Sample Storage and Handling

It is important to advise submitters and students how to better collect and pack samples, so that possible field and laboratory contamination can be avoided (Gillespie and Hedges Reference Gillespie and Hedges1984; Gupta and Polach Reference Gupta and Polach1985; Hedges Reference Hedges1992). Most facilities offer a checklist to assist submitters, which normally includes the following recommendations: (a) keep samples stored dry or frozen; (b) pack and clearly label each sample individually (preferably using Al foil and zipper bags); (c) if using vials, secure the lids; and most importantly, (d) never handle samples in or in close proximity to labs where tracer ¹⁴C might have been used. We go further and suggest that newer facilities access the history of the samples before accepting them for processing. Contamination by ¹⁴C tracer is a very serious problem, possibly leading to facility downtimes (Zhou et al. Reference Zhou, Wu, Lange, Lu, Cheng, Xiong, Cruz, Liu, Fu and Zhao2012), or even equipment losses (Jull et al. Reference Jull, Donahue and Toolin1990). Useful tips to avoid or recover from ¹⁴C tracer incidents can be found in Buchholz et al. (Reference Buchholz, Freeman, Haack and Vogel2000) and Zermeño et al. (Reference Zermeño, Kurdyla, Buchholz, Heller, Kashgarian and Frantz2004). However, the following might also apply:

• ∙ Avoid direct contact of samples, tools, glassware, chemicals, and etc. with the bare top surface of benches and hood areas. Place them in plastic trays (heavy duty and clear polypropylene trays), and cover the heavily used areas with heavy-duty aluminum foil. If contamination or suspicion of contamination occurs, both trays and foil can be easily replaced at a minimum cost.

• ∙ Separate tools and chemicals to be used in each lab or area, and avoid sharing them.

• ∙ Borrowed or shared instruments should be used wisely by ascertaining their history and users frequently.

• ∙ Dispose of used gloves when moving from one procedure to another, or between lab spaces.

• ∙ Clean laboratory spaces yourself as much as possible, since hired cleaners working in multiple labs or buildings can carry contamination from one space to another through the cleaning tools (baskets, brooms, mops, etc.).

• ∙ To conduct a swiping procedure choose small quartz (QZ) filters, such as 25 mm×0.4 μm, from SKC-West (Item #: 225-1825; skcwest.com), as after prebaking the filters at 900°C, they become brittle and larger filters tend to be harder to handle.

Pretreatments of Carbonaceous Materials

To produce graphite samples from carbonaceous raw materials, one should try to remove any contamination that is not associated with the sample itself. Besides being inherently impossible (Santos et al. Reference Santos, Southon, Drenzek, Ziolkowski, Druffel, Xu, Zhang, Trumbore, Eglinton and Hughen2010), this can be extremely complex and highly variable, as identifying and then extracting sample-specific carbon from unknown contaminants can be methodological or sample dependent (Turney et al. Reference Turney, Bird, Fifield, Kershaw, Cresswell, Santos, di Tada, Hausladen and Youping2001), especially when dealing with degraded raw materials (Bird et al. Reference Bird, Ayliffe, Fifield, Turney, Cresswell, Barrows and David1999a,Reference Bird, Moyo, Veenendaal, Lloyd and Frostb; Boaretto Reference Boaretto2009; Ascough et al. Reference Ascough, Bird, Francis and Lebl2011a,Reference Ascough, Bird, Francis, Thornton, Midwood, Scott and Apperleyb; Bird and Ascough Reference Bird and Ascough2012), or mixed carbon pools (e.g. Trumbore and Zheng Reference Trumbore and Zheng1996; Plante et al. Reference Plante, Beaupré, Roberts and Baisden2013). Therefore, some types of samples may require specific chemical and instrumental extractions, which will not be detailed here (for further information, check Taylor and Bar-Yosef Reference Taylor and Bar-Yosef2014 and Trumbore et al. Reference Trumbore, Xu, Santos, Czimczik, Beaupré, Pack, Hopkins, Stills, Lupascu and Ziolkowski2016).

