“…[samples were] reduced to graphite before AMS 14C analysis at the SUERC AMS Facility.”

To determine the rate of ice sheet retreat it is necessary to determine WHERE the ice sheet margin was WHEN. The WHERE is determined by studying landforms characteristic of glacier margins, such as moraines, erratic boulders, and sediment deposited both onshore and offshore by ice and glacial meltwater near the margins of the ice sheet. The WHEN is determined using different dating techniques depending on the material sampled. For BRITICE-CHRONO we are applying three different dating techniques, (1) optically stimulated luminescence dating (OSL) for sands, (2) radiocarbon dating (¹⁴C) for organic material like shell fragments picked out of sediment cores collected from the sea floor during the research cruises, and (3) surface exposure dating of rock using the terrestrial cosmogenic isotope beryllium-10 (¹⁰Be) in the mineral quartz.

Radiocarbon dating and surface exposure dating rely on being able to measure the abundance of extremely rare radioisotopes in the sample material using a technique called accelerator mass spectrometry (AMS) performed at the Scottish Universities Environmental Research Centre (SUERC) AMS Laboratory. What is extremely rare? Well, the natural abundance of ¹⁴C in modern carbon is 1 part per trillion (10^-12) which gets smaller with the age of the material being dated because ¹⁴C decays after the death of the organism. The ratio between the radioisotope ¹⁰Be and stable beryllium is roughly 10^-13 to 10^-14 for surface exposure dating of BRITICE-CHRONO samples. To put this in perspective, if you counted at one number per second it would take you about 3.25 million years to count to 10¹⁴. In other words, radiocarbon and surface exposure dating is only possible because we are able to count individual radioisotopes among trillions of almost identical stable isotopes.

In scientific papers these  measurements are often summarised in a few words, such as “…[samples were] reduced to graphite before AMS ¹⁴C analysis at the SUERC AMS Facility” or “¹⁰Be/⁹Be ratios were derived from measurements at the SUERC AMS Laboratory”. Neither sentence does justice to the complexity of the method and the efforts of a dedicated team of scientists and technicians. Here I will try to provide some insight into how the concentrations of the extremely low quantities of ¹⁴C are determined using accelerator mass spectrometry (AMS). For ¹⁰Be the process is similar but explaining the differences would add unnecessary complexity in what follows.

Accelerator mass spectrometry (AMS) is an ultra-sensitive technique for isotopic analysis in which atoms extracted from a sample are ionized (the process by which an atom or a molecule acquires a negative or positive charge by gaining or losing electrons to form ions); accelerated to high energies; separated according to their momentum, charge, and energy; and then individually counted after identification as having the correct atomic number and mass. The principle difference between AMS and conventional mass spectrometry (MS) lies in the energies to which the ions are accelerated. In MS the energies are thousands of electron volts (1 keV = 1.6 x 10^-16 J), whereas in AMS they are millions of electron volts (MeV). The practical consequences of having higher energies is that ambiguities in identification of atomic and molecular ions with the same mass are removed.

This is how we make the measurements at SUERC

Fig 1. Cathodes in sample wheel after measurement. The hole in the cathodes contains the sample and is 1mm wide.

Samples come to the AMS Laboratory in the form of pressed cathodes containing a few milligrams of graphite (for radiocarbon) and BeO mixed with Niobium for ¹⁰Be surface exposure dating. It takes a lot of time and effort to get from a sample collected in the field to the cathode stage, but that is another story.

The cathodes are loaded into a 134-position sample wheel together with standard materials that have known isotope ratios.

Fig 2. Sample wheel loaded in ion source.

The wheel is loaded into the ion source, the ion source is closed and pumped down to the same very high vacuum as the remainder of the beam line (steel tube) through which the ions produced in the ion source are going to travel (the correct term is drift, but it does not really convey the speed at which the ions move). We need a very high vacuum (comparable to conditions found in outer space) because we do not want our ions to collide with neutral atoms or molecules drifting anywhere along the 32 metres of beam line.

