkansson, Possnert, Wacker and Synal2016; Mollenhauer et al. Reference Mollenhauer, Grotheer, Gentz, Bonk and Hefter2021).
MICADAS and associated equipment of the 14C chronology laboratory of Lanzhou University: (a) MICADAS; (b) elemental analyzer (left) and AGE III (right); (c) iron-powder dispenser (left) and pneumatic sample press (right).
Herein, we present a status report of the MICADAS at Lanzhou University based on the test results for standard and background samples obtained over the last two years. Additionally, 15 sets of routine archaeological materials including charred seed, charcoal, and bone collected from archaeological sites in northwest China were selected, and an inter-comparison study involving specialist laboratories from the University of Arizona, USA (Donahue et al. Reference Donahue, Jull and Toolin1990), Isotoptech Zrt, Debrecen, Hungary (Molnár et al. Reference Molnár, Janovics, Major, Orsovszki, Gönczi, Veres, Leonard, Castle, Lange, Wacker, Hajdas and Jull2013), Peking University (Liu et al. Reference Liu, Ding, Fu, Pan, Wu, Guo and Zhou2007) and Lanzhou University was conducted. In addition to the evaluation of the dating results, details of the pretreatment steps for the charred materials and bone, and the graphitization processes are presented.
METHOD
Sample Pretreatment
The primary samples are typically pretreated to remove contaminants that may affect the dating results. Routine pretreatment methods are described elsewhere (Brock et al. Reference Brock, Ramsey and Higham2007, Reference Brock, Higham, Ditchfield and Bronk Ramsey2010; Knowles et al. Reference Knowles, Monaghan and Evershed2019), and acid-base-acid (ABA) procedures were used for charcoal and charred seed samples in this study. The basic method in use at the radiocarbon laboratory (Lanzhou University) was as follows: ultrasonically cleaned and charred samples were reacted first with an acid solution (1 M HCl, 10 mL, 60°C, 6 hr; process repeated 3–5 times until the solution was colorless or effervescence stopped indicating the complete removal of carbonate), and then the samples were rinsed with Milli-Q water to achieve a neutral pH (confirmed using pH paper). The samples were next treated with alkaline solution (0.5 M NaOH, 10 mL, 60°C, 2 hr; process repeated 2-3 times until the solution was colorless signifying complete removal of humic acids), and then the samples were rinsed with Milli-Q to neutral pH as previous. The third step was reaction with an acid solution (1 M HCl, 10 mL, 60°C, 2 hr) to remove the carbonate precipitate that may have been generated due to the reaction with atmospheric CO2 during the alkali reaction, and again rinsing the sample with Milli-Q water to achieve neutral pH. The samples were dried at 60°C in air for 1–2 days.
In the case of bone samples, the collagen material needs to be extracted for radiocarbon dating. Dense bones (∼1 g) were sectioned with an electric saw, and the surfaces of the bones were polished before cleaning in an ultrasonic bath. The samples were then placed in acid solution (0.5 M HCl, 15 mL, 4°C, 24 hr; process repeated 12–15 times until the samples were soft and/or effervescence had stopped indicating removal of the hydroxyapatite), and then rinsed with Milli-Q water to neutral pH. Next, the samples were treated with alkaline solution (0.125 M NaOH, 15 mL, 4°C, 20 hr) to remove humic acid in the event that the bones were inadvertently contaminated with humic acid, and then the samples were rinsed with Milli-Q water. The third step was a reaction in acid solution (pH = 3, 75°C, 48 hr), and the gelatin-like solution was filtered using a filter tube. Finally, the filtrate was frozen and kept in a refrigerator for 2 days, and then freeze dried in a lyophilizer to complete the collagen extraction process.
Graphitization
The AGE III system consists of an elemental analyzer and graphitization equipment (Figure 1d) and is connected to compressed air (for power and cooling), helium (carrier gas), oxygen (combustion-supporting gas), and hydrogen (reducing gas) gas supplies. Depending on the carbon content of the different substances, a certain amount of sample (1.8–2.2 mg of charred remains and 2.8–3.2 mg of collagen) is taken for graphitization, using the AGE III system as described elsewhere (Wacker et al. Reference Wacker, Němec and Bourquin2010a). A previously-cleaned iron powder catalyst (4–5 mg) was loaded into a quartz sample vial, which had previously been heated in a muffle furnace at ∼500°C to remove possible organic matter contaminants (Figure 1e). After processing in the AGE device (details as given by Wacker et al. Reference Wacker, Němec and Bourquin2010a), a full-sized sample would typically contain about 1 mg of carbon. The graphite materials were finally pressed into cathodes using a pneumatic press (Figure 1e).
