High Resolution ToF-AMS Analysis Guide

---

Introduction

The High Resolution ToF-AMS (HR-ToF-AMS) Analysis Guide is dedicated to the analysis of complex HR-ToF-AMS field data sets and is intended to establish standard practices in this area on a community-wide basis. It is assumed that users will have some familiarity with Igor and the community based ToF-AMS (Squirrel) analysis tool, Squirrel. Preliminary HR-ToF-AMS data can be generated while a project is underway without worrying about some details; this guide is intended for identifying all post-project analysis steps involved in generating final data.

A Message for Contributors

We want to encourage active participation by all ToF-AMS users in the evolution of the information contained within this wiki and welcome the addition of content that is beneficial to the community as a whole. However, please DO NOT delete any content from this page!! Significant time, effort, and deliberation has gone into the information contained in this page. Rather than deleting content, please feel free to voice your concerns by posting a comment to the discussion page where others can contribute (please be sure to include a topic to be referenced in responses).

HR-ToF-AMS Process Flowchart

m/z Calibration

• ions have to be isolated at the FWHM level, but not neccesarily at the tail level
• Doug: O+ is often bad, and we don't care about it. We do care about NH2+.

Squirrel Baseline

• Choosing baseline regions
  • Check several m/z at low (m/z 15), mid (m/z 50) and high (as high as you care), and see that the baseline regions are proper for all of these regions
  • Check for runs at start, middle, and end of experiment (or before and after instrument changes, etc.)

Peak Width

• The PW function is used to produce mathematical parameters which, once convolved with the user-determined peak shape (PS, see below), give a model of the expected measured signal as a function of mass for a given ion. The PS is simply a look-up table corresponding to the expected intensity of a peak as a function of the deviation in mass space from the centre, thus the m/z-dependence of the instrument model is contained with the PW function.

• The mathematical functions that can be used in the PIKA model of this m/z dependence are :

1) Line fit, PW = a + b * m/z 2) Power fit, PW = a + b * (m/z)^c

• Gaussian fits to isolated ions (ie those whose measured signal are not expected to contain interference from adjacent ions) are used to determine the FWHM as a function of m/z. Clearly, pre-requisites of an accurate determination of f(m/z) are an accurate mass calibration (defined here in a shamefully un-quantified manner) and selection of appropriate ions and runs for inclusion in the process.

• The process of determining peak width:

1) Calculate the gaussian fits for a range of ions 2) Choose todo 3) Choose ions to use 4) Determine PW as f(m/z) for each run in the todo wave 5) De-select problematic runs from the time-series 6) Calculate the average. This average flattens the time-dependence of your PW function to 2 or 3 parameters (depending on linear/power fit) which are declared to be the "final" peak width parameters for the entire todo wave. In consequent PIKA fitting, this single set of PW parameters are used for the HR fits on every run in the todo wave.

• 1) Gaussian fits
  • In Step 1, the PWPS panel finds gaussian fit parameters for a large range of ions for all the runs in a given todo wave. Only ions and runs fitted in this step can then be used to calculate PW. It is recommended that the user simply fit all the defaults for todo wave "all".

• 2) Choosing the todo for PW calculation
  • It is required to go through the remaining steps in the PW process for each subset of runs (ie, todo wave) for which you expect a different peak width within the instrument, for example:
    • V- or W-mode data, obviously
    • Step-changes in TOF setup (new MCP, filament => totally different tuning?)
  • For each of these subsets of your time-series, it is necessary in PW calculation to use a todo wave that encompasses a collection of runs which representatively reflects the magnitude and variation of PW across the entire subset. This might need to be the entire subset (e.g. "allV")... bear in mind that since the time-consuming step of generating the gaussian fits has already been completed, it is usually advantageous to include all the runs in PW calculation for each subset.

• 3) Choose ions to use
  • This is achieved through altering the binary masks for V- and W-mode displayed in the upper-right table <picture>. The PW calculation could be perturbed in the instance where results are included from:
    • ions where signal from adjacent masses broadens the signal within the gaussian fit-range (defined by "delta"),
    • ions with low signal-to-noise whose gaussian fit parameters may not be particularly well-constrained,
    • ions where the fit routine has failed to find the correct peak owing to imperfect m/z calibration.
  • The default list of ions for the gaussian fits attempts to exclude ions where one might expect interference. However, there are cases where the signal for individual ions may be artificially broadened, in particular in the case of poor instrument tuning (resolution). In the fit of PW vs. m/z this would manifest itself through points clearly deviating from the fit line. If a shoulder or double-peak is visible in the raw mass spectrum at this m/z, this is a clear indication that the ion should not be used in the PW determination.

• • In the case of a poor m/z calibration, the fit routine may miss the intended ion altogether:

<picture>

• • The user should utilise both the single run fits (taking note that the raw spectrum can be popped out) and time-series results to ensure such problematic

banded peak shape

Peak Shape

• ions have to be isolated at the tail level

Choose Ions

• Better to start simple, with fewer ions, and add ions as needed. Having too many ions in the fit can make the fit look good and have low residual for the wrong reasons
• Most people think that the CNO family of ions should NOT be used by default, and only added if really needed per users judgment.

• Web site for exact mass and isotopic calculations
• DeLaeter 2003 IUPAC Atomic Weights paper

• Curators of PIKA Ion List: Puneet Chabra (Chair), Delphine Farmer, Niall Robinson, Qi Chen

Group Ions into Families

• Users can define families for their own purposes
• Each ion can only be a member of one family

PIKA Fits

• Evaluating the fits
  • Look at the left side of m/z 48 and 64, which should not have any real ion to their left
  • Look at the right side of m/z 41 (C3H5+), which should not have any ions to their right

HR Results

• User should inspect the spectra of each family. Ions that greatly stick out are suspicious, and the user should go back to the PIKA fits and see whether they are real. This is particularly important for the CHN and CHNO ions, which tend to have a very small fraction of the UMR signal at most m/z's.

HR Batch & Frag Tables

This is not implemented yet, but we are targeting to release a version with these components at the 2009 AMS Users Meeting in Toronto. These tables will define the RIE and CE and the fragments to be used to calculate mass concentrations of species from HR data. The frag table is needed to e.g. apportion the H2O+ signal between sulfate, organics, air, and particle water, or the CO2+ signal between air and organics, etc. Right now the user has to do that manually.

PIKA Species Loadings

For the time being, the user has to do that by themselves.