Fig 3. Section through ionizer

We are now ready to perform some ion sorcery. We heat up a caesium reservoir to generate a Cs vapour in the space between the cathode holding our sample and a heated ionizing surface. Some of the vapour condenses onto the cooled cathode, some is ionised by the ionizer creating Cs⁺ ions that are accelerated and focused towards the sample because the sample cathode is at -5kV compared to the ionizer (Fig 4). The impact of the Cs+ ions on the sample surface causes sputtering of particles from the sample surface.

Fig 4. Cs+ focus lens at front of ioniser (Fig. 3). The sample cathode being ionised is located 1 mm in front of the hole in the lens

Some materials will preferentially sputter negative ions. Other materials will preferentially sputter neutral or positive particles, which pick up electrons as they pass through the condensed caesium layer, producing negative ions. All of this happens within 1 mm of the sample surface. The negative ions are removed from the sample surface and focused into the beam line by an extraction electrode (extractor) set at 15 kV. The extracted ion beam (shown in grey in Fig. 3) is spreading, just like the spreading of light from a torch (which is just another type of particle beam). This is known as beam emittance and it must be kept small to ensure high transmission to the detector. But I am getting way ahead of myself safe to say that we cannot get the emittance of the beam back to the original 1mm diameter at the sample surface and much of the experimental apparatus of the AMS is designed to focus the beam to specific places along the beam line.

Fig 5. Closed ion source. The metal rings around the beam line are part of the 45 kV injector. The apparatus to the left of the rings are for pumping the beam line and and focussing the beam.

So now we have extracted a negative ion beam from our sample material, and the ions are starting to move along the beam line. We accelerate the particles to 66 keV by exposing them to an additional 45kV in the injector (Fig. 5 and Fig 6). The particles pass through a spherical electrostatic analyser (ESA, Fig. 6), where particles with the incorrect energy over charge (E/q) are removed from the beam, before being injected into the accelerator via the injection magnet.

Fig 6. Schematic of SUERC 5MV tandem spectrometer.

The injection magnet separates particles based on momentum (= mass x velocity). For radiocarbon we set the magnetic field to allow particles with mass 14 through the magnet and into the accelerator, but we also need to put carbon-12 and carbon-13 into the accelerator to get the ratio ¹⁴C/¹²C and ¹⁴C/¹³C. Since both ¹²C and ¹³C have lower mass than ¹⁴C we use a magnet bouncing system (MBS; Fig. 6) to give the ion beam more energy so that ¹²C and ¹³C temporarily behave like ¹⁴C. Because ¹²C and ¹³C are much more abundant than ¹⁴C we inject the former two for only a few microseconds per second. While ¹⁴C enters the accelerator we measure the ¹³C current in a low energy Faraday cup and the same is the case for ¹²C when ¹³C is injected (Fig. 6. Inset A). ¹⁴C cannot be measured in a Faraday cup because there are far too few atoms to generate a current.

Unfortunately sample materials are not pure and therefore ion sources do not only produce the ions we want. However the next stage in the process takes care of many of the molecular isobars (= same atomic mass number) such as the hydrocarbons ¹²CH₂^– and ¹³CH^–, which have the same atomic mass number as ¹⁴C and are therefore also injected into the accelerator.

So far the system described is similar to conventional mass spectrometry. The next stage in the ion transport is what sets AMS apart. Up to now the ions have been energised by the ion source and injector to 66keV (the low energy end of the spectrometer), which means they are travelling at roughly 1000 km/s. Next they are accelerated to a speed of roughly 7500 km/s by exposing the negatively charged ions to a 4.5 MV positive charge at the terminal in the centre of the 8 m long accelerator tank (Fig. 6 & 7).

Fig 7. Accelerator pressure vessel known as the tank. It is filled with an insulating gas.

