XPS
X-ray Photoelectron Spectroscopy (XPS)
Last updated by: SDM, July 2016
Introduction
Casa XPS manual on share drive
Training
Contact Hugo Celio (hugo.celio@utexas.edu ) for training. The XPS is on main campus in the FNT Building (behind the boat bridge statue). Hugo's office and the XPS are on the second floor.
Training is similar to the MBE method of gradually decreasing supervision.
The XPS reservation system can be accessed here after you receive training and approval from Hugo.
The Casa XPS software is installed on the desktop computer at Emily's desk. You can obtain a copy from Hugo to install on university owned machines after training.
The Banerjee group has an XPS installed in situ with their tiny MBE. The Kratos XPS is generally recommended due to better accuracy / support from Hugo, but the Banerjee XPS is a backup option in case of samples sensitive to oxidation or time. Contact Anupam Roy. This XPS does not have the capability to visually align to an area on the sample once the sample is loaded, so is not recommended for transferred Bi films.
Reservations
The XPS is part of the Texas Materials Institute (TMI) on Main Campus in the FNT building. As with any instrument that's a part of TMI, you must book your equipment time on their webpage. Go to the following page to book time:
https://tmi.utexas.edu/core-facilities/equipment/kratos-x-ray-photoelectron-spectrometer-axis-ultra-dld/Before you can book time, you must
Complete training with Hugo
Create and NSF profile (link on reservation page)
Have an account number available so TMI knows who to charge for the reservation (get this from Seth)
Reservation windows are open up to 6 days in advance starting at midnight 6 days before hand. You need Hugo to operate the machine for sample loading/unloading and any other questions if something goes wrong, so do not book the machine on the weekends or outside regular working hours, i.e. 9am to 5pm.
Operating Procedure
Setup
Remove sample holder bar from chamber (Hugo)
Attach samples to sample holder bar with Cu or black tape depending on conductivity of sample. Sample size of ~ 1 cm x 1 cm is reasonable (you do this step)
Load sample (Hugo), pump down load chamber for several hours
Move sample bar into main chamber
Use manipulator tab to move z close to 1 and y close to 0. x position depends on your sample.
Populate Positions
Under x-ray tab, turn x-rays on, 13 kV, emission 10 mA (as of August 2015)
Move in x to sample 1 region of interest
Under analyzer tab, set analyzer to spectrum, hybrid, pass energy 80 (snapshot), slot (largest aperture)
Under acquisition, enter region name, turn on acquisition
Turn on charge neutralizer (optional- see if it improves signal. More useful for insulators)
Under manipulator, activate jog mode, optimize signal for sample 1 by moving in z
The CPS should be >1000 for a good signal/noise ratio
Go to table -> insert row to save this position
Move to second sample, repeat optimization of signal
After doing this for all samples, under acquisition, deselect active
Set up Scan
Create shortcut to data folder w/ name/date
Under manipulator, go to save positions to save table of positions
Under manager, go to create, then file -> load run to load a previous similar scan
Set up flow chart. Remember to turn x-rays, etc. off after scans finish, but not between scans
Unloading
Schedule with Hugo prior to leaving- make it clear whether you need your samples back.
Choose resume, submit to start scan
General XPS Strategy
Usually do a survey scan first- wide range scan, shows composition of first ~ 10 nm of sample. Good for identifying contamination or unknown phases, and for getting a qualitative idea of how strong each material peak is. This will help you determine the parameters for your single material scans
Set up single element scans. Always scan for the "adventitious carbon" peak at ~ 284.8 eV - the element positions may shift between scans of multiple samples. The adventitious carbon peak is used to realign them (What is adventitious Carbon?)
The XPS book in the lab helps you identify issues like various bonding / oxidation states of the same material.
XPS can tell you ratios with decent accuracy - i.e. this material is Bi2O3 vs BiO. It cannot give you an accurate absolute measure of the amount / thickness / etc. of the material.
Contact Hugo for peak fitting advice. Typically we use a Shirley background fit (XPS Background selection )
When using the software, make sure you use the relative sensitivity factors specific to the Kratos XPS, not for Casa XPS generally ( ). Choosing the RSFs in order to get the peak ratio that will confirm your theory is not science.
Angle-resolved XPS (a technique to determine layer thickness), only works if:
You know the absorption coefficient of all layers of interest
There's a significant contrast between them
The layers are continuous, with good interfaces
Using CasaXPS
Aligning to adventitious Carbon peak
If not already, make sure the "edit mode" is activated
Double Click C 1s for a particular time step or measurement day
Click on the graph and hit F8
Go to the "Calibration" tab
Click on the graph and draw a narrow box near the largest carbon peak, so that the left hand corner of the box is aligned with the main C1s peak position. This should fill in the “Measured” box with the measured peak
Change the "True" to the value you want, e.g 284.8 or 285, etc.
