- Email: [email protected]
Status of the Single Stage AMS machine at Lund University after 4 years of operation
Nuclear Instruments and Methods in Physics Research B 268 (2010) 895–897
Contents lists available at ScienceDirect
Nuclear Instruments and Methods in Physics Research B journal homepage: www.elsevier.com/locate/nimb
Status of the Single Stage AMS machine at Lund University after 4 years of operation Göran Skog *, Mats Rundgren, Pia Sköld Lund University, GeoBiosphere Science Centre, Quaternary Sciences, Geocentrum II, Sölvegatan 12, S-223 62 Lund, Sweden
a r t i c l e
i n f o
Article history: Available online 7 October 2009 Keywords: AMS Radiocarbon dating Single Stage AMS
a b s t r a c t The Lund SSAMS machine has been in routine operation since 2004. We present results from the last year of operation of the facility. The reference sample IAEA C7 and the ‘‘old” oxalic acid, OxI, were used as secondary standards. As primary standard for calculations the OxII was used. The background and long term stability of the facility are discussed. We also report on the quality of the 13C/12C ratio from the SSAMS system. Ó 2009 Elsevier B.V. All rights reserved.
1. Introduction The Lund SSAMS (Single Stage Accelerator Mass Spectrometer) machine has been in routine operation since October 2004. It is equipped with two ion sources and used for radiocarbon dating and for quantification of 14C-labelled substances with applications within radioecology, ecology, medicine and biomedicine. The machine operates at 250 kV with the gas stripper, high energy magnet and detector on a high voltage deck and with the injector and low energy magnet at ground potential. Supplier is National Electrostatic Corp. (NEC), WI, USA. Details of the Lund SSAMS system can be found in [1]. In [1], we reported problems with the Lund SSAMS system due to a current dependence of the stable isotopic ratios. A similar problem is reported from a compact 14C AMS system at Peking University [2]. To make a correct calculation of radiocarbon age it is necessary to normalize for the stable isotope fractionation. As some of the fractionation may originate from the AMS system (for example the ion source) it is important for an AMS system to be able to correctly measure all the three carbon isotopes (12C, 13 C and 14C). For the Lund SSAMS the problem lies in the 12C beam for which the transmission decreases with higher currents and thus the 12C/13C ratio increases with beam current. On the other hand the 14C/13C ratio seems to be independent of current. In [3] is reported that the 14C/13C measurement is reliable even with a 12 C current of 100 lA. In this paper, we show how the current dependence of the 13 12 C/ C ratio can be corrected. We also report on long-term mea-
* Corresponding author. Tel.: +46 46 2227885; fax: +46 46 2224830. E-mail address: [email protected] (G. Skog). 0168-583X/$ - see front matter Ó 2009 Elsevier B.V. All rights reserved. doi:10.1016/j.nimb.2009.10.058
surements of the pMC values for the reference sample IAEA C7 and the ‘‘old” oxalic acid, OxI, using the OxII as primary standard as well as long-term measurements of processed background samples. The results are from the last year of operation of the facility.
2. Results and discussion As seen in Fig. 1 there is a clear dependence of the 13/12 ratio to beam current. The data shown is from a typical run with about 400 measurements from ca. 35 cathodes. Each cathode was measured 11 times for 3 min. Besides the expected spread in 13/12 ratio because of different cathode material there is an overall increase of almost 10‰ per lA 12C+ above 7 lA. However, the trend can be fitted quite well to a second-order polynomial and thus it is possible to de-trend the measured data and normalize the 13/12 ratio. In Table 1, the calculated d13C values for anthracite and three standard samples are shown. The d13C normalization was performed using the OxII as a primary standard. The mean values are quite close to the consensus values showing that the method works properly. In a few cases there are ‘‘outliers”, but they may be explained as a ‘‘bad” cathode (for example, if there has been an isotopic fractionation introduced during the graphitization process) and is not necessarily indicating an error in the 13/12 measurements. During a normal run on the Lund SSAMS the number of 14C events, that are collected from a sample that is close to Modern (for example, OxI), is ca. 80,000. Normalization is made to four OXII samples placed evenly around the cathode wheel and altogether ca. 400,000 events are collected from the primary OxII standards. The error of the background is estimated to be ca. 0.1‰. If the uncertainty in sample, standard and background is taken into account this result in a counting precision of about 4‰ (or ca. ±32
896
G. Skog et al. / Nuclear Instruments and Methods in Physics Research B 268 (2010) 895–897 Table 1 Calculated d13C values for secondary standards and background samples from September 2007 to October 2008. All measurements have been normalized using the OxII standard normalized to d13C = 17.8‰. Date
Fig. 1. The 13C+/12C+ ratio as a function of the 12C+ current. The currents are measured in the Faraday cups positioned immediately after the high energy magnet.