When the negatively charged ions reach the terminal they pass through a gas stripper (Fig. 6). The collisions between the ions and the gas removes electrons from the ions thereby changing them from being negatively charged to being positively charged. The terminal voltage is set to remove at least three electrons because by this process molecular isobars of ¹⁴C (such as ¹²CH₂^– or ¹³CH^–) are destroyed due to the high instability of their positively charged forms, and atomic C⁺ ions such as ¹²C⁺, ¹³C⁺, and ¹⁴C⁺ can be separated due to their different mass to charge ratios. Once the now positive atomic ions emerge from the stripper they find themselves next to a very high positive charge (4.5 MV in the case of radiocarbon) and they accelerate away from this, hence the term tandem accelerator (two acceleration steps). The particles emerge from the tank at a velocity of greater than 17000 km/s (that’s equivalent to traveling around the Earth at the equator in just over 2 seconds), the high-energy part of the AMS.

The particles now enter another mass spectrometer, the analysing magnet (Fig. 6), where the ¹²C⁺, ¹³C⁺, and ¹⁴C⁺ are separated according to momentum. ¹²C⁺ and ¹³C⁺ currents are bent more than ¹⁴C⁺ and are collected and measured in Faraday cups, while ¹⁴C⁺ is allowed to continue on towards the gas ionisation detector. Even after all of this the ion beam still contains ions with incorrect mass, energy, or charge as a result of energy- or charge-changing collisions with system components or residual gas. Unfortunately these interferences mimic the path of the ions of interest. Some of them are removed by another electrostatic analyser (ECA) before the remaining particles arrive at the gas ionisation detector where the final identification and counting of atoms takes place (Fig. 8).

Fig 8. Detector (foreground) and electrostatic analyser (background).

As the ions enter the detector they are slowed down and stopped by passing through a gas. Each atom ionises some of the gas and the resulting electrons are collected, amplified, and digitised. In this way the path and location of each atom arriving in the detector can be determined and each arrival counted. The gas ionisation detector allows us to determine the atom species because heavier atoms travel further and deposit more energy than lighter atoms. This capability allows us to set the electronics to separately count ¹⁴C atoms and different interferences arriving in the detector (Fig 9).

Fig 9. Detector spectrum of one 6 minute measurement. Red dots (n=40933) are counted 14C events. Black dots (n=430) are scattered 14C and Lithium atoms (labelled) that have made it into the detector.

To make all of this possible requires every component of the accelerator mass spectrometer to operate together within very tight tolerances. Even when everything is working well, each time a new sample wheel is introduced into the ion source the conditions inside the ion source vary slightly and the machine has to be very carefully tuned to these new conditions prior to commencing the measurement of the precious samples. Once satisfied the AMS is operating within the necessary limits each sample is measured until measurement statistics are met (usually within 8 measurements), or the sample is exhausted. Each individual measurement lasts about 6 minutes. For high precision measurements it is not unusual to measure a sample for a total of 1 to 1.5 hours. Thus to complete the measurement of a sample wheel is a multi-day undertaking during which the AMS has to be continually monitored for any changes in the condition of multiple machine components that could compromise measurement integrity.

The ¹⁴C/¹³C ratio for each measurement is derived from the counted ¹⁴C atoms in the sample divided by the number of ¹³C atoms calculated from the ¹³C current in the high energy Faraday cup. The final ¹⁴C/¹³C ratio for each sample is the average of the combined ¹⁴C/¹³C ratio measurement for the sample, normalised to known standard materials that were measured in the same wheel. It is this final ¹⁴C/¹³C ratio that is used to calculate the radiocarbon age for the sample.

I hope the above summary provides a little bit of insight into what is meant when you come across sentences like “…AMS ¹⁴C analysis were made at the SUERC AMS Facility” in the literature.

