Sample Preparation

Spectra Evaluation

Selecting the Acquisition Conditions

Choosing optimal acquisition conditions for XRF analysis is a complex and critical part of the art. Selecting the proper acquisition conditions can mean the difference between measuring an element at PPM levels, or not seeing it at all. There are two fundamental principles that must be met to achieve optimal analysis conditions.

A) There must be a significant source peak above absorption edge energy of the element of interest. This may be either the K or L edge depending on which one is within the measurable range of the instrument with the preference usually going to K line measurements. The closer the source energy is to the absorption edge, the higher the intensity and sensitivity (counts/sec/PPM) will be for the element of interest. The ideal energy would be precisely at the absorption edge energy, but that is usually not possible.
B) The other fundamental principle is that the background x-rays within the element of interest region should be reduced as much as practical.

The difficulty is that these two principles work in opposition to each other, as the best sensitivity is often achieve when the background is highest, and the background is lowest when the sensitivity is worst. Add to this that the best theoretical detection limits are achieved when the sensitivity is highest, while the net count rate extraction, matrix corrections, and long-term analytical stability are best when the background is lowest. Optimal analytical performance is achieved by finding the best compromise between these two principles, given the instrument hardware.



Sample Preparation Equipment

Sample Cups
There are a wide variety of sample cups available with many features. To best compare them we will compare each feature, and show when it is beneficial.

Cups come in a variety of diameters from 20 mm to 40 mm. For the most part it comes down to individual user preference or the size of the hole provided by the manufacturer but there are some important criteria for selecting the diameter. Ideally the inner cup diameter should be larger than the spot size on the sample. A quick way to determine spot size is to put a drop of oil or water on a thin sheet of paper and analyze it for 10 minutes at very high power. The x-rays heat the area they strike causing the liquid to evaporate leaving a visible circle. Or just consult the instrument manufacturer.
The penetration depth of the x-rays for the elements of interest is also important. For characteristic x-rays above 10 keV, penetration depths of more than 1 cm are possible. With a 20 mm cup the x-rays may hit the cup wall. Larger diameter cups, 40 mm, are recommended for higher atomic number elements in transparent matrices to achieve the best performance. Alternatively the cup position must be tightly controlled.
The third criterion is reproducibility of the film. The larger the diameter the greater the range of film heights at the center and the higher the probability that some cups come out wrinkled. For ease of assembly and best reproducibility 32 mm or smaller cups are often preferred.
A 6.3 mm cup is offered by Spex that is useful when only a limited volume of sample is available.

Height selection is also a function of penetration depth. It is important to know the angle of the x-ray source when making this determination. If it is at a 45 degree angle there is generally no benefit to having a cup that is taller than it is wide. If the source points straight upward then a deeper cup is beneficial when analyzing high energy x-rays from a x-ray transparent sample.

Single Vs Double Open Ended
The obvious benefit to single open-ended cup is that it saves the step of putting a cover on the cup, and they are usually less expensive. The problem with them is that the amount of air that gets trapped varies causing the window film to bulge different amounts each time. Even after venting the film will remain stretched to a degree. A single open-ended cup will usually have poorer reproducibility. This is exacerbated by the fact the fact that a bubble can form over the pin hole causing the cup to pressurize and the film to bulge enough to affect the results. In many cases this error is minor relative to the benefits, so they are popular for analyzing high concentration elements in solution.
Double open-ended cups offer many other benefits. They offer a variety of choices for covering the sample. Even if a film cover is used, the pressure is distributed over both films causing less bulging. They are highly recommended for powdered samples since it allows the sample to be manually pressed. They are recommended when the highest degree of precision is required.

Collar Vs no Collar
Collars are intended to prevent wicking of the sample along the film and out of the cup. They are recommended for all liquid applications. Tape can be used instead of a collar. The Somar cup design uses a tall bottom retaining ring designed to eliminate the need for a collar and it works provided that the sample level is not above the top of the ring.

