FILIF Precheck - nthallen/keutsch-hcho GitHub Wiki
Purpose
Perform these instrument checks periodically (such as when the instrument is moved to a new location or field site) to ensure high-quality data.
Leak Testing
Leak testing is the bane of any instrument operator, but it's even more important for FILIF since not only will a leak dilute your sample flow, it will also contaminate the air due to indoor air containing levels of HCHO that are generally higher than outdoor levels.
Generally, spraying Swagelok fittings or any others joints with an acetone bottle will help detect most major leaks (for years, this has been the most common method the group has used for leak detection). If the PTRs are also present in the room, please check with their operator before spraying acetone since their instruments are highly sensitive to the molecule. Under no circumstance should you be spraying water on the instrument since it takes too long to dry (acetone evaporates almost immediately).
At a bare minimum, the leak rate should be less than 5 Torr/min and you should be able to achieve a pressure of 2 Torr or less when the vacuum pump is pulling on the instrument (needle valve and bypass MFC must be completely closed).
Throughput Determination
The throughput through the sample single pass cell should be greater than 90% when measured right before the input mirror into the cell and right after the output mirror at the exit of the single pass cell. A Thorlabs power meter with the UV photodiode detector is perfect for this.
Sample Throughput Test from July 19, 2023
Laser Setpoint @200, Syntax: laser set setpoint 200
NOTE: Upstream, I positioned the PM between the beam splitter and the turning mirror into the cell. Downstream, I removed the beam splitter and placed the PM just after the cell.
In: 2.09 Out: 1.85 ==> 89%
The beam shape/quality of the laser should be checked if the throughput isn't as desired. It should have a bright maximum in the center and then slowly fade as one leaves the center. Using a yellow Post-it note should help with viewing the beam shape.
Minimizing Background Counts
Back in July 2017, Singularity Black, an ultra-black carbon nanotube coating from NanoLab (Waltham, MA), was applied to all inside surfaces of both the sample and reference cells in order to minimize counts due to scattering photons. This coating easily scratches and rubs off, so DO NOT TOUCH THE COATING as the coating is expensive (greater than $5000). If there are minor defects, NanoLab does sell a field kit (with instructions) that can be used to repair the coating. In the event that one side of the detection cell was somehow missing the coating, your background counts would increase by at least 1-2 orders of magnitude even if your alignment was perfect.
CAUTION: DO NOT TOUCH THE CARBON NANOTUBE COATING AS IT'S EASY TO REMOVE AND IS EXPENSIVE!
The background counts can be minimized by using the two mirrors immediately preceding the single-pass cells to align the beam down the center of both cells. I would recommend having a plot in QNX of the raw counts from the single-pass cells to visually see if a slight adjustment in one of the turning screws actually lowers background counts.
Please note, make sure that you're not on an HCHO spectral feature when performing this check and that your power is relatively constant. In theory, if the laser stayed at exactly the same position, then this wouldn't be an issue when minimizing counts. In practice, the laser does tend to jump around a bit in terms of wavelength, enough so that it would be difficult to see if a lowering in counts is due to better alignment rather than the laser simply moving off the HCHO spectral line.
Power Meter Calibration
To obtain the calibration curve for the first power meter (right before the beam enters the single-pass cell), place the Thorlabs power meter right after the first beam splitter. Then, cycle through various powers on the laser so as to obtain a table of laser powers (in mW) and readings from the power meter itself (in Volts; monitor TS1_T in QNX). Obtaining about 10-15 different points over a wide range of laser powers is generally sufficient to obtain a good calibration curve. When performing the calibration, the Thorlabs power meter should not be moved. Before proceeding with any mixing ratio calibrations, input the results of the power meter calibration in TS.tmc following the directions here.
Note: The calibration is only valid for a particular alignment. If the alignment ever changes, you should perform another calibration.
Sample Power Meter Calibration - Operator Commands Highlighted
Recent Power Meter Calibration
Zero and Cal Factor Determination
Before data collection, a zero and cal factor determination should be performed. To do the zero, first pass ultra zero air from Airgas through a column filled with molecular sieves before sending the air into the instrument. It is advised that an overflow is put in place and that you flow 5000 sccm ultra zero air to prevent room air from accidentally leaking in from your overflow.
When flowing zero through FILIF, check onscreen whether the HCHO mixing ratio reported by FILIF is varying around zero. Of course, it's normal for FILIF to report negative mixing ratios due to variability, but the mean mixing ratio of HCHO collected over several minutes should be within 50 pptv of zero (or closer depending on the application). If the mean value is deviating far from zero ppbv, then I would first check to make sure that the power difference between the online and offline positions is minimized as much as possible in order to eliminate this as a possible reason for offline counts being higher than online counts (this is what causes a negative mixing ratio to be reported). If the algo is still reporting negative mixing ratios, then there's likely some additional laser scatter at the offline position (compared to the online position) that's adding just enough counts to force the mixing ratio negative. To fix this in the past, I would just adjust the mirror immediately prior to the entrance of the sample cell in order to fix the problem. If you touch or bump the mirrors prior to the power meter, you'd have to redo the power meter calibration from the previous step, which is annoying.
For the actual cal factor determination, use the same ultra zero air that you would use for a zero to dilute the 100 sccm flow from the HCHO (600 ppbv HCHO balance N2) gas cylinder. These cylinders can be special ordered from Airgas through Harvard's Airgas contact. Please note that it takes six weeks for the cylinder to be delivered, and it costs nearly $1000-$1500, so make sure to always use small flows and to close the cylinder immediately after use to prevent costly leaks.
Important: When doing a cal factor determination, flow gas from the HCHO cylinder for several hours before actually performing the calibration since it takes at least three hours for the surfaces to be passivated (such as in the gas regulator and span MFC). Please note that the span MFC is coated in Fluoropel to help prevent loss of HCHO to the surface, but this equilibration time is still required.