Accelerator Mass Spectrometer
Details of the MICADAS and the analytical approach are described elsewhere (Synal et al. Reference Synal, Stocker and Suter2007; Wacker et al. Reference Wacker, Bonani, Friedrich, Hajdas, Kromer, Němec, Ruff, Suter, Synal and Vockenhuber2010c). A full MICADAS magazine contains 39 cathodes including standards, blanks, and samples. In the routine test, typically 4-6 NIST SRM 4900C (OX-II) standards, one IAEA-C7 standard, one IAEA-C8 standard, and one or two phthalic acid (PHA) samples were placed in each magazine for analysis including data calculation, reproducibility monitoring, and background correction. Before commencement of a run, the ion source was optimized to produce a 12C- beam with an ion beam about 30∼50 μA. Typically, 10–15 passes and 15 cycles were required for each pass to acquire high precision results for the magazine (Wacker et al. Reference Wacker, Bonani, Friedrich, Hajdas, Kromer, Němec, Ruff, Suter, Synal and Vockenhuber2010c). The test sequence was defined by the software package Chameleon, and the dating results were calculated using the BATS software (Wacker et al. Reference Wacker, Christl and Synal2010b). Nemo software was used for real-time monitoring of the test data.
RESULT AND DISCUSSION
Long-Term Performance Data for MICADAS
The long-term results obtained at Lanzhou University over the last two years for the standard materials, OXII, IAEA-C7, IAEA-C8, and PHA by MICADAS are listed in Table 1; the results are also presented in Figures 2 and 3.
The pMC and δ13C values for the standard samples and the background samples.
N*: the number of samples measured.
The pMC values for OXII, IAEA-C7, and IAEA-C8 during a long-term test of MICADAS (Lanzhou University).
Plot comparing pMC values of the PHA to the time.
The pMC value obtained for OX-II (134.13 ± 0.48, n = 163) was consistent with the reference value (134.07 ± 0.05; Stuiver Reference Stuiver1983), and the pMC values for IAEA-C7 (49.54 ± 0.22, n = 42) and IAEA-C8 (15.10 ± 0.21, n = 36) also indicated a high degree of consensus with the reference values of 49.53 ± 0.12 and 15.03 ± 0.17 (Le Clercq et al. Reference Le Clercq, van der Plicht and Groening1997), respectively. In addition, the measured δ13C values were also consistent with the reference values for OX-II, IAEA-C7 and IAEA-C8, indicating that the MICADAS was stable and reliable for measurement of the prepared samples. The measured pMC values for OXII, IAEA-C7, and IAEA-C8 by MICADAS at Lanzhou University were found to be comparable with the values obtained by Dongguk University (Republic of Korea) (Lee et al. Reference Lee, Park, Kong and Kim2020), and the Centro Nacional de Aceleradores (Spain) (Guerra et al. Reference Guerra, Arévalo, García and García-Tenorio2019). The background level for the MICADAS at Lanzhou University was also found to be comparable with the AMS value reported by Beijing University (Liu et al. Reference Liu, Ding, Fu, Pan, Wu, Guo and Zhou2007). Also, the δ13C values for the above-mentioned standard samples, obtained by MICADAS, were noted to be close to the reference values, which further confirmed the stability of measurement for MICADAS at Lanzhou University.