Check the "Regions" box
Click the remaining region scans from you measurement, e.g Bi 4f and O 1s, (hold down ctrl, and select the other boxes)
Select "Apply to Selection" and all the species region charts should now be shifted/aligned to the specified carbon peak value
Check that all the carbon peaks are aligned
Select/highlight the entire carbon data block row
Overlay (F2)
If one of the curves is a little off, it will make any other curves in that dataset/timestep off by the same amount, so fix it!
Curve fitting
Click on the graph
Click F7 (or Quantify)
In the "Regions" tab
"Create"
RSF = sensitivity factor. These values are machine specific, so ask the machine operator for these values Some known RSFs you should use for the TMI XPS are:
C 1s --> 0.278
Bi 4f --> 9.14
O 1s --> 0.78
Ask Hugo about other species' RSFs
Start and End = range of the data over which the software will attempt to fit a curve
You can adjust the area with your mouse. It will show up as either grey bars or a Blue area depending on the version of CasaXPS
Can also manually enter values
BG type = fit to use for the background/noise
Linear
Shirley --> pick this one, acts as ~step function when it encounters a peak
In the "Report spec tab" and change select regions
Click on the active window
Copy (ctrl + c)
Copy all to clipboard
Copy to clipboard and open in MS Excel
In "Components" tab
"Create"
Again, you must adjust the RSF of each component/column created
C 1s --> 0.278
Bi 4f --> 9.14
O 1s --> 0.78
Defaults by using a Gaussian multiplied by a 30% Lorentzian, i.e. "GL30"
Change this by typing in a different value, e.g. "GL65"
Adjust the width of the curve by changing the FWHM
Narrower/sharper --> lower FWHM, e.g 0.6
Wider/broader --> higher FWHM, e.g. 1.1
Fit components on the curve
Might need to fit multiple components
Can use the residual fit button
Click and drag peaks to fit the data
Try to have column A through whatever go from left to right so it is easier for you to know which fitting
Use the "Residual On <-> Off" button to see how close your curve fits the measured data
Back to "Reports Spec" tab
Regions and comps
After setting up the fits for the current data set, e.g. all the carbon curves from multiple timesteps/measurements, highlight all by holding down ctrl and click "Regions and Comps"
Copy all -- use for the other peaks in the other data sets
Relative Sensitivity Factors specific to Kratos XPS (from Hugo)
C 1s 0.278
N 1s 0.477
O 1s 0.78
S 2p 0.668
Si 2p 0.328
Ti 2p 2.001
Co 2p 3.59
Sr 3d 1.843
Pd 3d 5.356
Ag 3d 5.987
Mn 2p 2.659
Cu 2p 5.321
Al 2p 0.193
Sb 3d 8.627
Li 1s 0.025
Bi 4f 9.14
Relative Sensitivity Factors for Banerjee XPS (default library)
Bi 4f 20.7163
O 1s 2.93
C 1s 1.0
Fitting Half Doublets
The Kratos and Vision library sensitivity factors are defined as applying to the entire whole of a spin orbit doublet. This is important when considering p, d, and f levels. If the entire doublet is not included in the quantification region, the sensitivity factor must be adjusted according to the following list.
For example, half of the Sb 3d doublet can potentially lie under the O1s peak in a sample with surface oxide. It's recommended to fit the visible half of the Sb 3d doublet with the adjusted RSF rather than trying to attribute some fraction of the O1s peak to Sb.
SF = normal sensitivity factor for whole doublet L: transition level E: element
ELs - SF
ELp - SF
ELp3/2 - SF*(2/3)
ELp1/2 - SF*(1/3)
ELd - SF
ELd5/2 - SF*(3/5)
ELd3/2 - SF*(2/5)
ELf - SF
ELf7/2 - SF*(4/7)
ELf5/2 - SF*(3/7)
Note - this is a first order approximation. If accuracy is very important, you are encouraged to develop your own sensitivity factors using samples with known surface compositions. It is also a good idea to confirm compositions using a second method if possible.
Notes about data analysis
The following calculation is performed by CasaXPS when determining the percent composition of some element "i", the area under the curve for i, and the RSF for that i.
\begin{composition}
\end{composition}
%composition_i = (100%) x ((area_i}/(RSF_i))/(sum over all i for area_i/RSF_i)The %composition for the same species, e.g. two curves of metallic Bi, should be added together to get the actual %composition. These can be compared against other %composition values for other species, like the adding together two Bi oxide peaks.
Trouble shooting
For machine issues, call Hugo first