Adventitious Carbon

 

What is Adventitious Carbon ?
Can we use it for Charge Referencing BEs from Insulators ???

by B. Vincent Crist (Sr. Editor at The XPS Library)

The chemical composition of “adventitious carbon”, which has been widely used for charge referencing, has never been definitively determined by any analytical method, even though ToF-SIMS and GC–MS could be used to reveal more about this type of carbon. Adventitious carbon is also known as Air-borne Molecular Contamination (AMC) which has carbon compounds and a few inorganic compounds – all of which are Air-borne.

It is very  important to note that the ratio of the different types of adventitious carbon chemical state moieties (hydrocarbon, alcohol, ether, ketone, ester, acid, carbonate) changes in accordance with the basic chemical nature of the substrate (e.g. metal, glass, ceramic, oxide, polymer) and the origins of the adventitious carbon.

Adventitious hydrocarbons are convenient as a rough or crude reference energy, but there are many variables that you must accept or deal with if you use adventitious hydrocarbons as a reference energy.

Adventitious carbon will contain different types of carbon depending on the contaminations that exist in the “production” room, the “packaging” room, the “storage” room, the contamination inside the plastic bag or plastic box used to store the material.

The type and nature of adventitious carbon also depends on your location around the world.  Are you near a forest that releases all sorts of pine oils or other tree products?  Are you located in a major city that has all sorts of air pollution that will eventually attach itself to a freshly made surface which is your product. Are you located near the ocean or a big lake?  These are also sources of airborne organic matter that can attach itself to your surfaces.  Is your company making chemicals?  Some of the gases from chemicals will move around.  You may not smell them, but they might be there in the PPM or PPM level.

It is interesting to note that the ratio of the different types of adventitious carbon chemical state moieties (hydrocarbon, alcohol, ether, ketone, ester, acid, carbonate) changes in accordance with the basic chemical nature of the substrate (e.g. metal, glass, ceramic, oxide, polymer). The as-received fully passivated surface of a metal often has 40–60 at.% of adventitious carbon on it, whereas polished glasses and ceramics have 20–40 at.% of adventitious carbon and polymers have only 1–10 at.%.  The C (1s) spectra shown below reveal the various levels and types of adventitious carbon that collect on different materials.  The Magnesium and Lead native oxides collect CO2 gas from the air and form Carbonates.  The formation can be direct or the native oxide may form a Hydroxide that then adsorbs the CO2 from the air.

Charge referencing is performed by mathematically correcting all experimental peak values to a suitable reference energy, which in the case of non-conductive materials, is normally the C 1 s BE of the covalently bonded, non-ionic, hydrocarbon (component) moieties (C-H, C-C, C=C, CxHy) that exist on the “as-received” surface of nearly all materials. This hydrocarbon moiety (component) is the dominant form of adventitious carbon on all materials if the sample has not been recently ion etched, fractured in vacuum, or specially treated to remove the adventitious carbon.

 

Adventitious Carbon C (1s) Spectra on As-Received Native Oxides, Fresh Exposed Bulk and Semiconductor Wafers – Examples

 

 

 

 

 

Using C 1s to “Charge Reference” BEs from Non-Conductive Materials

 

Be Very Careful

 

A fundamental problem of uncertainty and error exists for non-conductive materials (insulators) which represents by far the most common type of material analyzed by XPS. To use BEs measured from a non-conductive material, the XPS user must first use a method to compensate, offset or correct for the charge induced energy shift that occurs when the X-rays and/or a low voltage beam of electrons strike the surface of the area being analyzed. This method is commonly called “Charge Referencing”.

Charge referencing is performed by mathematically correcting all experimental peak values to a suitable reference energy, which in the case of non-conductive materials, is normally the C 1 s BE of the covalently bonded, non-ionic, hydrocarbon (component) moieties (C-H, C-C, C=C, CxHy) that exist on the “as-received” surface of nearly all materials. This hydrocarbon moiety (component) is the dominant form of adventitious carbon on all materials if the sample has not been recently ion etched, fractured in vacuum, or specially treated to remove the adventitious carbon.

This fundamental problem exists in all numerical databanks, and is due to the 0.4–0.6 eV range in the reference energy of the hydrocarbon moiety C 1s peak BE that is used for charge referencing non-conductive materials. Table 8 shows the range of C 1s reference energies still in use today. The hydrocarbon moiety C 1s reference energy that is used for charge referencing raw spectra obtained from non-conductive materials depends completely on the initial training of the operator, author, or scientist.

Each beginner is trained to believe (accept) that the C 1s reference energy for the hydrocarbon moiety is one of the following values:

284.6,  284,7,  284,8  or  285.0 eV

At this time, there is no standard reference energy for the C 1s BE of the hydrocarbon moiety or any pure non-conductive hydrocarbon material, and there is no known method for ensuring a true Fermi level contact between the surface of a non-conductive material and a suitable reference material or a user deposited layer of some pure non-conductive hydrocarbon material.

The 0.6 eV range of hydrocarbon moiety C 1s reference energies (284.6–285.2 eV) shown in Table 8 is based on a summary published by Swift and the work done between 1970 and 1980 by Malmsten, Schön, Johansson, Richter, and Wagner (referenced in Swift [30,34]), who reported hydrocarbon moiety C 1s BEs found by peak-fitting the C 1s signals attributed to the hydrocarbon moieties that exist on the “as-received” surfaces of several noble metals: Ag, Au, Cu, Pd, Pt (Table 8). The decision to use only noble metals was not defined in the original publication.