14
C-years). Similarly, the number of collected events for a sample close to one half-life (for example, C7 with a consensus age of 5644 BP) is ca. 40,000, which results in a counting precision of about 5.3‰ (or ca. ±42 14C-years). The internal precision from single runs (with 11 individual measurements for each cathode) mostly agrees with the counting statistics. However, the external precision between different cathodes and between different runs is not automatically at that level. One way to test the accuracy of the measurements is to look at the statistical variations during long-term measurements of secondary standards. Table 2 shows the calculated pMC values for OxI and IAEA-C7 standards during more than 1 year of operation. The calculations were based on measurements of the 14/13 and 13/12 ratios using OxII as primary standard in both cases. The 13/12 ratios were first de-trended and normalized as described above. The agreement between the mean values and the consensus values are excellent. Each data point is the mean value of two cathodes on the sample wheel. Thus the errors should be multiplied by the square root of 2 to represent the error in the measurements of a single cathode. From the table we conclude that the error for OxI is 5.2‰ corresponding to ±42 14C-years and the error for C7 is 6.6‰ corresponding to ±50 14C-years. Thus, in the interval 0–5500 BP we conclude that the precision of the Lund SSAMS machine is close to ±50 14Cyears. This is slightly higher than what we get from counting statistics only (see above). The background for the SSAMS machine is higher than is normally reported for bigger machines. From measurements at our old 3 MV AMS [4] we know that our processed anthracite is at least better than 0.3 pMC. Thus we may conclude that the relatively high background is not introduced during the pre-treatment or graphitization process, but is caused by the machine probably due to ions scattered off residual stripper gas away from the stripper canal. The lower background values seen for the C1 marble could be explained by molecular beams (13CH or 12CH2) of lower intensity for the C1 sample introducing fewer scattered ions. A good vacuum is very important to suppress the background count rate. The Lund SSAMS was delivered with one turbo pump at each ion source, one turbo pump between the low energy electrostatic analyzer and the low energy magnet, two turbo pumps at the stripper and one turbo pump at the high energy electrostatic analyzer (see Ref. [1] for a detailed picture). From the SUERC SSAMS
Anthracite
OxI
C7
C1
07-09-12 07-09-12 07-09-12 07-09-24 07-09-24 07-09-24 07-10-08 07-10-08 07-10-08 07-10-16 07-10-16 07-10-16 07-10-26 07-10-26 07-10-26 07-11-06 07-11-06 07-11-06 07-11-20 07-11-20 07-11-20 07-12-03 07-12-03 07-12-03 07-12-12 07-12-12 07-12-12 08-01-21 08-01-21 08-01-21 08-01-28 08-01-28 08-02-07 08-02-07 08-02-07 08-02-25 08-02-25 08-03-05 08-03-05 08-03-05 08-03-31 08-04-03 08-04-03 08-04-03 08-04-07 08-04-16 08-04-16 08-04-16 08-05-12 08-05-12 08-05-12 08-05-28 08-05-28 08-05-28 08-06-10 08-06-10 08-06-10 08-08-11 08-08-11 08-08-25 08-08-25 08-08-25 08-09-04 08-09-04 08-09-24 08-09-24 08-10-01 08-10-01 08-10-01
25.2 19.9 27.3 22.6 23.0 20.4 25.3 23.1 26.4 29.0 29.3 28.9 25.6 23.7 24.1 24.2 25.7 18.9 25.1 19.6 17.3 24.3 22.0 28.6 24.7 25.0 24.3 23.0
19.4 21.6
10.6 10.1
17.2 17.3
8.7 10.6
19.8 18.9
13.7 13.5
20.1 21.5
14.4 13.8
21.1 22.5
12.3 15.2
22.7 22.7
12.2 15.6
18.7 16.2
9.3 11.1