Standard sample pretreatments include physical and/or chemical steps to remove any unwanted material before conversion to CO₂ (by combustion or acid hydrolysis), and have been extensively covered in the literature (especially in the proceedings of the International Radiocarbon and Accelerator Mass Spectrometry Conferences). However, recommendations are given as to the following steps.

Organic Fractions

Carbon buried in soils and sediments may have several fractions (total organic carbon, soil organic matter, humin or refractory carbon, humic and fulvic acids) and/or biomarkers of different ages. One must choose the carbon fraction of interest to be dated, and then select the appropriate physical and chemical procedure to isolate it. Many examples can be found in the literature (e.g. Ingalls and Pearson Reference Ingalls and Pearson2005; Rosenheim et al. 2008; Shah et al. Reference Shah, Mollenhauer, Ohkouchi, Eglinton and Pearson2008; Boaretto Reference Boaretto2009; Druffel et al. Reference Druffel, Zhang, Xu, Ziolkowski, Southon, Santos and Trumbore2010; Bird and Ascough Reference Bird and Ascough2012; Birkholz et al. Reference Birkholz, Smittenberg, Hajdas, Wacker and Bernasconi2013; Martin et al. Reference Martin, Ingalls, Richey, Keil, Santos, Truxal, Alin and Druffel2013; Taylor and Bar-Yosef Reference Taylor and Bar-Yosef2014; Trumbore et al. Reference Trumbore, Xu, Santos, Czimczik, Beaupré, Pack, Hopkins, Stills, Lupascu and Ziolkowski2016 and references therein).

• ∙ Most organic samples (e.g. charcoal, wood, macrofossils, etc.) can undergo an acid/base/acid (ABA) treatment, including some protein samples (feathers, animal and human hair) if the alkaline washes are diluted and performed at room temperature. However, the typical procedure involves a series of warm acid and alkaline washes, and rinsing with deionized water. Santos and Ormsby (Reference Santos and Ormsby2013) made a compilation of several ABA treatments to evaluate their weaknesses and strengths, and found that most of them are suitable even when ¹⁴C ages are close to the ¹⁴C limit (>50 ka BP).

• ∙ Organic samples that are difficult to clean may require a more aggressive treatment, such as wet oxidation (also termed ABOX, e.g. acid, base, oxidation treatment) with or without the stepped-combustion procedure (Bird et al. Reference Bird, Ayliffe, Fifield, Turney, Cresswell, Barrows and David1999a; Santos et al. Reference Santos, Bird, Pillans, Fifield, Alloway, Chappell and Hausladen2001; Brock and Higham Reference Brock and Higham2009; Haesaerts et al. Reference Haesaerts, Damblon, Nigst and Hublin2013), stepped-combustion alone (McGeehin et al. Reference McGeehin, Burr, Hodgins, Bennett, Robbins, Morehead and Markewich2004), or hydropyrolysis (Ascough et al. Reference Ascough, Bird, Meredith, Wood, Snape, Brock, Higham, Large and Apperley2010). Regardless of the chemical strategy adopted, note that those harsh digestions may require several milligrams of organic material (more than usual); therefore, the risks of the procedure should be evaluated carefully (Higham et al. Reference Higham, Brock, Peresani, Broglio, Wood and Douka2009; Santos and Ormsby Reference Santos and Ormsby2013).

• ∙ ABA pretreatments (as well as holocellulose extractions, Southon and Magana 2010) can be carried out in large batches when using 13×100mm disposable borosilicate glass tubes coupled with heat blocks up to ~200°C maximum. Samples needing heating can be securely placed in these heat blocks on bench tops, with culture tubes covered by ventilated plastic caps (Kim-Kap disposable closures, natural, 13 mm for plain-end glass culture tubes). To expedite the procedure and reduce centrifuge usage between chemical supernatant removals, one can try disposable fine-tip pipettes (Thermo Scientific; Samco; 5.0 mL Extended Fine Tip Transfer Pipets).