Cup Covers
There is a lot of variation in cup covers. It is common to use no cover at all for pressed powders, and some brave technicians will not cover solutions or powders.
Window film is a common cover but it is labor intensive to install and causes both the top and bottom films to bulge slightly affecting reproducibility. It is recommended to put a pinhole it to equalize pressure. If the top film gets wet it is possible for a bubble to form across it holding in additional pressure. Cups with and without this bubble will read slightly different in some difficult applications.
A solid snap on cover is easier to install, but it flexes the bottom window as much as the single open ended cup and is prone to having problems with bubbles over the vent hole if it is allow to get wet. The solid covers area few cents more expense than a simple ring.
Solid vented covers are an excellent choice. They are easy to install, do not pressurize the cup and only cost a few cents more than a simple ring and the same as a solid cap. The holes should be large enough that even when wetted, bubbles do not form over the surface.
A variety of baffled cap designs are offered. Each is designed to slow the evaporation process while preventing the cup from becoming pressurized. Some are designed with the intent of causing the sample to reflux. These designs are slightly more expensive than a solid vented cap. Many operators have failed to find any benefit versus a solid vented cap, while others swear by them. They may be worth experimenting with when analyzing highly volatile samples.

Inverted Cups
Cups are offered that allow the samples to be inverted. One vender, Somar, designed the cover so that it slightly depresses the Teflon window film covering the sample. The cup is filled to the top and the Teflon film placed over it leaving no air bubbles. The cups can then be inverted for use in instruments with downward facing optics. The problem with this design is that the film height changes from cup to cup, but many instruments with downward facing optics have movable sample stages that can be used to position the sample reproducibly enough for quantitative analysis. Other innovative designs have also been produced for this purpose.



Sample Preparation of Films

Window (aka Sample Support) Film
A variety of window films are available to fill a variety of needs. Decisions on the type of film to use are made based on factors such as cost, x-ray transmission, chemical resistance, reproducibility, and ease of use.

The most common types of films are addressed below.

Mylar is often the first choice people make because it is low in cost and has a high tensile strength so it gives very reproducible results. It has worse transmission characteristics than most other films, but because it is so strong it is available in a thin, 1.5 um form. The 6.3 um material is often used when analyzing most high Z elements, while the 3.6 um form is used for elements between Al and Ca. Mylar’s weakness is that it has poor resistance to acids, which eliminates a broad class of solution applications. The 1.5 um film is difficult to handle, and is not widely used for routine analysis.

Polypropylene is the next most popular film. It is the next least expensive film after Mylar and is resistant to most acids. It also has better x-ray transmission than Mylar. It usually comes in a 6.3 um thickness and is useful for analyzing elements from Al up in the periodic table. Polypropylene’s weakness is that it has poor tensile strength and stretches readily, meaning it has poorer reproducibility than Mylar. Also, while it is resistant to hydrocarbons it absorbs them, softens and wrinkles over time, so it is a poor choice for use with oils and solvents.

Prolene and Spectrolene
These films have similar chemical resistance properties to polypropylene, and so are useful for acid analysis. They are also a little better than polypropylene for hydrocarbon analysis. They have superior x-ray transmission and are available in 6 um and 4 um thicknesses. These films are highly recommended when analyzing low atomic number elements such as Na and Mg in solutions. Because of these properties Spectrolene and Prolene have become the favored general-purpose film in many XRF laboratories. The 4 um film can still be handled reasonably well by an average technician. The major drawback is the comparatively high cost of these films.

Polycarbonate is similar in cost and transmission to polypropylene. Its advantage is that it’s tensile strength is similar to Mylar, but with better transmission characteristics. And it comes in 5 um and 2 um thicknesses. It has very good reproducibility, close to Mylar. It is also highly resistant to hydrocarbons, so it is the preferred film for measuring low atomic number, Na-Cl, elements in oils and solvents. This material is known to tear easily, particularly when assembling Somar style cups. The 2 um film, like other very thin films, is difficult to work with on a routine basis.

