Detective Work With Fourier Transform Infrared (FTIR)

By Michael Sepe, Materials Analyst

The learning curve for using infrared spectroscopy to identify materials is worth climbing just for the technology’s versatility.

Infrared spectroscopy, often abbreviated in the literature as FTIR for Fourier Transform Infrared, is a versatile technique for identifying materials, detecting localized contamination, and checking the composition of unwanted substances that appear on mold surfaces. With today’s instruments, a probe can be brought into direct contact with the area of interest on a molded part or on a swab wiped across a mold surface to determine the makeup of a sample. Here are four examples of how to use this technique.

Case #1: Real-world Experience
The first case involves identifying the composition of an elastomeric earplug. Figure 1 shows an infrared spectrum for a core sample from the molded part. Cutting into the part and examining the core material is an important consideration in performing FTIR using direct contact methods. The surfaces of molded parts may have picked up a variety of surface contaminants in the process of being shipped and handled. Looking at an interior location avoids the problem of mistaking a surface contaminant for a real ingredient in the compound.

To the uninitiated, the results from an FTIR scan are not terribly helpful. The technique is useful because the chemical bonds that hold molecules together vibrate—stretching, rotating, and rocking at particular frequencies. Many of these frequencies are in the infrared region of the electromagnetic spectrum, particularly for organic materials like polymers and their additives. Infrared radiation directed at a surface or passed through a film will be absorbed at the wavelengths that correspond to the frequencies of these vibrations.

The peaks in an absorbance spectrum, like the one in Figure 1, represent the wavelengths where the vibrating chemical bonds absorb infrared radiation. These correspond to particular types of bonds.

Another method involves measuring the amount of radiation transmitted through a material. In this case, the amount transmitted declines at the wavelengths where a substance absorbs, and the results appear to be upside-down as they are reported in percent transmittance instead of percent absorbance. However, the wavelengths at which the peaks occur for a given type of chemical bond are unchanged regardless of which method is used.

For those who struggled through organic chemistry, spectral interpretation was part of the coursework. Students were given tables that detailed the spectral location for different chemical bonds and were then presented with a spectrum of an unknown material. Using the table, they had to identify the unknown substance.

Out in the real world today, we use libraries of known substances and computer-search these libraries to match the unknown to something in the library. Like most technological advances, this has both benefits and problems. The libraries can provide quick answers, while a manual interpretation can be very time consuming.

However, many polymer compounds are a mixture of two or more materials. The computer does not know this and will attempt to account for everything it sees with a single match. It still takes a skilled analyst to look at the results and sift through the details. It also helps to have some knowledge of the commercial environment since this can provide a reality check on the information that the matching software provides.

Figure 2 shows the match made to the elastomeric earplug. Most of the absorption peaks are accounted for by a polymer identified chemically as a polyisoprene:styrene. A good commercial match can be made with the Kraton D materials and a check of the hardness showed that it was in the range expected for this family of materials.

However, there are two absorption bands that appear in the sample spectrum that are not contained in the library match. The software returns a match with poly(dimethylsiloxane). Translation: silicone. Unfortunately, FTIR alone cannot determine if this represents a separate silicone polymer mixed with the primary material, a silicone functionality chemically incorporated into the Kraton as an integral part of each molecule in the compound, or a silicone oil added as a lubricant or release agent. For this type of detail, more sophisticated techniques are needed. However, armed with this much information, a call to a material supplier may result in some commercial options.

Case #2: Identifying Inclusions
This case involves clear inclusions embedded in a molded polyethylene part. These inclusions caused weak spots in the parts and also plugged the hot tips in the mold. The inclusions were sufficiently distinct to enable a positive identification without interference from the surrounding polyethylene. Figure 3 shows a spectrum for the clear inclusion and provides a perfect match with a styrene-methyl methacrylate (SMMA) copolymer. This same spectrum was obtained for the material pulled from the hot tip.

Very often the first question posed to the analyst when reporting results like this is, where did the material come from? Unfortunately, only the people involved in the manufacturing process can answer this. The test can tell the client what it is looking for. But the client must then check its plant operation, material handling procedures, and incoming material quality to identify the source. Any suspect material found during this investigation can be tested to see if it’s a match.

Case #3: Hot Runner Buildup
One of the biggest problems in processing is buildup in hot runner systems and on mold surfaces. Substances associated with this type of problem contaminate the molded parts, clog vents, stain mold surfaces, and can be a major cause of rejected product and downtime. Identifying these compounds can be very difficult without the use of techniques like FTIR. If the offending material can be scraped off the mold details or even wiped onto a cotton swab with a solvent, FTIR can very often identify the problem.

Figure 4 shows a spectrum for an oily residue building up inside a hot runner system after a relatively short run time. Material surrounding the oily residue was waxy and discolored. The result shows a reasonably good spectral match with polypropylene, which was the material being processed through the hot runner system.

However, there is one glaring difference between the spectrum for the residue and that of the known polypropylene: a very large absorption peak that appears in an area where polypropylene has no chemical bonds. This region is associated with the presence of substances that contain carbon double-bonded to oxygen (C=O), known in organic chemistry as a carbonyl. This type of bond is present in a variety of compounds including esters, aldehydes, ketones, and organic acids.

The breadth of this spectral band tells us that most or all of these substances are present in the residue. As it turns out, these substances are all byproducts of the oxidative degradation of polypropylene. The oil and the discolored waxy material represented different and very severe stages of polymer degradation, indicating either very poor stability in the polymer or very poor temperature control in the hot runner system.

Case #4: Check the Process
Figure 5 shows a spectral result for a residue found on a mold surface and in the vents in a mold running a nylon material. A very good spectral match is found with a member of a chemical family known as long-chain amides. These are frequently used as lubricants in many types of polymers, including nylon. But a much more involved analysis would need to be performed on the raw material in order to confirm that this is the source of the plateout. Since this additional testing is quite expensive, it’s a good idea to look first at the molding process and the condition of the mold with an emphasis on venting.

It is important to note that additives of this type melt and even boil at typical melt processing temperatures for polymers like nylon. Running a process on the high side of the melt temperature range is more likely to drive additives out of the material. Normally, these byproducts exit the mold through vents in the tool. But venting in injection molds is often inadequate. Vents frequently do not extend around enough of the parting line and may be too shallow to perform their intended function.

This is particularly true of escapes, the deeper channels designed to carry the gases that run ahead of the melt front out to atmosphere. Escapes can be any depth; they do not connect to the parting line and so will not cause flash. Most material suppliers recommend escape depths of .030-.040 inch spaced about 2-3 inches apart around most of the cavity perimeter. But there are a lot of molds out there with escapes that are .005-.010 inch deep and that are only cut along the side of the cavity judged to be the last place to fill. A little knowledge can be a dangerous thing. Before you call up your material suppliers, wave an FTIR scan in their faces, and harangue them about too much lubricant in their material, check your process and your mold. This type of pain is often self inflicted.

June, 2006 - Reprinted with permission from Injection Molding Magazine. Copyright © Canon Communications LLC.