Methodological comparison of extraction methods

The main characteristics of the ideal extraction method are given in Table 3.47, which at the same time are also criteria for comparison of sample preparation techniques. It is unlikely that a unique best method can be defined, which is analyte and matrix independent. Extraction is affected by polymer functionality, molecular weight and cross-linking. Selective extraction of some additives is basically not possible. Hence, the goal of an ideal extraction would be the complete extraction of all additives from the polymer for subsequent chromatographic separation. 

Experimental comparisons may suffer from a lack of optimal conditions for all methods considered or may be based on biased evaluation. It is frequently noticed that results quoted by the preferred extraction technique compare extremely favourably with existing extraction technology. Also, lack of prospects of using CRMs is not helpful for comparisons. However, it appears that for a given infrastructure (R D vs. plant laboratory) and need (routine vs. occasional operations), and depending on the mix of polymeric matrices to be handled, some preferences may clearly be expressed. 

If the focus is on a wide variety of additives in polyolefins only the OSM/TSM procedure of Marcato and Vianello [210] appears to be far reaching, despite the fact that this total design approach must have required considerable effort in terms of method development. On the other hand, in comparison considerably more method development and optimisation appears to be necessary for SFE of a single polymer/additive matrix, but even then still without the demonstrated benefit of a total approach [322]. Where analyte integrity is of greater concern recourse should be taken to less 

Soxhlet extraction is well established, and generally exhaustively extracts all additives. The selection of extraction solvent can make large differences to the extraction time. The generally long extraction times followed by concentration steps may determine losses of volatile or thermally labile components. Because this form of extraction is one of the oldest and still widely used in industry, it is the standard to which many of the newer extraction technologies (which are likely to determine future applications) are referred. However, it should be realised that extraction mechanisms may be different, and thus comparisons are sometimes irrelevant. 

It is not uncommon that extraction techniques are unfairly compared. Appropriate interlaboratory studies are few. Soxhlet and sonication extraction (EPA methods 3540 and 3550, respectively) were compared in an interlaboratory study (129 participants) for PCBs in soil. Results from laboratories using Soxhlet extraction were significantly more accurate than those obtained using sonication, especially at higher concentrations, but with equal precision [196]. This is rationalised by the observation that the Soxhlet procedure presents the sample with fresh solvent so that the extraction solvent is never saturated, unlike the sonication procedure. Sonication is very sensitive to the solvent polarity, nonpolar solvents producing considerably less accurate results than polar solvents. It is not as sensitive to clean-up procedures as 

---