Kapton (polyimide)
Kapton film has a distinct yellow appearance. It comes in a thick 7.6 um film and has poor x-ray transmission characteristics. It is also the most expensive of the XRF window films. With all these negatives it is not hard to imagine that it is rarely used. Kapton is however resistant to highly aggressive acids, such as concentrated SO4, HF, and aqua regia, that destroy other films.

Teflon, Tefzel, PFA
The most common use of Teflon is as a microporous film that is used to cover cups of loose powders when they need to be analyzed in a vacuum. It allows the air to escape without allowing powder to be sucked into the vacuum pump, and distributed everywhere in the analysis chamber.
Tefzel and PFA films are not available in very thin films, 12.5 um is the thinnest size. They are not very x-ray transparent, and are not a stock item for most XRF film venders. They can be used with some acids when even Kapton is not good enough, and they can be used as a window lining in some on-line applications.

Cups with solid Be windows are available for analyzing very light elements. They are very expensive costing a couple hundred dollars each. They must be cleaned between analyses, and can be destroyed by most acids. They are seldom used for general analytical work.

Specialty Films
Numerous other specialty films have been used, but are generally limited to special laboratory methods.



Sample Preparation of Liquids

Liquid samples can be the easiest or most difficult samples to work with depending on the composition and stability. A sample cup is filled about ¾ full and then presented to the analyzer.

The problems with liquid samples are that they evaporate, stratify, and precipitate. The liquid may attack or be absorbed by the window film, wick up and out of the cup.
Because of these issues liquid samples should be freshly prepared, preferably immediately prior to analysis, although some liquids are stable for a day or more. Solutions should be well mixed prior to pipetting them into a sample cup. The sample should be taken from the center of the container since some components may concentrate on the walls. There are some techniques that can be used to overcome most of the other problems.

All liquids have some vapor pressure at standard pressure and temperature. Some low vapor materials such as water or mineral oil may be stable for a day or more, while gasoline and other volatile liquids may only be stable for a few seconds. Highly volatile samples should be prepared and analyzed one at a time and the time from pouring to starting the analysis should be consistent within a few seconds.
A cover or cap may be placed over the cup to reduce evaporation, but the film will bulge causing poor reproducibility. Single end cups and caps for double-ended cups can be punctured with a pin, and some have a snap off device that leaves a hole. Window film covers can also be punctured to relieve pressure. Most cup manufactures also make baffled cups that are designed to minimize evaporation and in theory the sample.

Liquids stratify when they either contain two or more immiscible liquids like oil and water, or if they contain several molecules that have different density such as a crude oil. Various organo-metalics are also higher in density and tend to settle out over a long period of time. Most liquids need only be well mixed prior to pipetting it into a cup, and they will be stable during analysis. Others will stratify during analysis and must be analyzed quickly.
Immiscible liquids are a special challenge. Sometimes it is possible to produce a meta-stable emulsion. Others times it may be necessary to analyze the two components separately. Lastly solidification techniques can be attempted.

As with stratification, liquid samples that precipitate may be well mixed and analyzed one at a time if the particles stay suspended long enough. Normally the supernatant liquid is pipetted off the top and measured separately from the precipitate, which should be dried, weighed, and measured as a powder. Solidification techniques have also been used successfully with rapidly precipitating samples.

Wicking happens when the liquid is hydrophilic with respect to the window film. The liquid is drawn along the film, out, up, and over the bottom retaining ring. This can also happen even with a hydrophobic liquid because of gravity, when the liquid level is above the height of the bottom retaining ring. One cup vender attacks this problem by having a taller bottom ring, while others use collars that slide up the outside of the cup holding the window film above the sample height. Another technique that works in a pinch is to use tape around the outside of the cup to hold the window film up.