20.5 23.2
16.6 16.9
16.0 14.2
21.5 19.8
21.4 23.4
17.4 16.9
14.6 23.9 19.4 18.2 24.9 23.9
20.5 21.9 19.7 18.3
16.1 17.5 12.5
18.9 20.6 22.2 22.3
13.0 13.0 13.2 17.9
26.4 26.0 28.2
23.2 14.8
21.0 9.5
30.1 28.9 28.2 23.2 26.0
17.4 13.8
2.5 12.7
20.0 20.1
15.8 11.4
19.3 21.0 23.5 22.7 23.8
17.2 23.0
10.7
22.2 22.0
16.3 19.1
3.0
24.9 28.3 25.2 17.0 26.7 16.4 23.1 20.4 29.8 22.7 24.7 32.6
18.4 27.9 24.9 14.9
12.2 17.2 14.1
0.1
21.0 20.9 23.9 25.7 18.9 20.0
15.3 14.1 18.3 21.0 0.1 10.9
3.0
Mean Standard deviation
24.1 3.7
20.3 3.0
13.9 4.3
1.1 3.0
19
14.5
2.4
Consensus
28.4 24.9 19.1
4.8
0.8
G. Skog et al. / Nuclear Instruments and Methods in Physics Research B 268 (2010) 895–897 Table 2 pMC values for secondary standards and background samples from September 2007 to October 2008. OxII was used as primary standard. Date
Anthracite
OxI
C7
07-09-12 07-09-24 07-10-08 07-10-16 07-10-26 07-11-06 07-11-20 07-12-03 07-12-12 08-01-21 08-01-28 08-02-07 08-02-25 08-03-05 08-03-31 08-04-03 08-04-07 08-04-16 08-05-12 08-05-28 08-06-10 08-08-11 08-08-25 08-09-04 08-09-24 08-10-01
0.86 0.79 0.78 0.82 0.74 0.71 0.98 0.83 0.63 0.63 0.55 0.64 0.64 0.63 0.68 0.69 0.62 0.88 0.79 0.86 0.85 0.72 0.56 0.52 0.60 0.70
104.25 104.23 103.59 103.89 103.52 103.88 104.63 103.73 104.33 104.49 104.48 104.19 104.19 104.14
49.22 49.59 49.49 49.95 49.61 49.59 49.25 49.49 50.03 49.54 49.75 49.65 49.29 49.57
104.40
49.64
103.69 103.58 104.83 104.36 104.41 103.78 104.15 103.41 103.94
49.43 49.19 49.88 49.62 49.51 49.32 49.78 49.49 49.78
Mean Standard deviation
0.72 0.12
104.09 0.38
49.57 0.22
103.98
49.53
Consensus
897
All the turbo pumps in operation: 0.27 pMC. One ‘‘extra” turbo pump (just after the low energy magnet) turned off: 0.36 pMC. Both the ‘‘extra” turbo pumps turned off: 0.72 pMC
C1
The last result is in good agreement with the data in Table 2 indicating that an upgrade of the Lund SSAMS with two extra turbo pumps will improve the background. 3. Summary
0.40 0.47 0.46 0.66 0.82 0.56 0.18
To date (October 2008), about 2700 samples for radiocarbon dating have been analyzed at the Lund SSAMS. At present the performance of the system is as follows: precision (for samples younger than 5000 radiocarbon years): ±50 14C-years, background for processed anthracite: ca. 40,000 BP, minimum sample weight: ca. 0.25 mg of carbon. The background of the Lund SSAMS is higher than for larger AMS systems. An upgrade of the system with two more turbo pumps, which is planned for the winter 2008–2009, will improve the background. Note added in proof After the upgrade we now (Oct 2009) measure a background of ca. 0.35 pMC or ca. 45000 BP for processed anthracite. Acknowledgements The equipment was funded by the Knut and Alice Wallenberg Foundation and by the Swedish Research Council.
machine is reported much lower background values [3]. Their machine is equipped with two ‘‘extra” turbo pumps (just after the low energy magnet and just before the high energy magnet). A test run with a SSAMS machine at the NEC factory [5] equipped with the ‘‘extra” turbo pumps gave the following results for the alpha machine blank:
References [1] [2] [3] [4]
G. Skog, Nucl. Instr. and Meth. B 259 (2007) 1. K. Liu et al., Nucl. Instr. and Meth. B 259 (2007) 23–26. S. Freeman et al., Nucl. Instr. and Meth. B 266 (2008) 2225–2228. G. Skog et al., in: Symposium of North Eastern Accelerator Personnel, Lund, Sweden, 22–25 October 2001, SNEAP XXXIV, 2002, p. 48. [5] R. Loger, personal communication.
