Ready, Set, Sublimate

Hello folks, apologies for being slow to update during the season.  Now that I’m back home and basking in the summer heat, I wanted to take a minute to go over one of the critical objectives that we embarked upon this past season – the Field Sublimation device!

Photo by Ben Hmiel

The system as it was assembled in the field

First, allow me to explain the rationale behind the system.  The overall goal of the project at summit is to characterize the production and retention of cosmogenic in-situ 14C through the firn column.  To put that more simply, we’re trying to find out how much 14C is produced by nuclear reactions from cosmic rays in the upper part of the ice sheet, as well as how much actually comes to be trapped in the air bubbles as opposed to just diffusing out of the firn matrix.  We’ve already discussed the large-volume Ice melting system in another blog post, but one big limitation to highlight of that system is that it can only characterize the 14CO and 14CH4 content in the Ice – not 14CO2.  This is because melting of ice can lead to extra CO2 production from carbonate dust present in the ice (especially significant in Greenland – less so in Antarctica). So with only the Melter, we’re left with an incomplete picture of 14C production as 14CO2 is a major contribution of the overall 14C content.  By sublimating the, ice we skip the liquid phase during the extraction and prevent the aqueous carbonate equilibrium reactions from producing extra CO2.  We’re able to do this by keeping the Ice below the ‘Triple-Point’ on the water phase diagram, enabling the conversion directly from Ice -> vapor just by keeping the Ice cold and under vacuum.

Other laboratories have attempted to measure 14CO2 by dry extraction methods, however since the cosmogenic 14C is produced within the ice matrix itself (not just the air bubbles), results suggest that dry extraction methods fail to liberate all of the in-situ CO2 leading to inconsistent results.

Great, so we need to sublimate the Ice if we want to extract 14CO2.  Then why on earth would we bring our system to the field – isn’t it easier to build a complicated extraction system in a laboratory with stable temperatures and electricity not provided by generators?

Well, the reason for that is two-fold.  First, we want to sample the firn column to determine the amount of in-situ 14C that is retained through to bubble close off.  Firn samples are known to trap bubbles over time following drilling and shipment back home.  This would lead to atmospheric air being trapped (which is high in 14CO2) and mistakenly part of our sample.  Thus for the firn samples we plan to collect – the only option is to extract the 14CO2 in the field immediately after drilling.  Additionally, once an Ice sample is drilled, it is subjected to intense 14C production at the surface of the ice sheet at a high altitude, high-latitude location like Summit (and even more at higher altitude during the airplane shipment back home).  Thus, there would be significant post-coring production of 14CO2 that would need to be accounted for and subtracted out.  That being said – we still did collect some Ice samples in the field for the purpose of testing this effect.  I plan on running them in Rochester during the months to come in order to see how my field results compare; however to get the best data, we aimed to sublimate each sample with hours or a few days of being drilled.

OK, so with the justification and science explained, let’s get into what a typical sampling day for me would entail in the field.  Just after waking up, I would sleepily crawl out of my tent to head into Plan Q to check on the system from the overnight evacuation (while also hoping that my generator did not run out of fuel overnight).  After performing a few checks, I head to breakfast while setting up a test to test for any leaks of ambient air that could contaminate my sample. Technically there is always a leak of ambient air in any vacuum line, however it is a matter of whether or not the amount is significant in terms of the analytes being measured.  During the course of the sublimation, I need to make sure that the leak rate into the vessel and surrounding lines was less than ~1×10-5 sccm (standard cc’s/min) to avoid trapping any significant amount of extraneous CO2.  To think of it another way, that equates to a cube of air 6mm on each side (roughly the size of a corn kernel) entering the vessel every hour – that’s not very much air!

Photo by Ben Hmiel

Here is the vessel empty without insulation and ethanol in the moat. Beneath the plate lies another chamber where the Ice is loaded

Once the system passed all the morning tests, I was ready to load my sample for the day into our large glass sublimation vessel above.  The design of this system allows for an Ice sample to be loaded onto a chilled pedestal from below.  There is then an array of IR lamps that irradiate the sample while it is under vacuum.  Then above the Ice sample, we designed a built-in ‘moat’ of Ethanol chilled with a probe chiller.  The moat serves to re-condense the bulk of the water vapor right away, allowing the gases and 14C to travel through the rest of the vacuum line and minimizing the amount of water vapor co-transferred.  This glass vessel is rather cumbersome and difficult to work around, but amazingly it worked well for the season and only suffered one breakage (luckily we had a spare made).  Thanks to Allen Scientific Glass for making this rather complicated device for us.