• ∙ After the last acidification, rinse the sample well to get rid of most of the salt to prevent it from etching the combustion tube. Use double-tube combustion (e.g. an open 6-mm inside of a 9-mm-OD QZ tube) if salt cannot be rinsed off, such as dissolved organic carbon (DOC) samples from freeze-dry down or any other materials that contain salts and tend to erode the QZ tube during combustion.

• ∙ Bone, dentine, and enamel require special extraction procedures prior to combustion as well as quality control assessment by the C:N atomic ratio, and the % carbon and % nitrogen using an elemental analyzer coupled with an isotope ratio mass spectrometer (EA-IRMS) have been extensively discussed elsewhere (DeNiro Reference DeNiro1985; Stafford et al. Reference Stafford, Jull, Brendel, Duhamel and Donahue1987; Ajie et al. Reference Ajie, Kaplan, Slota and Taylor1990, Reference Ajie, Kaplan, Hauschka, Kirner, Slota and Taylor1992; Stafford et al. Reference Stafford, Hare, Currie, Jull and Donahue1990; Hedges and van Klinken Reference Hedges and van Klinken1992; Brown et al. Reference Brown, Nelson, Vogel and Southon1988; Burky et al. Reference Burky, Kirner, Taylor, Hare and Southon1998; van Klinken Reference van Klinken1999; Taylor et al. Reference Taylor, Smith and Southon2001; George et al. Reference George, Southon and Taylor2005; Brock et al. Reference Brock, Bronk Ramsey and Higham2007, Reference Brock, Geoghegan, Thomas, Jurkschat and Higham2013; Boaretto Reference Boaretto2009). Nevertheless, two simple initial tests can indicate the state of preservation of bone and teeth samples:

1. (a) Does the sample feel hard under contact with implements and tools? The strength of a bone is derived from the presence of the collagen that binds the crystalline component together. Soft samples may give low gelatin extraction yields and highly fractionated results.

2. (b) Does the sample smell burnt when ground by an instrument (such as a Dremel coupled with diamond points)? Good-quality bone or tooth material, when in contact with a high-speed burr or cutter, often smells, indicating the presence of proteinaceous material (Santos et al. Reference Santos, Fabra, Laguens, Demarchi, Southon, Taylor and George2006; Beaumont et al. Reference Beaumont, Beverly, Southon and Taylor2010).

• ∙ It would be helpful to know whether a bone sample would yield sufficient collagen for any kind of isotopic analysis and warrant additional sample preparation such as ultrafiltration by Millipore Centriprep filters (Brown et al. Reference Brown, Nelson, Vogel and Southon1988; Beaumont et al. Reference Beaumont, Beverly, Southon and Taylor2010). After a bone sample has undergone demineralization and hydrolyzation, it is possible to ascertain whether a sample may contain enough collagen to require further preparation. Introducing air using a clean pipette into the gelatin solution should create bubbles of gelatin. If the bubbles do not readily pop, it signifies long strands of collagen present; if bubbles pop immediately or none form, this would signify broken or shorter chains of collagen that may or may not survive ultrafiltration (Hector M de La Torre, University of Alberta, Edmonton, Canada, personal communication).

Removal of Secondary Carbonates

• ∙ X-ray diffraction (XRD) and scanning electron microscope (SEM) imaging may help to identify secondary calcite precipitates, and therefore determine sample quality prior to dating (Grothe et al. Reference Grothe, Cobb, Bush, Cheng, Santos, Southon, Edwards, Deocampo and Sayani2016 and references therein).

• ∙ To remove secondary carbonates (Chappell and Polach Reference Chappell and Polach1972) from shells, corals, speleothem, and foram samples, a leach in diluted HCl may be applied. The strength of the leaching procedure can vary (e.g. 10%, 30%, or 50%) depending on sample appearance under the microscope and/or the amount of material available (Santos et al. Reference Santos, Southon, Druffel-Rodriguez, Griffin and Mazon2004). These acid-etching treatments are based on the assumption that secondary carbonates are preferentially soluble than the original calcite. For an alternative method, check Douka et al. (2010).