Contents lists available at ScienceDirect
Nuclear Instruments and Methods in Physics Research B journal homepage: www.elsevier.com/locate/nimb
Status of the Single Stage AMS machine at Lund University after 4 years of operation Göran Skog *, Mats Rundgren, Pia Sköld Lund University, GeoBiosphere Science Centre, Quaternary Sciences, Geocentrum II, Sölvegatan 12, S-223 62 Lund, Sweden
a r t i c l e
i n f o
Article history: Available online 7 October 2009 Keywords: AMS Radiocarbon dating Single Stage AMS
a b s t r a c t The Lund SSAMS machine has been in routine operation since 2004. We present results from the last year of operation of the facility. The reference sample IAEA C7 and the ‘‘old” oxalic acid, OxI, were used as secondary standards. As primary standard for calculations the OxII was used. The background and long term stability of the facility are discussed. We also report on the quality of the 13C/12C ratio from the SSAMS system. Ó 2009 Elsevier B.V. All rights reserved.
1. Introduction The Lund SSAMS (Single Stage Accelerator Mass Spectrometer) machine has been in routine operation since October 2004. It is equipped with two ion sources and used for radiocarbon dating and for quantification of 14C-labelled substances with applications within radioecology, ecology, medicine and biomedicine. The machine operates at 250 kV with the gas stripper, high energy magnet and detector on a high voltage deck and with the injector and low energy magnet at ground potential. Supplier is National Electrostatic Corp. (NEC), WI, USA. Details of the Lund SSAMS system can be found in [1]. In [1], we reported problems with the Lund SSAMS system due to a current dependence of the stable isotopic ratios. A similar problem is reported from a compact 14C AMS system at Peking University [2]. To make a correct calculation of radiocarbon age it is necessary to normalize for the stable isotope fractionation. As some of the fractionation may originate from the AMS system (for example the ion source) it is important for an AMS system to be able to correctly measure all the three carbon isotopes (12C, 13 C and 14C). For the Lund SSAMS the problem lies in the 12C beam for which the transmission decreases with higher currents and thus the 12C/13C ratio increases with beam current. On the other hand the 14C/13C ratio seems to be independent of current. In [3] is reported that the 14C/13C measurement is reliable even with a 12 C current of 100 lA. In this paper, we show how the current dependence of the 13 12 C/ C ratio can be corrected. We also report on long-term mea-
* Corresponding author. Tel.: +46 46 2227885; fax: +46 46 2224830. E-mail address: [email protected] (G. Skog). 0168-583X/$ - see front matter Ó 2009 Elsevier B.V. All rights reserved. doi:10.1016/j.nimb.2009.10.058
surements of the pMC values for the reference sample IAEA C7 and the ‘‘old” oxalic acid, OxI, using the OxII as primary standard as well as long-term measurements of processed background samples. The results are from the last year of operation of the facility.
2. Results and discussion As seen in Fig. 1 there is a clear dependence of the 13/12 ratio to beam current. The data shown is from a typical run with about 400 measurements from ca. 35 cathodes. Each cathode was measured 11 times for 3 min. Besides the expected spread in 13/12 ratio because of different cathode material there is an overall increase of almost 10‰ per lA 12C+ above 7 lA. However, the trend can be fitted quite well to a second-order polynomial and thus it is possible to de-trend the measured data and normalize the 13/12 ratio. In Table 1, the calculated d13C values for anthracite and three standard samples are shown. The d13C normalization was performed using the OxII as a primary standard. The mean values are quite close to the consensus values showing that the method works properly. In a few cases there are ‘‘outliers”, but they may be explained as a ‘‘bad” cathode (for example, if there has been an isotopic fractionation introduced during the graphitization process) and is not necessarily indicating an error in the 13/12 measurements. During a normal run on the Lund SSAMS the number of 14C events, that are collected from a sample that is close to Modern (for example, OxI), is ca. 80,000. Normalization is made to four OXII samples placed evenly around the cathode wheel and altogether ca. 400,000 events are collected from the primary OxII standards. The error of the background is estimated to be ca. 0.1‰. If the uncertainty in sample, standard and background is taken into account this result in a counting precision of about 4‰ (or ca. ±32
896
G. Skog et al. / Nuclear Instruments and Methods in Physics Research B 268 (2010) 895–897 Table 1 Calculated d13C values for secondary standards and background samples from September 2007 to October 2008. All measurements have been normalized using the OxII standard normalized to d13C = 17.8‰. Date
Fig. 1. The 13C+/12C+ ratio as a function of the 12C+ current. The currents are measured in the Faraday cups positioned immediately after the high energy magnet.