Derek

An Engineers Apprentice

By Claire Mellett and Jenny Gales (British Geological Survey)

A typical day for us usually involves sitting behind a desk staring at a computer in the basement of the BGS Edinburgh office. As marine geologists we are tasked with mapping the seabed and sub-seabed for government and commercial interests. Fundamental to this is an understanding of how geological processes such as ice, rivers, wind, waves and tides have shaped the seabed over long periods of time. Our field area is inaccessible to us as it is drowned beneath sometimes thousands of metres of water and we rely on remote sensing data such as bathymetry and seismic to image the seafloor and make our interpretations. Once we have “guestimated” geological conditions we need to prove them with physical samples and this is where the BGS Marine Operations team comes in.

When carrying out our own research we focus on finding the most suitable site that will provide an answer to whatever question we are asking and we don’t spend too much time thinking about how the sample is recovered. Luckily we have a BGS Marine Operations team comprising electrical, mechanical and design engineers that can build and adapt equipment to meet our expectations. However, ignorance isn’t always bliss and by understanding how different rigs work and the logistics involved in transporting, fitting and fixing equipment on different vessels all around the world, we will have knowledge of how our data was collected and the limitations of its use. Claire: “I thought I would be fairly useless as a member of the operations team given that I am a typical “pen pusher” but I went in with an open mind willing to try anything. As the weeks have gone by I find it easier to lift the barrels meaning I must be getting stronger. I also seem to have started a scrap metal collection as I keep finding bolts and washers in all my pockets. This apparently proves your worth an engineer (according to Garry, one of the BGS engineers)”.

This slideshow requires JavaScript.

After a while we decided to formalise our training so made ourselves engineer’s apprentices. As part of the apprenticeship we came up with list of skills that needed to be developed. These include tasks like winch operation (which is the most stressful part of the apprenticeship), vibrocore assembly, vessel awareness (Claire: “I can now distinguish the bulkhead from the deckhead”), health and safety and vibrocore driving. This last skill is obviously very important as when carrying out this task you get the comfiest seat in the container right next to the heater (which also reclines for when you’re on night shift). Additional skills every seafaring apprentice must have include coffee and tea making (including biscuit acquisition) to keep the team going on twelve hour shifts, rope skills (Jenny: “we can now both tie a rolling hitch with two half hitches to get the core liner out of the barrel”) and radio etiquette which varies greatly depending on accents. The final part of the training is tool recognition. We are getting good at this although there appears to be a nomenclature issue depending what tradesman you get e.g. a “toffee hammer” is apparently the same as a “quarter pound ball pein hammer”.

This slideshow requires JavaScript.

We do have it easy in comparison to the rest of the BGS engineering team as when things break down, which is likely to happen when you’re at sea for a long time, they surprise us by just fixing things. As engineers, this is their job, but it still gets us each time they make something work. For example, we are running low on core catchers as the geology keeps destroying them so we decided to just make some. A bit of improvisation and some welding and we have a new supply of core catchers, voilà!

The work day for an engineer’s apprentice is so refreshing yet tiring. We are outside all day which is delightful when the sun (or moon) is reflecting off a reasonable calm sea with big white fluffy clouds on the horizon. Even when the rain is horizontal and the waves are crashing over the deck, we still look forward to getting out to work. When compared to the often solitary life of a scientist where you exist in your ideas, it is a welcome change to be working outside as part of a team of engineers and ship’s crew physically collecting the scientific data you spend most our time working on. Claire: “I must add here that the ship’s crew on board are all extremely patient with helping us in our training (especially when it comes to winch operation!)”.

We keep being asked if we prefer being a scientist or part of the operations team on a research cruise and it’s a difficult question to answer. Claire: “It is a bit of a holiday for me being an engineer’s apprentice as it is not my profession, therefore all the pressure is on our trainers (Iain and Mike’s) shoulders. I do appear to spend a large part of the day laughing (usually at myself) which is a sign I am enjoying the work. However, if I had a 90 m research vessel at my disposal, as a scientist, I can only imagine the fun I would have!”.