• ∙ We use 3-mL glass silica-coated Vacutainer blood collection vials (Becton-Dickinson) for direct leaching and hydrolysis of carbonates. The main advantages of using these vials are their reduced costs and less acid waste, as used vials filled with a small amount of acid can be disposed directly into a waste basket (Santos et al. Reference Santos, Southon, Druffel-Rodriguez, Griffin and Mazon2004). The drawback is that this brand is no longer manufactured by Becton-Dickinson. An alternative product is the 3-mL non-silicone-coated glass tube from Monoject (Kendall). However, one must be aware that the red rubber septum is a bit different from the made by Becton-Dickinson; therefore, from our experience, the operator must be extra careful when puncturing the septum to avoid leakage.

• ∙ If you are planning on letting acidification hydrolysis to linger for several hours when using blood vial types such as Vacutainers (Yokoyama et al. Reference Yokoyama, Miyairi, Matsuzaki and Tsunomori2007), consider buying the uncoated glass vials, as it is unclear whether the graphitization reaction could be jeopardized by the breaking down of the rubber or silica coating under the hot acid vapor. Further tests on this subject are needed.

• ∙ Acid fumigation method for getting rid of carbonates from soils or sediments may not work for ¹⁴C analysis when the carbonate content is high, especially with dolomite [CaMg(CO₃)₂] in your sample. Even though the method may work for δ¹³C, ¹⁴C is much more sensitive to dead C contamination (Elisabet Nadeu, Campus Universitario de Espinardo, Murcia, Spain, personal communication). Also, if the acid fumigation method is to be employed, the fumigation time should not exceed 24 hr to minimize blank carbon incorporation (Komada et al. Reference Komada, Anderson and Dorfmeier2008).

Glassware and Reagents

Preparation, baking, and storage of glassware, glass tubes, and reagents are part of the preparation steps necessary during sample processing.

A) Cutting and making tubes for combustion of organics is relatively simple. The tubes can be produced from high-temperature fused silica (such as QZ or Vycor) tubing.

• ∙ Vycor by Corning is a borosilicate type of glass with 96% of silica and a temperature limit of 1200°C. It is generally cheaper than pure quartz. However, early evaluations in its quality/background indicate that they may introduce a higher blank (Vogel et al. Reference Vogel, Nelson and Southon1987). We have not noticed this phenomenon in recent years.

• ∙ To evenly score glass tubes, consider using glass tubing cutters (such as Griffin-Type Cutters Z150770-1EA from Sigma-Aldrich) rather than four-edged glass scoring knifes. Glass tubing cutters cut tubing up to 1-1/2 in. (38 mm) in diameter.

• ∙ It is also easier and more efficient to make two tubes at once by flaming off a tube in the middle of the desired length, constantly rotating the tube until a thick wall at the tip is formed. The opening should be slightly polished for better performance, and to avoid the sharp edges cutting the O-rings in the vacuum line attachments.

• ∙ Glass or QZ tube “wormholes” (bubbles in glass that are formed when the glass is overheated) can be avoided by adjusting the flame to a slightly lower temperature. Sometimes wormholes can be visible under a microscope, or felt when blowing air into the tube. Avoid using these defective tubes to prevent sample loss.

B) Some laboratories apply specific cleaning procedures for their glassware using soap, lab-grade detergent, or even diluted acid rinses. In fact, most glassware can be sterilized by just subjecting it to baking close to the melting point and letting it cool down slowly, followed by storage in clean areas (such as wet cabinets). The combination of cleaning with detergent and baking are not recommended. Alkalis (used in soap-making) at elevated temperatures will etch glass (refer to glassware vendors for details), and may unintentionally affect the background, when residues are allowed to bond with the glass tubes.

• ∙ Borosilicate glasses (such as Pyrex, Duran, Kimex, or Simax) can easily stand baking at 500–550°C, whereas fused quartz and Vycor can be baked at >950°C.