14
C-years). Similarly, the number of collected events for a sample close to one half-life (for example, C7 with a consensus age of 5644 BP) is ca. 40,000, which results in a counting precision of about 5.3‰ (or ca. ±42 14C-years). The internal precision from single runs (with 11 individual measurements for each cathode) mostly agrees with the counting statistics. However, the external precision between different cathodes and between different runs is not automatically at that level. One way to test the accuracy of the measurements is to look at the statistical variations during long-term measurements of secondary standards. Table 2 shows the calculated pMC values for OxI and IAEA-C7 standards during more than 1 year of operation. The calculations were based on measurements of the 14/13 and 13/12 ratios using OxII as primary standard in both cases. The 13/12 ratios were first de-trended and normalized as described above. The agreement between the mean values and the consensus values are excellent. Each data point is the mean value of two cathodes on the sample wheel. Thus the errors should be multiplied by the square root of 2 to represent the error in the measurements of a single cathode. From the table we conclude that the error for OxI is 5.2‰ corresponding to ±42 14C-years and the error for C7 is 6.6‰ corresponding to ±50 14C-years. Thus, in the interval 0–5500 BP we conclude that the precision of the Lund SSAMS machine is close to ±50 14Cyears. This is slightly higher than what we get from counting statistics only (see above). The background for the SSAMS machine is higher than is normally reported for bigger machines. From measurements at our old 3 MV AMS [4] we know that our processed anthracite is at least better than 0.3 pMC. Thus we may conclude that the relatively high background is not introduced during the pre-treatment or graphitization process, but is caused by the machine probably due to ions scattered off residual stripper gas away from the stripper canal. The lower background values seen for the C1 marble could be explained by molecular beams (13CH or 12CH2) of lower intensity for the C1 sample introducing fewer scattered ions. A good vacuum is very important to suppress the background count rate. The Lund SSAMS was delivered with one turbo pump at each ion source, one turbo pump between the low energy electrostatic analyzer and the low energy magnet, two turbo pumps at the stripper and one turbo pump at the high energy electrostatic analyzer (see Ref. [1] for a detailed picture). From the SUERC SSAMS
Anthracite
OxI
C7
C1
07-09-12 07-09-12 07-09-12 07-09-24 07-09-24 07-09-24 07-10-08 07-10-08 07-10-08 07-10-16 07-10-16 07-10-16 07-10-26 07-10-26 07-10-26 07-11-06 07-11-06 07-11-06 07-11-20 07-11-20 07-11-20 07-12-03 07-12-03 07-12-03 07-12-12 07-12-12 07-12-12 08-01-21 08-01-21 08-01-21 08-01-28 08-01-28 08-02-07 08-02-07 08-02-07 08-02-25 08-02-25 08-03-05 08-03-05 08-03-05 08-03-31 08-04-03 08-04-03 08-04-03 08-04-07 08-04-16 08-04-16 08-04-16 08-05-12 08-05-12 08-05-12 08-05-28 08-05-28 08-05-28 08-06-10 08-06-10 08-06-10 08-08-11 08-08-11 08-08-25 08-08-25 08-08-25 08-09-04 08-09-04 08-09-24 08-09-24 08-10-01 08-10-01 08-10-01
25.2 19.9 27.3 22.6 23.0 20.4 25.3 23.1 26.4 29.0 29.3 28.9 25.6 23.7 24.1 24.2 25.7 18.9 25.1 19.6 17.3 24.3 22.0 28.6 24.7 25.0 24.3 23.0
19.4 21.6
10.6 10.1
17.2 17.3
8.7 10.6
19.8 18.9
13.7 13.5
20.1 21.5
14.4 13.8
21.1 22.5
12.3 15.2
22.7 22.7
12.2 15.6
18.7 16.2
9.3 11.1