Photo by Ben Hmiel

Here is the system in operation with ethanol in the moat cooling the inner chamber.

Loading the Ice into the vessel is a delicate process that takes two people to perform.  By the time I was well rehearsed in the procedure, I was able to minimize the time that the vessel was open (and subjected to ambient CO2) down to ~5 minutes.  Once the Ice was in the vessel and the flange assembled, we begin by evacuating the ambient air out of the vessel.  This is a key part of the procedure that takes ~45 minutes allowing for a few flushes of CO2-free air to make absolutely sure that any ambient CO2 (including that which I breathe into the poorly ventilated building) is not collected during the extraction.  We also begin to sublimate a small amount of Ice off of the surface to ‘pre-clean’ it from any CO2 that may have adsorbed onto it.  CO2 is known to be a very “sticky” gas, meaning it readily adsorbs onto surfaces – making it a pain-in-the-neck for all scientists trying to measure it!

Photo by Xavier Fain

View from below with the lamps irradiating the ice block in the lower chamber

Following the evacuation it’s time to sublimate the Ice!  I set the lamps to ~25% power and monitor them over the course of the day.  The amount of time it takes to sublimate a sample depends on the CO2 concentration, sample mass and geometry.  In general for fully closed Ice, I can sublimate a 1.5Kg sample for ~7 hrs to extract ~20μg of CO2 that will be sent to ANSTO for graphitization and measurement by AMS.  The lamp voltage is adjusted from a pair of Rheostats mounted on the device.  Generator power is fairly inconsistent and unstable, so I would have to monitor and adjust the power based on my readings of the pressure inside of the vessel.  Too little lamp power and I’m not sublimating enough Ice, too much and the vapor pressure increases to an extent that I may end up pushing the system above the triple point and melting it slightly – neither are desirable scenarios.

The rest of the system is rather simple in comparison to the vessel.  Downstream of the vessel sits a pair of glass traps that are chilled to -90 in an Ethanol bath to condense any excess water vapor that is not trapped by the moat in the main vessel.  It is imperative to remove as much water as possible beyond this point, since water vapor behaves similarly to CO2 in our vacuum system and it is possible to mistakenly identify water as CO2 during a pressure reading in the manometer.

Photo by Ben Hmiel

Molecular sieve trap held under liquid nitrogen

Following the water trap is a small loop trap held under liquid nitrogen to trap the CO2 during the extraction.  At liquid nitrogen temperature (-196C), CO2 will freeze under vacuum while allowing the bulk Air (mostly N2 and O2, some noble gases as well) to pass through unaffected.  Other trace species (Eg. N2O) will also freeze at this temperature, however their concentration is generally too small to be significant.  Following the CO2 loop trap is one final glass trap, this time containing ~3g of molecular sieves that are also held under liquid nitrogen during the extraction process.  When cold, the molecular sieves will adsorb the majority of gases they are exposed to, save those with small molecular diameters (Eg. Ne, He).   This allows us to trap the remaining air that is released from the Ice during the extraction process.  This is needed in order for us to measure the concentration of CO2 we collect.  Additionally, the mole Sieves serve as a ‘pump’ – creating an area of low pressure that drives diffusion of the gases out of the sublimation vessel downstream towards the CO2 trap.      While it is not that strong of a ‘pump’ – over the ~7 hour extraction time it is certainly sufficient to get the majority of gases out of the vessel.

Photo by Ben Hmiel

Manometers for measuring CO2 and air content

The only part of the line left to explain are a series of manometers for measuring the pressure of the bulk Air collected on the sieves as well as the pure CO2 that is distilled from the liquid nitrogen trap.  Once all the measurements are made, the CO2 is manipulated into one side of the line where I flameseal it into a sealed glass ampoule.  This part is a bit nerve-wracking, however with plenty of practice it becomes fairly routine.  I’m proud to say that I have never botched a flameseal on a real sample (either here or on the 14CH4 extraction line back home) – though that says nothing about how many failed while I was practicing.  The glass vials will eventually be shipped halfway around the world to Australia to be measured for 14C on the AMS (Accelerator Mass Spectrometer).

I hope that gives everyone an informative view of what I did up in Greenland this past season.  Once we get some results from this system I’ll be sure to share it on the blog!



Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s