• ∙ The term “wet cabinet” refers to a desiccator acrylic cabinet from Fisher brand with vented shelves. A neoprene gasket and door latches keep it well sealed. By adding a small plastic tray with 10–20 mL 1N NaOH solution on the bottom of the cabinet, one can trap the atmospheric CO₂ inside the cabinet, while helping to reduce the static from the baked glassware. The alkaline solution should be replaced frequently (every 15 days or so). These air-tight cabinets with NaOH trays (solution is poured directly on the polypropylene tray) are also very useful for storing graphite until measurement can be performed. However, be aware that filament graphite does change slowly with time (see details in Beverly et al. Reference Beverly, Beaumont, Tauz, Ormsby, von Reden, Santos and Southon2010).

• ∙ If one can afford to purchase larger Pyrex and/or quartz dressing jars (with loose lids made of metal or quartz), smaller glassware items can be baked directly inside them. Remove jars still warm from the furnace and store immediately into the wet cabinets. This procedure is effective and keeps the glassware sterilized longer, and is far less time consuming. Thus, it simplifies the overall transferring, baking, and restoring of smaller prebaked glassware items.

• ∙ QZ tubes for combustion can also be stored inside a wet cabinet without losing their sterility if they are tightly wrapped in clean aluminum foil until use, or if baked together with reagents inside them, in a standing position.

C) Dealing with reagents

• ∙ We chose to use rods/wire of cupric oxide (CuO) from Fisher. For surface cleaning, we bake a small quantity of material at 900°C for 3 hr in a small ceramic crucible, and transfer it to a prebaked glass vial while it is warm for storage. We repeat the procedure at least every 15 days. However, before baking make sure to remove the very fine powder with a metal sieve #200. By removing the fine powder, less CuO becomes attached to the inner wall of the QZ tube during loading, and therefore does not interfere with the torch sealing procedure.

• ∙ To remove sulfur and/or other undesirable gases after combustion that may poison graphitization (such as Cl), a small amount of prebaked silver (Ag, as powder, sheets, or wire) can be added into the QZ tubes. We prefer to use annealed Ag wire (1.0 mm×0.04 in. diameter, 99.9% metal basis from Alfa Aesar), which is easy to cut into small pieces, and prebake in its own storage glass vial at 500°C until it is ready to be used. If desired, those can be baked together with CuO directly into the QZ tubes at 900°C before combustion of organics. Silver cups (used in EA-IRMS) may also be used during combustion of those “problematic” organics. Although we have not attempted the procedure ourselves, colleagues at the Australia National University reported good success. We suggest to preferably prebake the cups at 300–400°C before using them, and to test the procedure using a ¹⁴C-free blank material to account for any possible exogenous C added by the addition of the silver cups.

• ∙ For the sealed-tube Zn reduction method (Xu et al. Reference Xu, Trumbore, Zheng, Southon, McDuffee, Luttgen and Liu2007), prebaking the assembled reaction tubes with all reagents (Zn, TiH₂ in the bottom of a 9-mm-OD Pyrex tube, and catalyst Fe inside a 6-mm-OD culture tube sitting above Zn/TiH₂) at 300°C in air has shown to be an efficient way to reduce graphitization background (Khosh et al. Reference Khosh, Xu and Trumbore2010).

Combustion of Organics

• ∙ As an extra effort to minimize blanks, we normally combust samples within 6-mm-OD QZ tubes. Larger surface areas, even if they are prebaked, are still prone to pick up more exogenous carbon.

• ∙ To produce enough CO₂ for 1 mg C of graphite, weigh ~60 mg of CuO directly into the QZ tube with the help of a spatula. This amount of CuO is stoichiometrically much higher than the amount necessary to convert 1 mg of OC into CO₂ (Santos et al. Reference Santos, Southon, Drenzek, Ziolkowski, Druffel, Xu, Zhang, Trumbore, Eglinton and Hughen2010). Larger amounts of CuO may introduce larger backgrounds, as the prebaking procedure cannot remove all C embedded in CuO rods (Vandeputte et al. Reference Vandeputte, Moens and Dams1998).

