The Materials Analyst, Part 10: Sorting out mixed materials

In this age of recycling, the effective use of reclaimed materials is becoming a potential strategic advantage in reducing cost. Most molders utilize at least some of the scrap that they generate in-house. Some may purchase reclaim streams from outside their own operations, particularly when the reclaim comes from a market such as medical where purity is virtually assured. A medical molder may be prohibited from reusing its reground polypropylene or acrylic and may find a willing market in molders who make less critical products.

For some years now, enterprising compounders have been producing a variety of compounds in materials like nylon and PET polyester using reclaimed material from the fibers
industry. And even the majors have gotten into the act, incorporating postindustrial and postconsumer scrap into some of their standard compounds. They even advertise this fact as the practice of recycling gains legitimacy and is sometimes mandated by some large end users.

While the success of some recycling efforts is well publicized, a great deal has also been written about the fluctuating economics of recycling. If excess capacity develops for a given polymer, prices for the virgin material fall. This forces the price of the recyclate down, reducing the economic incentives to collect scrap--particularly the postconsumer materials that need to go through an extensive process of sorting, cleaning, and granulating. Most people involved with recycling agree that if reclaimed material can be used in mixed form, the economics get much better. The problem comes in identifying what is in the mixed material.

Reclaim Identification

One of our customers came to us with just such a problem. He was purchasing brush fiber reclaim and using it to produce a compound that was then molded into a line of hardware accessories. Through some development work of his own, he had discovered that certain mixed systems could be readily reprocessed and molded without much difficulty. The primary polymer present in the waste stream was PET polyester. Also present were small amounts (less than 10 percent) of polypropylene and polyethylene. These did not present any particular difficulty for the molded product.

However, sometimes PBT polyester would also be present. If the PBT were incorporated at low loadings, the molded parts would still perform satisfactorily. However, when the PBT content exceeded 10 percent in conjunction with the presence of the polyolefins, the product would become brittle. In order for the reclaimed material to be usable, the PBT had to be screened. The problem was compounded (no pun intended) by the fact that the supplier of the reclaimed material denied that any PBT was present in his recycling stream. So when problems with brittle product would arise, they were always blamed on processing.

Melting Point Measurement

In past articles we have discussed the utility of DSC (differential scanning cal-orimetry) to identify various polymers by their melting points if they are semicrystalline and by their glass transition temperatures if they are amorphous. In this case, all of the materials in question were semicrystalline and had well-defined melting points. Screening the materials was simply a question of heating the samples provided to us and observing the melting points.

Figure 1 shows the melting points for pure samples of LDPE, polypropylene, and both PBT and PET polyester. These form the fingerprint against which the unknowns are compared. Our customer had eight separate samples that needed to be evaluated.

Results in initial heating of the first sample are shown in Figure 2. Three features stand out. First there is a noticeable step in the curve near 65C. This is usually a glass transition, and in this case, the temperature is consistent with the T_g of PET polyester. The second event is a small exotherm between 105 and 135C. Finally, the main event occurs between 200 and 260C with a peak at 245C. All three events indicate that the sample is PET polyester. The step transition is the glass transition; the small exotherm is a
recrystallization.

Polymers like PET and PPS are semicrystalline; however, they crystallize slowly. If materials like these are quenched rapidly, they will freeze in the amorphous state. When they are reheated the crystallization will occur once the material reaches a certain temperature. The temperature of this recrystallization and the melting point are typical for PET polyester.

So far, so good; the first sample was PET. However, one of the materials
we were looking for was low-density polyethylene. LDPE has a melting point of 105 to 115C, depending on its exact density. Whenever a material is heated by DSC and it exhibits recrystallization, it's a good idea to heat it a second time to be certain that the first heat finished the crystallization process. Figure 3 shows the results of the second heat and reveals a surprise. A melting event that corresponds to a low-density polyetheylene is present on second heat. It had been hidden by the recrystallization during the first heat because the two events occur at the same temperature. However, the percentage of LDPE was below 10 percent, and the material was usable.

Finding the Culprits

Most of the other materials sent in for evaluation were a different story. The results for the next sample are shown in Figure 4. In this test we clearly see the LDPE with its characteristic broad melting event that peaks at 106C, and the PET melting event is present, peaking at 247C. But we also see a strong melting event that starts below 200C and peaks at 219C. This corresponds well with the melting point of the PBT polyester that our customer was so intent on avoiding. It can be seen that the two melting points are close together and the PBT melting process is not quite complete when the PET melt begins. This makes it difficult to quantify the degree of contamination, but an approximation was possible by comparing the areas under the curves with those of pure polymer.

In this sample we estimated that the material was one-third PBT polyester. When molding was attempted, the parts, in fact, did not work. The test results for the next sample are shown in Figure 5, p. 48. Here we see another mixed system; polypropylene with a melting point of 166C is followed by another combined melting event with both PBT and PET present. Notice this time, however, that the PBT content is much lower. In this case we estimated that it made up 4 percent of the sample. While our customer was initially hesitant to accept the material, a molding trial did prove the material to be usable.

The worst case came from the sample shown in Figure 6. No polyolefins were found in this sample, but the combination of PBT and PET polyester is clearly evident. In this sample, the area under the curve for the PBT is actually greater than for the PET and we estimated the sample to contain 55 percent PBT. After evaluating several other samples we came across a sample that was actually pure PBT.

When this was shown to our customer's supplier, the supplier finally accepted the possibility that PBT was getting into its recycling system and took steps to do its own screening using the same techniques. By successfully eliminating the PBT entirely from the system, our customer was then able to go back to molding good parts out of mixed recylate.