Tutorial_and_Recs

Jose's 2018 Int. Aerosol Conf. Tutorial on Oxidation Flow Reactors can be found at this link and below. Any input is welcome. 

In the tutorial many recommendations about how to run OFR experiments and report data were proposed, which are reproduced below. Any input is welcome, and I will update them in the future with the input I receive. It is understood that not all recommendations will be relevant to all experiments.

1. General

• • Use “OFR” in prez + pubs, mention specific reactor in methods

  • Subscribe to PAM Users list: https://groups.google.com/d/forum/pamusers 

  • Submit your PAM and other OFR papers to Andy lambe@aerodyne.com or Jose jose.jimenez@colorado.edu 

2. Physical Design and Operation

• • Always monitor reactor stability after steps and before measurements

  • Take the peak in SOA production as the organic PAM

  • How to run the lights

    • Cycle the lights as fast as possible

    • Keep in mind that keeping UV constant does not get you constant OHexp!

    • Cycle no faster than ~x3 the residence time in reactor, or there can be carryover

      • In our ambient studies, ~5 steps spaced in log(OHexp) over 2 hrs (true OHexp will fluctuate in time)

    • ALWAYS include a step with lights off

    • If sampling fast-changing sources (fire, car), try to run at “constant” age of 1-3 days (after accounting for suppression)

    • ALWAYS do zeros daily (or more) 

      • Humid zero air for contamination

      • Clean w/ methanol & run reactor for ~1 day at the start

      • Filter in front & pressure test for leaks (plumbing gets very complicated)

        • Careful with plastic filters, they can generate a lot of SOA

  • 2D variation of UV field

    • If optical depth > 1, use 4 lamps in PAM

    • Evaluate the issue for other OFRs

  • Residence Time Distribution

    • Try to approach plug flow or laminar flow (less wall contact, more uniform reaction time)

    • Download RTD figure from PAMWiki and plot your RTD on top

    • Pay attention to inlet & outlet details, and quantify by measuring RTD

  • Reactor Material

    • Use glass reactors if particles not charged (e.g. from nucleation) OR remove charged p. OR if don’t care about quantification of amount of particles

    • Especially avoid glass for combustion sources, that make highly charged particles

    • Use reactors with a conductive surface if particle losses are important for your experiments (but may conflict with vapor loss recommendations!)

    • Pay attention to the charging of particles by UV lights if working with combustion or metal particles

  • Vapor delays in tubing

    • Use Pagonis et al. (2017) results for your systems (taken as an approx. lower limit delay for non-Teflon tubing

    • Run w/o an inlet if you can, or otherwise as short as possible with high flow

    • From sources, dilute w/ laminar sheath flow to avoid wall contact (see later about need for dilution)

    • Heating can help a little but does not solve problem

    • There is a need for creative solutions for each situation

    • Evaluate problem experimentally if you can

  • PAM walls stickiness

    • Try to run at RH > 50% to reduce stickiness (also good for chemistry)

    • Be aware of limited memory effects (in particular NH3 – Palm’16)

    • Systematically investigate coatings (conductive Teflon?)

  • PAM temperature

    • Always operate OFR at ambient outside temperature, not lab T

    • Always cool your lights with a sheath flow

3. Quantification of Aerosol Formation and SOA

• • Use acc. coeff. ~1, supported by PAM and chamber work (Palm et al., 2016; Krechmer et al., 2017)

  • Apply the LVOC correction algorithm (posted in PAMWiki)

    • Use a time and OHexp dependent correction, with average of input and output cond. sink

    • Use the cond. sink at ambient humidity

    • Note that wall timescale will depend on your RTD

  • Inject a seed if under low CS conditions (clean sites or experiments)

  • Verify that correction is working

    • E.g. SO2 + OH -->  H2SO4 (Need good OHexp quantification)

4. Quantifying Oxidant Exposure

• • Use global average OH (24-hr) of ~ 1.5 x 106 molec. cm-3 to convert between OHexp and photochemical age

  • Learn by heart what OHexp and OHR are and always quantify them for your experiments

  • Understand the results of OFR kinetic studies

  • Measure OHexp from decay of a species

    • CO and VOCs if possible

    • Beware of tubing delay issues (even at c* ~ 106 ug m-3)

    • Careful with VOC interferences in PTRMS

    • Careful with SO2 at higher RH (line losses)

  • Use estimation equations when available (in PAMWiki)

    • With a multiplicative factor determined by fitting your calibration data (typically x2-3 for Penn State lamps)

    • Estimation equations don’t work for OFR185 with Aerodyne lamps (yet), but they do work for all OFR254

  • If you are doing something unusual, consider running a full model

  • Be aware that it is typically unrealistic to capture OHexp better than ~x2

  • Be aware of the impact of OH supression

    • Need to run OHexp calibration experiments covering the same ranges of UV, humidity, and OHR than your experiments of interest

    • Doing calibrations at low OHR and applying them to experiments with high OHR can lead to errors in OHexp of orders-of-magnitude

  • Use the model output maps for experimental planning

5. Relevance of OFR chemistry to the Atmosphere

• • Be aware that if the OFR chemistry strongly deviates from atmos., very different molecules will form, and the trends that you measure (e.g. mass, light absorption, toxicity, CCN…) can be totally wrong compared to real atmospheric conditions

  • Run at near ambient concentration levels (OHR <~ 100 s-1)

  • Be aware that photolysis of gases has lower importance in OFR than in the atmosphere. If replicating atmospheric photolysis is important to you, consider adding blacklights

  • Photolysis of particle material by 185 and 254 nm can be important at high UV settings (e.g. unrealistic brown carbon "bleaching"). Do control experiments if this is important to your experiments

  • Be aware of the RO2• fate in your experiments

  • Be aware of the multiple additional issues for high-NO experiments

  • Dilute and humidity exhaust from sources as needed (e.g. cars by x100, biomass burning by x1000)

  • Evaluate non-OH processes in the reactor and adapt as needed (e.g. add H2SO4 seed to capture IEPOX-SOA

And email jose.jimenez@colorado.edu with any suggestions for additions, corrections, clarifications etc.