• ∙ When dealing with static and loading of very small pretreated samples (such as charcoal), try make use of a funnel made of a weighing paper, so that the sample can be loaded directly into the prebaked QZ tubes. An Al foil wrapped around the quartz tube is also helpful to lower the static, while weighing samples using a balance.

• ∙ To better deal with static issues in general, an antistatic gun (Zerostat 3 Antistatic Gun) may be used. It is available online from several suppliers.

• ∙ To expedite reagent loading into combustion tubes, we use small spoon-shaped medical devices, such as Chalazion curettes. These are metallic spoons sold online by several suppliers in several shapes and sizes. Once one’s consistency in using the spoons is calibrated and verified, the reproducibility of the reagent weight by using the small spoon alone (without actually using the balance) is relatively good.

• ∙ When samples are known to be naturally loaded with sulfur, several pieces of annealed Ag wire may be used so that S is kept as AgS (see case study in Santos et al. Reference Santos, Bird, Pillans, Fifield, Alloway, Chappell and Hausladen2001).

• ∙ When dealing with powdery samples, QZ wool may be used. Once samples and reagents have been loaded, place a small ball (~5–6 mm wide) in the very top of the QZ tube, then proceed with slow evacuation, and sealing with a torch at ~2–3 cm under the QZ wool plug. Although QZ wool may be baked >900°C before use, its large surface area and strong absorptive properties cannot prevent CO₂ and volatile organic carbons (VOCs) from air to be trapped back into it as soon as it cools off. Thus, it is not recommended to leave the QZ wool inside the combustion tubes. A cheaper and less brittle material option is glass wool. However, the glass wool sterilization must be done at <550°C to avoid melting it.

• ∙ When weighing fine particulate and powdery material (such as sediments or wood powder), burn the neck of the QZ tube just after the sample has been loaded and before attaching it to the vacuum line for evacuation. This simple procedure can combust off the organic matter left from weighing samples, preventing SiC to be formed and easing the sealing of the QZ tube under vacuum (Southon and Magana Reference Southon and Magana2010).

• ∙ Once samples are loaded, and combustion tubes are evacuated and sealed, lay down the QZ tubes horizontally and spread out the inside material to lower the risk of sample and reagents eroding the tube from inside, or exploding inside a muffle furnace.

• ∙ To isolate the combustion tubes in a muffle furnace and prevent a chain-reaction disaster (if one tube avertedly explodes), ~25.5-mm-OD (~1-in.) Sillimanite mullite tubes with open ends can be used. They are suitable for operating in temperatures up to 1700°C. Alumina ceramic tubes are a cheaper alternative, and can be found at the required length from some online vendors.

• ∙ Some labs prefer to identify sample tubes by their particular position within the muffle furnace after combustion (e.g. Marzaioli et al. Reference Marzaioli, Borriello, Passariello, Lubritto, Cesare, D’Onofrio and Terrasi2008). On the other hand, combusted samples can also be identified directly by their IDs written onto the QZ tubes by using a high-temperature marker that can withstand up to ~1100°C. Once QZ tubes are combusted at 850–900°C, the marker ink tends to fade to a lighter shade, but is still visible to the naked eye. Markers are available from different manufacturers; however, we have had good experience with the Dykem High Temperature Tip Marker (sold online by several vendors).

Hydrolysis of Carbonates

For carbonate samples, CO₂ can be produced by in-vacuum acid hydrolysis. To ensure that evacuation of the carbonate vessel occurs before acid reaction, a two-arm flask may be used. However, we choose to perform the in-vacuum acid hydrolysis directly in a 3-mL blood collection vial by piercing the rubber septum with a needle syringe (Hamilton Series 1000 5-mL Luer-Lock syringe #1005):

• ∙ Put a small drop of water around the needle after punching for a better seal during evacuation of a vial, and wipe off the water before pulling the vial off the needle that is attached to a vacuum line to prevent water getting into the line through the needle.

• ∙ For injecting ~0.8 mL of 85% phosphoric acid through the septum of an evacuated vial, put a drop of phosphoric acid on top of the rubber septum first to prevent air leaking in.