This 0.6 eV range in the “hydrocarbon moiety C 1s reference energy” is large enough to cause the mis-assignment of chemical state structures that exist within or on the surface of, not only nonconductive materials, but also carbon bearing materials (polymers) both of which constitute most of the materials analyzed by XPS. The effect of this 0.4–0.6 eV range is shown in the Periodic Table shown as Fig. 3. The reader is referred to the Range, Standard Deviation and Confidence Levels for the non-conductive materials which reveals the potential for mis-assigning chemical states.

The chemical composition of “adventitious carbon”, which is widely used for charge referencing, has never been definitively determined by any analytical method, even though ToF-SIMS and GC–MS could be used to reveal more about this type of carbon

It is interesting to note that the ratio of the different types of adventitious carbon chemical state moieties (hydrocarbon, alcohol, ether, ketone, ester, acid, carbonate) changes in accordance with the basic chemical nature of the substrate (e.g. metal, glass, ceramic, oxide, polymer). The as-received surface of a metal often has 40–60 at.% of adventitious carbon on it, whereas polished glasses and ceramics have 20–40 at.% of adventitious carbon and polymers have only 1–10 at.%. (Unpublished results by the author.)

To characterize the range of C 1s BEs that occur for native oxides, the author measured a series of old, naturally formed, native oxides that behaved conductively in an XPS system equipped with a monochromatic beam of X-rays and a low voltage beam of electrons (flood gun). A table of the hydrocarbon moiety C 1s BEs derived from those samples is provided as Table 9. The charge neutralizer (flood gun) was turned off during all measurements.

The same native oxide samples were removed, scraped clean in air, re-entered to the analysis chamber, strongly ion etched to remove residual surface contaminants, and left in the XPS analysis chamber under the X-ray beam overnight for >14 h allowing the surfaces to develop a new steady state native oxide and a new layer of carbon contamination that originated from the gases that are residual to the UHV analysis chamber and the contamination that existed elsewhere on the surface of the sample. (The instrument used for this study was an SSI X-Probe, equipped with a monochromatic Al K1 source of X-rays and a cryo-pump on the analysis chamber having a base pressure of 10−10 torr.)

The average of the C 1 s BEs measured from those native oxide samples is 285.4 eV for the BE of the hydrocarbon moieties that exist on naturally formed, native oxides (Table 9).

 

 

The average of the C 1 s BEs measured after ion etching the same samples is 284.9 eV for the hydrocarbon moieties that on the surface of ion etched metals after many hours (>14 h) in cryo-pumped UHV (Table 9). Based on these statistics, the 284.9 eV value, which is free from any native oxide effects, is probably a more reliable number to be assigned to the hydrocarbon moiety. The higher 285.4 eV number from native oxides is probably due to unexpected charging effects or surface dipole moments, neither of which have been studied.

 

Fig. 6. Box plot of C 1s BEs from old, naturally formed native oxides showing quartiles and the median value, 285.4 eV.

Fig. 7. Box plot of C 1s BEs from ion etched elements that captured gases from UHV showing quartiles and the median value, 284.9 eV.

 

 

 

Adventitious Carbon

by M. Biesinger

The C 1s spectrum for adventitious carbon can be fit by defining a peak constrained to be 1.5 eV above the main peak, of equal FWHM to the main peak (C-C, C-H). This higher binding energy peak is ascribed to alcohol and/or ester functionality (C-OH, C-O-C). Further high binding energy components can be added if required. For example: C=O at approximately 3 eV above the main peak and O-C=O at 4 to 4.5 eV above the main peak. One or both of these peaks may have to be constrained to the FWHM of the main peak if they are poorly resolved.

Spectra from insulating samples can then be charge corrected by shifting all peaks to the adventitious C 1s spectral component (C-C, C-H) binding energy set to 284.8 eV. There is certainly error associated with this assignment. Swift [1] lists a number of studies showing errors ranging from ±0.1eV to ±0.4 eV.  “Newer” studies (late 1970’s) range from ±0.1 to ±0.3 eV. “Older” studies (late 1960’s to early 1970’s) were in the ±0.4eV range – however, reproducibility and resolution of the spectrometers of the time may have played a role.  Barr’s [2] work from 1995 states that error in using adventitious carbon is ±0.2 eV.  Our work [3] in 2002 also suggests error in the ±0.2eV to  ±0.3eV range.  Experience with numerous conducting samples (1995 to present) and a routinely calibrated instrument have shown that the C 1s signal generally ranges from 284.7 eV to as high as 285.2 eV [4].

For organic systems, especially polymers, it is convenient to charge correct to the C-C, C-H signal set to 285.0 eV. This makes for easier comparison to the polymer handbook [5] which uses this number for charge correction.

References:
[1] T.L. Barr, S. Seal, J. Vac. Sci. Technol. A 13(3) (1995) 1239.
[2] P. Swift, Surf. Interface Anal. 4 (1982) 47.
[3] D.J. Miller, M.C. Biesinger, N.S. McIntyre, Surf. Interface Anal. 33 (2002) 299.
[4] M.C. Biesinger, unpublished results
[5] G. Beamson, D. Briggs, High Resolution XPS of Organic Polymers – The Scienta ESCA300 Database Wiley Interscience, 1992.