A mass spectrometer for high resolution paleoclimatology at the University of Arizona

Jump to:
 History
 Wet lab
 IRMS lab
 Safety
 HTG pyrolysis


History.  With support from NSF's Major Research Instrumentation Program and co-PIs  Warren Beck,  Julie Cole,  Malcolm Hughes,  and Jonathan Overpeck, we have acquired a suite of instruments (scanning micro-XRF, particle size analyzer, ICP-AES, digital densitometer, and continuous flow light stable isotope ratio mass spectrometer) for developing high resolution records of past climates and environmental conditions.  The last of these instruments is my responsibility and main interest.  We can measure oxygen and carbon on solid organics, carbonates, and water samples, with potential for H/D, N and S.  Our specialty, however, is oxygen isotopic analysis of wood cellulose in support of projects in tropical isotope dendroclimatology.  I bought the instrument and peripherals for online preparation of samples in spring 2004; it was delivered in Summer 2004, and was installed and commissioned in spring 2005 into temporary space courtesy of Joaquin Ruiz, Mark Baker, John Chesley, Julie Cole, Dave Dettman, and Jon Patchett of Geosciences.  In summer 2007 we consolidated and moved the lab to Rm. 801 of the Gould-Simpson Building - again courtesy of the Geosciences Dept.  We have specialized in oxygen isotope analysis on celluloses (see below).  Other associated equipment in the lab includes a Sartorius microbalance, a rotary microtome, a digital scanner, and equipment for doing basic chemical extractions (microcentrifuge, vacuum desiccator, oven, pipettes).  Contact me if you are interesting in visiting and using these facilities.









wet lab, Rm. 801Wet lab.  Wet lab side of the room prior to installation of the second fume hood and toxic gas venting.   Equipment for performing sampling, chemical extractions, and preparation for stable isotope analysis includes a digital scanner and networked linux workstation, rotary microtome, three microcentrifuges, micropipetters, a vacuum desiccator, lab oven, and a Sartorius microbalance.  Plenty of working space and storage is provided by a desk and a 4' x 8' table courtesy of UA surplus.


















mass spectrometry lab IRMS Lab.  Mass spectrometry portion of the room.  This is a Thermofinnigan Delta Plus XP with Conflo III continuous flow interface. Peripherals are a Gasbench II water/carbonate preparation system, Costech Elemental Combustion System (ECS), and Costech High Temperature Generator (HTG).  Temperature is kept at about 18-21 degrees C via chilled room air input and exhaust through the hood vent on the roof of the building.  Humidity ranges from 20% to 60% noncondensing.  The hood and gas cabinet vent lines keep the room at negative pressure, which meets Chemical Safety Level-3   (CLS-3) designation.   Power to the analyzer, pumps, water/carbonate preparation peripherals, elemental analyzer and computer is provided by a Toshiba 1600EP series uninterruptible power supply with 230V/30A/50Hz single phase output.  This gives us about 30 min. of  emergency power, as well as clean power to electronics.   Power to the HTG is by an unprotected buck-n-boost transformer providing the HTG with 220V/8A/60Hz from the overhead power line.  Remote access to the IRMS is provided by ethernet; the subnet is protected behind a Cisco firewall router.










gas cylinder farm Safety. Compressed gas cylinder tank farm and safety measures.  Tanks are either chained or supported on a stand.  Because we have carbon monoxide in the lab, there are several necessary special safety precautions for CSL-3 designation.  The 150L CO tank is housed in a toxic gas cylinder cabinet (center) which is exhausted to the roof via the fume hood exhaust line.  We achieved about ~220 linear ft/min through the gas cabinet window.  The CO line to the ConFlo is a single piece of 1/16" stainless steel. tubing.  Prior to exiting the gas cabinet, the CO line runs into a DC electrical solenoid valve provided by Teqcom Industries; the valve is normally closed (open when powered).  It is powered through a  DC circuit on a Nyont Systems Ventalert airflow sensor which is normally open when airflow in the vent line is greater than 70% of normal.  In this configuration, either low airflow or a power outage results in the solenoid closing and CO returning to zero at the ConFlo in about 5 minutes.  Just to be sure we know what's going on, we also have a CO alarm, and we periodically test the flow sensor and monitor calibrations.  I hope we never have an alarm situation.







pyrolysis reactors and induction heaterHTG pyrolysis.  Center: Costech Analytical elemental analyzer and high temperature generator (HTG) used for carbon and oxygen isotopic analysis of organic materials.  The EA is standard issue with options for 50 and 100 position autosampler turrets. The HTG is a radio frequency (1MHz) induction heater that takes a 3 inch region of a pyrolysis reactor to 1400+ degrees C in 35 seconds or so.  As of Fall 2005 we have optimized the system for dilution on mode (samples larger than 300ug) by creating a reactor (left) composed of a quartz outer tube (Costech), boron nitride spacers of my own design (Ferro-Ceramic Grinding), and a molybdenum foil crucible (ESPI Metals). This combination of materials appears to minimize the blank CO signal by minimizing susceptor and carbon mass, as well as heating/exchange with the quartz tube.  The system combines economy, simplicity and throughput with the complete conversion and precision of high temperature pyrolysis: on standard materials, 300ug or larger, precision is <0.3 permil with respect to SMOW).  We have sent the reactors to Ottawa (Paul Middlestead) and Edinburgh (Jason Newton) for further evaluation of the results.  Right: We are currently experimenting with modifying this reactor design for similar precision but with small sample mode (dilution off) measurements on as little as 25-50ug of cellulose.   In this case we have found we need to go to a low oxygen reactor sleeve material made of silicon nitride (right).   A paper describing the system and results is submitted to RCM; preprint is here.  Email me for more information.






dilution on sample chromatogramHere's a dilution on chromatogram; blank signal is typically 10mV or less on all collectors.  Click on the screenshot for a larger version.














dilution off sample chromatogramHere's a dilution off chromatogram; blank signal is typically 100 mV or less on all collectors.  Click on the screenshot for a larger version.













Back.