(Selected photography by Alex Ingle)

A room without a view……

By Elke Hanenkamp (MSCL Operator)

Enter my lair

Six o’clock in the morning on board the RRS James Cook somewhere on the edge of Malin Sea in 1500m of water, and my shift as the MSCL operator starts right now. The dayshift (midday to midnight) is still fast asleep and the nightshift (midnight to midday) scientists are eagerly (or maybe more fatalistically) awaiting my arrival. The beginning of my shift marks the start for them that cores can finally be split and described soon (meaning more work for them), therefore I have been jokingly nicknamed “the harbinger of cores”.

My role during this expedition is to collect physical properties data (density, porosity etc) from the vibro and piston cores before they are split on board. I am operating a Geotek Multi-Sensor Core Logger (MSCL) in a containerised lab (also known as “the container cave”, I am in there all the time holed up with the cores). So the obvious question is – what is happening behind the closed door of the container? After the cores come aboard, they are cut into sections and labelled, and then stored for at least 6 hours inside the container to equilibrate to ambient temperature. Only after this period, the cores will be measured on the MSCL, because some of the sensors are temperature sensitive. It is not possible to prop the door open during the measurements, fluctuations in temperature would influence the data. That’s why I am holed up in the container most of the time, every so often delivering already measured cores to the scientists for splitting or taking newly labelled cores into the container.

The Multi Sensor Core Logger is a quite versatile core measurement system, equipped with four sensors – Gamma Density, P-Wave Velocity, Non-Contact Resistivity and Magnetic Susceptibility. While the core is pushed past the stationary sensors, it is scanned, and data from all four sensors is collected at once when the core pauses at a measurement point (in this case every 2 cm). Sequential core sections are loaded on to the logger, this way a complete core can be logged in a continuous process while the data is displayed graphically in real time on the computer. Typically, with measurements being done every 2 cm, a 1 m section can be logged within 15 min, but overall measurement time for one whole core depends on the amount and length of each individual section the core is cut into earlier. The shortest core section we had so far measured only 21 cm. The amount of cores sections measured each day highly varies, but a couple of days ago, 45 sections were measured on the MSCL within my 12 hour-shift, with a total length of a little bit over 41 m (a new record).

This slideshow requires JavaScript.

The MSCL gives us a non-destructive way of analysing cores before they are split and sampled. The measurements can help to characterise the physical nature of the individual cores, e.g. lithology, density, porosity, and will be used in combination with core descriptions and various geochronological data to better understand the timing of ice sheet recession. The high-resolution dataset from the MSCL should also allow us to make correlations between individual core sites in the Celtic, Irish and Malin Seas fringing the North Atlantic.

A view of the world

Cruise 1: meet our Marine Mammal Observer

By Marian McGrath

Hello everyone! It’s been one week since we joined the RRS James Cook in Southampton, even though we didn’t actually leave the port till Friday the 18th due to technical problems with the vibro corer. My role on board is as the Marine Mammal Observer (MMO). The role of the MMO is to ensure the safety and protection of marine mammals from man-made noise pollution in the ocean. This can damage or kill cetaceans which have very sensitive hearing. The Marine Mammal Observer (MMO) is required by law to be aboard any vessel which is carrying out seismic surveys within Irish waters. On this vessel, Sub Bottom Profiler seismic equipment and Multibeam echosounder equipment are being used. In unprotected marine areas an MMO is required to carry out a 30 minute pre Multibeam echo sounder and Sub Bottom Profiler watch followed by a 20 minute watch during the soft start. Sound activity cannot commence until the MMO gives clearance after the 30 minute watch. If marine mammals are spotted within 500m range of the equipment during this watch then a further 30 minute watch is undertaken till marine mammals have left the mitigation zone. If no marine mammals were seen within this time then a soft start would commence. Once the ramp up procedure is started there is no need to stop the equipment during night time hours. The Multibeam and Pinger systems remain active during the survey unless we are on a coring station for longer than an hour in which case they are switched off. They are also turned off during the mid-cruise port call in Killybegs.

