Physics & Astronomy

Woods Hole National Ocean Sciences AMS Facility

Brett E. Longworth, Mark L. Roberts, Patricia Long, Karl F. von Reden,

Brad E. Rosenheim, Ernst Galutschek, and Susan Handwork

National Ocean Sciences AMS Facility

Department of Geology and Geophysics

Woods Hole Oceanographic Institution

Woods Hole, MA 02543-1539

The National Ocean Sciences Accelerator Mass Spectrometry (NOSAMS) Facility at the Woods Hole Oceanographic Institution provides radiocarbon analyses to the ocean science research community. Applications include dating of sedimentary deposits, studies of carbon-bearing materials in sediments and marine water column, studies of paleocirculation and ventilation and detailed studies of circulation and carbon cycling in the modern ocean. A gradual shift in sample origin has been experienced over the last five years. The World Ocean Circulation Experiment (WOCE) initially provided almost all of our samples, whereas now samples arrive from a diverse group of client submitters from all of the aforementioned applications.

NOSAMS is now home to two AMS systems dedicated to ¹⁴C: our twenty-year-old 3MV tandetron workhorse, and a newly operational 500kV compact pelletron system. Having two AMS systems will open new avenues of research on the new system and allow a series of upgrades to the tandetron without disrupting sample throughput.

Tandetron

Our AMS system was built by US-AMS Corporation, based on a 3MV Tandetron accelerator with two recombinator/ injector legs sharing a switching magnet. A US-AMS IS-60 sample ion source (GIX 846) is located on the first recombinator/injector leg and the second leg has been dismantled while future development is under consideration. The IS-60 ion source is our production workhorse as progress continues on the developing CFAMS system (see below). The accelerator has been in continuous operation since the last SNEAP meeting. In this time we have run about 9500 graphite targets, of which 5500 were reportable samples (3600 client, 1000 ocean circulation, and 900 outside graphite), and the remainder are divided among standards and R&D samples.

The primary AMS has seen minor improvements and few problems this year, resulting in our highest throughput ever. The main improvements were installation of a new digital source camera system, and solving of several nagging problems with the vacuum system, source dark current, and source electronics. The new camera produces a much more detailed image of the target cartridge during sputtering, making it much easier to adjust the spot size and position while getting a sample run started. Replacement of several aging seals in the source and recombinator has resulted in incremental improvements to source vacuum. One of our Physical Electronics (now Gamma Vacuum) ion pumps was refurbished due to sputtered-through elements. Our problem with cooling water conductivity was solved by a combination of addition of a new deionization unit and a period of downtime for the CFAMS high energy magnet. Our only major failures this year were the aging 386-based PC that runs our sample changer, and the RF terminal charging controller. Building a suitably primitive replacement PC and transferring software was a very involved process, especially since both floppy drives had failed. Check your backups with the drives they need to work on! The RF controller failure was the result of 95° F temperatures in the lab due to an air conditioning failure. Thanks to having a spare on hand, a quick replacement had us operational again in short order.

We have run several experiments with our source this year, including efficiency tests, high current experiments, and trials of new source and target configurations. Running known target masses to exhaustion shows that our total AMS efficiency has improved; we are now at about 1% total efficiency and 2.6% source efficiency for both standard and small-diameter targets. Thanks to incremental improvements to the ion source, our average beam currents during normal operation have increased by about 50% (from ~500nA to ~750nA at the HE ¹²C cup). Modeling predicts that our source should be able to produce 2µA on the high-energy ¹²C cup with small changes in beam energy, but in practice, we have not seen much above 1.2µA. Experiments with front-loaded sample cartridges with a recessed face have confirmed that our source will require modification to the cathode immersion lens to run this style of sample with good results. Due to a client error, we tested the limits of our ¹³C split cup with a sample containing 15 times the natural abundance of ¹³C. This sample showed that our source has very little between-sample memory, but our picoammeters were overloaded, and later inspection showed some melting of nylon parts in the cup.

Continuous Flow Accelerator Mass Spectrometry System

We recently finished building a new AMS system that is primarily designed for the analysis of ¹⁴C from a continuously flowing stream of sample gas. A key component of the new system is a gas-accepting microwave-plasma ion source first built at Atomic Energy of Canada Limited, Chalk River, Ontario, Canada and further developed at NOSAMS. A description of the source was presented at the 2004 SNEAP conference at McMaster. The system also has a NEC 134 sample MC-SNICS for ‘traditional’ AMS measurements. The accelerator is a National Electrostatics Corporation (NEC) model 1.5SDH-1 with a maximum terminal potential of 500 kV.

The compact AMS system is completed, and has been successfully operated with the solid sample sputter source. The performance of the system with the sputter source has readily met expectations with excellent efficiency and ion current. Numerous test-wheels have been analyzed, and measurement precision is state-of-the-art for AMS ¹⁴C measurements: replicate standards agree to order 0.2% in ¹⁴C/¹²C ratio. Moreover, ion currents on small samples (~ 50μg) exceed 60% of full sized (1 mg) samples with very little difference in measured (vs. expected) ¹⁴C/¹²C ratio. Tests currently underway to examine the performance of the new system on even smaller samples have revealed that good counting precision (of order of 1%, corresponding to ~60,000 ¹⁴C counts) can be achieved on samples as small as 20 μg. Such exceptional system performance can be attributed to the large aperture optics that was necessitated by the highly divergent beam coming from the gas ion source.

We anticipate that with modest improvements in source design that are currently underway, we can regularly achieve or even better this performance. These source improvements include an immersion lens and a redesigned vacuum-jacketed Cs reservoir and delivery tube. The cesium focus electrode was replaced with a target shield or immersion lens. Whereas the cesium focus electrode operated at an intermediate voltage between cathode and source potential, the immersion lens operates at cathode potential. The immersion lens provides the same (or perhaps better) functionality as the cesium focus electrode while eliminating one power supply.

Problems during initial operation of CFAMS have been relatively minor. The most troublesome has been failure of the terminal fiber-optic communication bundle, likely due to terminal sparking, and breakage of the lexan driveshaft that provides power to the accelerator terminal. A Group3 Hall probe at the injection magnet has also failed several times, presumably due to sparking from the flight tube to the probe. Two 60kV isolation transformers from Total Recoil Magnetics failed before we switched to a unit from Stangenes, which has been in operation for several months with no trouble. Oil leaks have required us to replace the shaft seals on our forepumps every 6-12 months. Varian is aware of the problem and is looking into a solution. We have been extremely happy with the PXI control hardware from National Instruments. It has performed admirably at source potentials, recovering from source sparks with ease.

An improved gas ion source is currently under construction. This new source will be used to directly monitor the radiocarbon content of a continuously flowing gas stream. The source was designed in conjunction with Atomic Energy of Canada, Ltd., Chalk River, Canada. Design aims were increased ion source efficiency and improved ion optical characteristics compared to the prototype source that has been previously discussed at SNEAP. One notable design feature was the implementation of a tandem accelerating charge-exchange canal, which will permit higher ion beam energy, lower beam divergence, and operation of the source at ground potential. The last attribute enhances the utility and safety of source operation. The current status of the gas ion source is that the main components have been machined, and assembly is now underway. We anticipate initial testing of the gas ion source in the next two to three months.