• ∙ Heat may enhance the acid hydrolysis reaction, expediting the CO₂ production in the vial headspace. However, do not exceed the reaction time duration unnecessarily, as the heated acid vapor may break down the rubber or inner coating of vials (see also the Removal of Secondary Carbonates section for further details).

• ∙ Prior to CO₂ extraction, a needle attached to the graphitization vacuum line is partially inserted into the septum of the tube, so that the dead volume between the loading portion of the line and needle tip can be evacuated (Santos et al. Reference Santos, Southon, Druffel-Rodriguez, Griffin and Mazon2004).

• ∙ Quantitative extraction of DIC from natural waters by using CO₂-free gas bubbling through the water for ¹⁴C analysis by AMS has long been used by many laboratories (McNichol et al. Reference McNichol, Gagnon, Jones and Osborne1992; Kwong et al. Reference Kwong, Povinec and Jull2004). Recently, a rapid method utilizing a headspace-extraction approach has been developed by Gao et al. (Reference Gao, Xu, Zhou, Pack, Griffin, Santos, Southon and Liu2014). A syringe is used to remove the CO₂ from the headspace after acidification and the CO₂ is then injected into a vacuum line for purification and graphitization. The main advantage of this method is that water is isolated from the vacuum line, thus speeding up the process significantly. The small fractionation introduced by the headspace extraction can be completely corrected for by using the online AMS δ¹³C.

CO₂ Extraction from Canisters and Molecular Sieve Traps

• ∙ U-shape traps with small glass beads in them can significantly increase the trapping efficiency of CO₂ in liquid nitrogen. Also, the use of flow controlling devices is highly recommended for extracting CO₂ from atmospheric samples.

• ∙ Releasing CO₂ sampled by the molecular sieve trap is usually done by heating the trap to 500–600°C. Pumping down the trap first before heating can reduce tailing of CO₂ release. Avoiding water entering during sampling and avoiding overheating during extraction can prolong the life of molecular sieves.

• ∙ Capillary or flow-rate control devices attached to the inlet of an air canister can be used for taking a time-integrated air sample.

• ∙ Canisters, capillary tubes, and flow controlling devices can be purchased from Alltech or Fisher.

Vacuum Lines

• ∙ To protect vacuum lines from undesirable powder, rather than install larger inline strainers that can also restrain the vacuum speed, try inline metal filters. Lay down porous discs (2μm porosity×1/2'' OD frits from Supelco; item# 58255) in strategic places in the vacuum line, to prevent sample particles (when evacuating combustion tubes) or CuO powder (when releasing CO₂ in a graphitization bellows/cracker) from getting into the vacuum line. The strainer fits loosely between the O-ring and the 1/2'' Ultra-Torr metal body. To better hold it in place, try to use it with a fat O-ring. The strainer needs to be cleaned (by blowing it with air) or replaced from time to time. The frequency of cleaning or replacement will depend on the vacuum line usage.

• ∙ If possible, consider building a vacuum line that is relatively simple and short/small in overall volume, to reduce pump-down time duration to reach the vacuum desired. Also, avoid long and relatively “skinny” tubing to avoid limiting the ultimate pressure. Eliminate unnecessary sections and valves to lower the chance of leakage. When in operation, limit the volumes exposed to ambient air pressure by isolating portions of the vacuum line, and thus speeding up the pumping.

• ∙ Beware that certain types of valves have a small dead volume that could trap gases, such as Ball valves from Swageloc. Opening and closing the valve a few times helps to get rid of the unwanted trapped gases.

• ∙ When dealing with extraction of CO₂ in air or other samples with high non-condensable gases, use U-traps with glass bead inserts to slow down pumping speed and increase sample extraction yield.

• ∙ Use an LN₂ trap between an oil mechanical pump and the remaining vacuum line to prevent oil vapor from back-flowing into the line. An LN₂ trap between the turbopump and the line can also speed up pumping, especially for samples that contain lots of water.