This slideshow requires JavaScript.

Marine mammal observations are carried out from the bridge. This gives the best view point of both sides and in front of the vessel. The equipment is always started during daylight hours to allow for MMO watches to be carried out prior to soft starts. Observations are undertaken using a reticular binoculars, a range finder and also by the naked eye. Distance to marine mammals is determined using this reticular binoculars and height above sea level. To determine the range one of the divisions present in the binoculars is placed on the horizon. A formula is then used to determine the distance of the mammal from the ship. The formula is: Distance (m) = (height of eye above sea level (m) x 1000/ no. of mils down from horizon). Throughout the duration of the survey, watches are undertaken throughout the day and any sightings are logged in a computer supplied by The Irish Whale and Dolphin Group. This will feed into a database which is constantly updated regarding location and numbers of various species. Throughout the day recordings are taken of precipitation, sea state, visibility, ship speed, water depth, cloud cover, latitude and longitude, wind speed and direction. So far on this survey Common Dolphins have been seen near the shelf edge of the Celtic Sea. First 4 adult dolphins were seen on the 21st July and later the same day 11 adults and one calf were seen.

Rapid retreat of the Irish Sea Ice Stream – just out in the Journal of Quaternary Science

A new paper has just been published by Richard Chiverrell and a hefty team of Britice-Chrono co-workers (James Scourse, Katrien van Landeghem, Chris Clark, Colm O Cofaigh, Dave Evans, Danny Mccarroll, Colin Ballantyne) presenting the first Bayesian integration and modelling of all the dating control for the marine sectors of the largest ice stream that the last British-Irish Ice Sheet ~ 24,000 years ago. The modelling shows very rapid retreat for this marine-terminating ice stream over greater distances (650 km) and timescales (8000 years) than is available from short term (decadal) observations of present day ice stream margins. The modelling shows this retreat 24,000 years ago was rapid and linked with climatic warming, sea-level rise, mega-tidal amplitudes and reactivation of meridional circulation in the North Atlantic. But, significantly the pattern of retreat appears uneven with a pulsed pattern of retreat attributed to the passage of the ice stream between normal (sloping away from the ice margin) and adverse (sloping towards) ice bed gradients and changes in the geometry or marginal constriction of the ice stream. To read more click here.

The methodology and application kind of formed an important test case for Britice-Chrono as we attempt to constrain rates of and controls on marine ice stream retreat over millennial timescales for eight ice stream radiating out from the last British-Irish Ice Sheet. The methodology outlined in the paper will underpin and be used as a guide for our data collection for the wider British-Irish Ice Sheet. It would be quite good fun to play around with some of the available chronology for other ice streams…..

For news and updates on Britice-Chrono see our Twitter site and everyone please get tweeting or twittering!.

Chris helps the Yorskhire Post make sense of the latest sea level paper re ice sheets and a possible link to the latest flooding

article by Chris Bond;
http://www.yorkshirepost.co.uk/news/features/why-antarctica-meltdown-may-mean-more-floods-for-us-1-5183666

We should remember to contact Chris Bond, he seemed very keen on profiling our project at a later date

Chris gave the LQL, no 90 on our project

Chris delivered the 90th London Quaternary Lecture, hosted by the University of London; ‘The BRITICE-CHRONO project and it’s role in forecasting deglaciation of polar ice sheets’ in Nov 2012. Turned out to be a very nice occasion and with wine thrown in and a great talk on how bloody hard it is to get good records of Eemian ice from Greenland ice core records – by Sune Rasmussen.

Had very useful chat with John Lowe, who had some ideas of targets we could usefully date (more on this later). He was also very supportive of our project.

Only my 2nd visit to Royal Holloway