Alternative Liquid Sample Preparation Methods
Since many liquid samples are inherently unstable there are a number of alternative methods for stabilizing the samples. There are also a number of pre-concentration techniques that are available that have been used with some success.

Liquid Sample Solidification
Solidification is one useful method when the sample contains immiscible liquids such as oil and water, or it the precipitates rapidly. Several materials including, cellulose, carbon, gelatin, and alumina, have been used successfully as solidifying agents. Lower atomic number solidification agents are preferred when analyzing low Z elements. The ideal mixing ratio must be determined experimentally.

Thin Film Preparations
Popular method to attempt, is to deposit a sample on a thin sample support and then measure it either wet or after drying. By reducing the sample thickness it is possible to reduce the background while keeping the elemental net intensities high, thus improving the detection limits. Drying the sample helps even more since much of the background due to backscatter is due to the hydrogen and other low atomic elements in the base matrix.
Common support materials include filter paper and IR cards. A few, typically 5-50, microliters of sample are be pipetted onto the support, and allowed to soak in and distribute. The method has also been done using an atomizer to spray the sample. It can be presented wet provided the moisture content and distribution is consistent from sample to sample or else it can be dried. The primary problem with these techniques is that they are often not reproducible enough for routine laboratory work.

Sample Concentration
There are several concentration/thin films techniques that have been used for XRF analysis. The simplest involves filling up a sample cup and dying it. This works best when the matrix is a highly volatile solvent.
If the elements of interest are particulates in suspension then a filter can be used to filter a large volume of fluid. Then the filter can be analyzed wet or dried. One instrument vender uses this method to analyze trace metals such as iron in nuclear reactor coolant.
Another method involves the use of ion exchange filters or resins. A large amount of fluid can be moved through an ion exchange medium removing the ions of interest. The medium can be measured to determine concentrations of the elements of interests. Filters can be analyzed wet or dry. It is important to select an ion exchange medium that does not interfere with the XRF analysis.


Sample Preparation of Loose Powders

The simplest approach to analyzing a loose powder sample is to simply fill a sample cup approximately ¾ full and analyze it without any additional preparation, or simply tapping the cup on a clean surface to pack it to a more consistent density.

This method is satisfactory in same cases where the reproducibility requirements are not very strict. The greatest advantage with this method is that it is easy. Spinning the sample may improve the consistency of the measurement by averaging over a larger area.
There are many shortcomings with this technique; the bulk density may not be consistent, grain size variation cause the readings to vary and finer grains can be forced to the surface during tap packing. Because of these problems this method works best with homogenous material that has been dried and ground to a uniform grain size. Of course, these recommendations apply to every sample presentation method.

Manually Pressed Powders:
Several venders offer a sample press, a modified arbor press, that is designed to press a sample within a sample cup. The sample cup is first filled ¾ full or more and then positioned under the press and compacted to 15 to 30 Newton-meters. Higher pressures can be achieved depending on the press. The cup must be held down firmly to prevent the film from bulging, and it is usually a good idea to place a Kimwipe or other clean disposable material under the cup. Place another Kimwipe over the cup before pressing if the sample material is likely to stick to the press. If the material compresses a lot it may be necessary to refill and press 2 or 3 times to achieve a reasonable sample depth.
Manually pressed samples can still be prepared quickly since pressing generally takes less than a minute. Because tap packing is unnecessary this method eliminates the problem with finer particles settling. Ultimately, by giving the samples more uniform bulk density it is possible to achieve analytical reproducibility that is often no more than 10% worse than with a hydraulically pressed sample.