20.5 23.2
16.6 16.9
16.0 14.2
21.5 19.8
21.4 23.4
17.4 16.9
14.6 23.9 19.4 18.2 24.9 23.9
20.5 21.9 19.7 18.3
16.1 17.5 12.5
18.9 20.6 22.2 22.3
13.0 13.0 13.2 17.9
26.4 26.0 28.2
23.2 14.8
21.0 9.5
30.1 28.9 28.2 23.2 26.0
17.4 13.8
2.5 12.7
20.0 20.1
15.8 11.4
19.3 21.0 23.5 22.7 23.8
17.2 23.0
10.7
22.2 22.0
16.3 19.1
3.0
24.9 28.3 25.2 17.0 26.7 16.4 23.1 20.4 29.8 22.7 24.7 32.6
18.4 27.9 24.9 14.9
12.2 17.2 14.1
0.1
21.0 20.9 23.9 25.7 18.9 20.0
15.3 14.1 18.3 21.0 0.1 10.9
3.0
Mean Standard deviation
24.1 3.7
20.3 3.0
13.9 4.3
1.1 3.0
19
14.5
2.4
Consensus
28.4 24.9 19.1
4.8
0.8
G. Skog et al. / Nuclear Instruments and Methods in Physics Research B 268 (2010) 895–897 Table 2 pMC values for secondary standards and background samples from September 2007 to October 2008. OxII was used as primary standard. Date
Anthracite
OxI
C7
07-09-12 07-09-24 07-10-08 07-10-16 07-10-26 07-11-06 07-11-20 07-12-03 07-12-12 08-01-21 08-01-28 08-02-07 08-02-25 08-03-05 08-03-31 08-04-03 08-04-07 08-04-16 08-05-12 08-05-28 08-06-10 08-08-11 08-08-25 08-09-04 08-09-24 08-10-01
0.86 0.79 0.78 0.82 0.74 0.71 0.98 0.83 0.63 0.63 0.55 0.64 0.64 0.63 0.68 0.69 0.62 0.88 0.79 0.86 0.85 0.72 0.56 0.52 0.60 0.70
104.25 104.23 103.59 103.89 103.52 103.88 104.63 103.73 104.33 104.49 104.48 104.19 104.19 104.14
49.22 49.59 49.49 49.95 49.61 49.59 49.25 49.49 50.03 49.54 49.75 49.65 49.29 49.57
104.40
49.64
103.69 103.58 104.83 104.36 104.41 103.78 104.15 103.41 103.94
49.43 49.19 49.88 49.62 49.51 49.32 49.78 49.49 49.78
Mean Standard deviation
0.72 0.12
104.09 0.38
49.57 0.22
103.98
49.53
Consensus
897
All the turbo pumps in operation: 0.27 pMC. One ‘‘extra” turbo pump (just after the low energy magnet) turned off: 0.36 pMC. Both the ‘‘extra” turbo pumps turned off: 0.72 pMC
C1
The last result is in good agreement with the data in Table 2 indicating that an upgrade of the Lund SSAMS with two extra turbo pumps will improve the background. 3. Summary
0.40 0.47 0.46 0.66 0.82 0.56 0.18
To date (October 2008), about 2700 samples for radiocarbon dating have been analyzed at the Lund SSAMS. At present the performance of the system is as follows: precision (for samples younger than 5000 radiocarbon years): ±50 14C-years, background for processed anthracite: ca. 40,000 BP, minimum sample weight: ca. 0.25 mg of carbon. The background of the Lund SSAMS is higher than for larger AMS systems. An upgrade of the system with two more turbo pumps, which is planned for the winter 2008–2009, will improve the background. Note added in proof After the upgrade we now (Oct 2009) measure a background of ca. 0.35 pMC or ca. 45000 BP for processed anthracite. Acknowledgements The equipment was funded by the Knut and Alice Wallenberg Foundation and by the Swedish Research Council.
machine is reported much lower background values [3]. Their machine is equipped with two ‘‘extra” turbo pumps (just after the low energy magnet and just before the high energy magnet). A test run with a SSAMS machine at the NEC factory [5] equipped with the ‘‘extra” turbo pumps gave the following results for the alpha machine blank:
References [1] [2] [3] [4]
G. Skog, Nucl. Instr. and Meth. B 259 (2007) 1. K. Liu et al., Nucl. Instr. and Meth. B 259 (2007) 23–26. S. Freeman et al., Nucl. Instr. and Meth. B 266 (2008) 2225–2228. G. Skog et al., in: Symposium of North Eastern Accelerator Personnel, Lund, Sweden, 22–25 October 2001, SNEAP XXXIV, 2002, p. 48. [5] R. Loger, personal communication.