The sensitivity of the instrument is reflected by the count rate for the background sample PHA. The long-term mean pMC value for PHA was 0.29 ± 0.07, which corresponds to about 47,000 years. After several continuous runs, the ionizer and lens in the ion source usually become coated with cesium compounds, the reionization of residual carbon of different ages might cause the higher blank. The test results presented in Figure 3 give support to this interpretation of the signal responses. After cleaning the ion source (the insulated section and other parts, and/or replacing with new components), the observed pMC value for PHA usually decreased significantly and the instrumental sensitivity increased (Figure 3). As the test sequence continued, the instrument sensitivity tended to show a clear downward trend with time (Figure 3). Typically, the ion source was cleaned after analyzing 5–10 magazines (or 300–400 hr of ion source contact work) continuously. Additionally, after each reinstallation of the ion source, the cathode position was fine-tuned to ensure that the cesium ion beam was precisely aligned, thus ensuring optimal response and stability of the ion beam.
Inter-Comparison Results
Four sets of inter-comparison materials were measured by AMS at the University of Arizona, Isotoptech Zrt, Peking University, and Lanzhou University, respectively, and the results are listed in Table 2. Given that the charcoal particles may originate from different tree ring positions and the charred seeds may also be a mixture of seeds, both of which may lead to different ages, the charcoals and charred seeds were ground further in an agate mortar to reduce possible inhomogeneities in the particles. The charred seeds and charcoal were pretreated with powdered material (“a”) and raw material (“b”). We finished the all physical treatment and send the samples (bones, powdered materials “a” and particle materials “b”) to the University of Arizona and Isotoptech Zrt, the ABA pretreatment, graphitization and AMS test were conducted by themselves. The other two groups of samples were pretreated and graphitized by the laboratory at Lanzhou University and then tested at Peking University and Lanzhou University.
Inter-comparison results from 4 AMS laboratories (AA: University of Arizona; DeA: Isotoptech Zrt; BA: Peking University; LZU: Lanzhou University).
The results for the three sets of collagen samples analyzed in the four laboratories were in good agreement (Figure 4), the difference between the maximum and minimum values being between 70 and 140 years. The results for the three sets of charred seed materials also showed good agreement (Figure 4), the difference between the maximum and minimum values being between 95 and 150 years. The 14C ages for charcoal 7 and charcoal 8 were also very consistent, the difference between the maximum and minimum values being between 90 and 110 years. The results for charcoal 9 did show significant differences. This particular sample was collected from the Luowalinchang archaeological site in the northeast Tibetan Plateau (Chen et al. Reference Chen, Dong, Zhang, Liu, Jia, An, Ma, Xie, Barton, Ren, Zhao, Wu and Jones2015). Previous 14C dating results indicated that this site experienced long periods of human activity, and the charcoal materials collected from there may not have originated from the same period, and this could account for the significant differences in the test results for the different laboratories.
The 14C ages determined by the four laboratories (the error bar of value is covered for the point size is large).
According to the four sets of data for each sample, we calculated the average values and calculated the difference between each data point and the average values to evaluate the degree of dispersion of the test results, these being shown in Figure 5. Except for samples, charcoal 9a and charcoal 9b, which exhibited very large differences in the pMC deviation values due to sample heterogeneity, inspection of the other data revealed a high degree of consistency, indicating that the MICADAS (Lanzhou University) provided accurate results. It can, therefore, be concluded that the pretreatment and graphitization procedures adopted for use with MICADAS were reliable and fit for purpose. At present, the laboratory can undertake in-house the total AMS 14C dating process consisting of pre-treatment, graphitization, and measurement, the service mainly supporting research in archaeology and paleoenvironments.
The degree of dispersion of the test results for the four laboratories.
CONCLUSION
This paper describes the performance of MICADAS at Lanzhou University and the results for an international inter-comparison study. The results for long-term operation of the AMS instrument demonstrated that our laboratory generated accurate and reliable 14C dating results. Moreover, the pretreatment and graphitization procedures employed produced data that were in agreement with independent protocols used at Arizona University and Debrecen Zrt. Future work will include refinement of current protocols and addition of novel sample preparation methods and gas injection. Overall, the results have demonstrated that the AMS system at Lanzhou University produced reliable radiocarbon data comparable to those of internationally recognized laboratories.
DECLARATION OF COMPETING INTEREST
The authors declare that they have no conflict of interest.
ACKNOWLEDGMENTS
This work was jointly supported by the Program of Introducing Talents of Discipline to University (BP0618001), and the Second Tibetan Plateau Scientific Expedition and Research Program (STEP) (2019QZKK0601).