Hydraulically Pressed Powders:
The preferred method for analyzing powders or samples that are usually ground to a powder to make them more homogeneous is the hydraulic pressed pellet. Several manufacturers make hydraulic presses that are capable of pressures ranging from 10 to 50 tons or more. The press uses a die set to contain and form the sample during pressing. There are a few types of indie sets and a lot of variations to the methods incorporating a lot of individual creativity, but one basic outline follows.
The sample is first dried and ground to a fine consistency, 400 mesh or better is recommended. Remember that the x-ray wavelengths are still substantially smaller than the particles. Next the sample is usually blended with a binder that helps hold the pellet together, although some materials don’t require it. Selection of the best binder for a given material is an art form itself and is discussed in more detail on the binder page. A ring and puck mill or mixer mill is very useful for both the grinding and blending steps.
To prepare the press, the die set must first be cleaned with methanol or other solvent. The backing, usually an aluminum cap, is inserted into the die. A specific weight of sample is then poured into the cap, usually 5-10 grams. It is important to keep the mass constant because the sample may not be infinitely thick at high x-ray energies. Next a polished pellet is placed over the sample to produce a smooth finish. The plunger goes in after that, and then the die set is positioned in the press. Follow the press instructions for pressing the sample to 10-20 tons, and holding it for a period of time usually from 10-100 seconds. Different materials produce better pellets at different pressures, so finding the best pressure may require some experimentation. Once the best pressure is determined however all samples of that type must be compacted to the same pressure and hold time to achieve optimal analytical results. The pellet is then removed from the die set, taking care not to crack it in the process.
The finished pellet is uniform in composition, density, and mass per unit area, and has a smooth finish. These are all highly desirable traits for an XRF sample, so many operators will spend the needed five to ten minutes it takes to prepare the sample. The only remaining disadvantage with pelletized samples is that there are still matrix affects due to the grain size being larger than the x-rays.



Sample Preparation : Solids

Sample Preparation of Plastics

Plastic samples present some special challenges since they come in numerous forms, powders, pellets, fibers, sheets, and variously shaped solids, and cannot be prepared by most of the usual methods.

Plastic Powders:
Plastic powders like other powders can be presented in a sample cup either loose or manually pressed. Plastics are resistant to hydraulic pressing, unless heated to 100 to 150C before pressing or pressed using heated platens. Manually pressed samples are recommended for most applications.

Plastic Pellets, Beads, or Regrind:
Plastic Pellets can also be analyzed loose in a sample cup as long as the pellets are fairly uniform in shape. Using a sample spinner helps with the analysis. It is also possible to heat pellets to 100 to 150C and press them into a pellet cap, or press them by normal methods. For most applications loose pellets presented in a spinning sample cup will give good results. Coarse regrind material may also be analyzed in this manner.

Plastic Fibers:
Fibers are particularly tricky to work with. They can be coiled or balled up in a cup, but the reproducibility is generally not very good. It helps somewhat to insert a weight in the cup on top of the sample. An alternative approach that often works better is to win the fibers on a spool. If the wound spool is thick and fairly uniform the results may be satisfactory. Ultimately most fibers must be ground to a powder using a freezer mill or other suitable plastic grinding apparatus and presented in powder form.

Plastic Films, Sheets and Fabric:
Plastic films, sheets and fabric, can often be analyzed simply by cutting them to size and placing them in the analysis position. Several layers of thin films and sheets can measured to increase the net counts for elements with higher energy x-rays. It is also useful to weight some types of films or fabrics with a flat piece of solid metal or a metal ring, provided a metal is selected that doesn’t interfere with the analysis. It is important that the part of the sample in the analysis region be uniformly flat and smooth, or else there will be reproducibility errors.

Solid Plastic Samples:
Solid plastic parts usually make excellent samples. It is only important to remember that the surface must be uniformly flat and smooth in the analysis area. Many plastics shrink during curing plastic pieces giving them a concave surface. As long as all the parts have the same shape and can be position reproducibly, it may still be possible to analyze them without flattening the surface in some way.
It is important to note that the composition of the surface skin is sometimes slightly different than in the interior of the part, since the molding process forces some additives to the surface. To measure the average interior composition it may be necessary to grind or sand off the surface layer. It is then important to make sure that the surface finish is repeatable.