Field-Flow Fractionation - The Universal Separation Principle for Particle and Macromolecule Characterization

Dr. Tino Otte, Dr. Thorsten Klein, Dipl.-Chem. Evelin Moldenhauer
Postnova Analytics GmbH, Max-Planck-Straße 14, DE-86899 Landsberg / Lech

High molecular weight polymers, bio molecules or particles in a range of a few nanometers up to several microns are more and more important in science and technology. Many applications are based on the specific properties of these materials. For this reason, an exact knowledge of the molar mass or particle sizes distribution is essential for later application as well as for optimization of the production processes or quality control. In addition the knowledge about the chain structure of macromolecules or biopolymers as well as about the shape of particles is often necessary. For this reason those samples, which often can be very complex systems, have to be separated according to parameters such as e.g. hydrodynamic volume or molecular weight. After the separation process the resulting narrow dispersed fractions are detected with different techniques. Thus, an optimal separation is necessary to ensure that the obtained values are correct, otherwise the best detector will deliver wrong or at least partially erroneous results [1-3]. So far the separation of very large or highly adsorbing material was partially realized by chromatographic techniques such as e.g. GPC/SEC, HPLC, or non-chromatographic methods like e.g. the ultracentrifugation, solubility fractionation or microscopy.

The chromatographic methods are mostly based on the separation of the analyte inside a porous stationary phase, this leads to particularly high shear forces and there is a permanent risk of adsorption present. The most chromatographic techniques show high performance for low or medium size material but for higher molar masses or increased particle diameters they are strongly limited and deliver insufficient or even no separation. Offline techniques such as ultracentrifuge, solubility fractionation or microscopy offer better results but they are mostly not universally applicable, very expensive or the separation/analysis processes are extremely time consuming and complex.

The characterization of polymers and particles by field flow fractionation offers the possibility of a universal separation with high resolution [4]. Also very large structures can be analyzed [5] under shear free conditions and without filtration by porous material or unwanted adsorption on the surface of a stationary phase. The analyte species, which are dissolved or dispersed in the solvent, will be separated in an empty channel. This separation can be according to size, chemical composition or density. An external field force, which acts perpendicular to the carrier-flow, is the source of sample separation. Depending on the kind of FFF-technique a cross-flow (Asymmetric Flow FFF/AF4), temperature (Thermal FFF/TF3) or a centrifugal field (Centrifugal FFF/CF3) is used.

The cross-flow in AF4 leads to an accumulation of the analyte species at the channel bottom. The size-dependent diffusion abilities of particles or molecules lead to an arrangement in different layers of the parabolic flow profile inside the channel. As a result the analyte components with small hydrodynamic diameter will elute first and the larger molecules or particles will elute later.

In Thermal FFF a temperature gradient is used for the separation of the sample. Since the thermal diffusion coefficient also depends on the chemical nature of the analyte this method can be used for separation according to both, size and chemical composition. As a result components with the same hydrodynamic radius but different composition can be separated by TF3.

The Centrifugal FFF utilizes a centrifugal field which is realized by rotation of the FFF-channel. This method allows an improved separation according to size and density. Consequently the separation of particles with the same diameter and different density can be performed. The high selectivity for dense material causes an improved retention also of very small particles despite their high diffusion coefficients. These advantages make the CF3 to an ideal method for the analysis of compact particles.  
All FFF separation fields can be adjusted in their strength. A gradient function of any shape can be applied. Thus the separation efficiency can individually be adapted with regard to each separation problem. A tailor-made calibration curve can be applied. In FFF the separation is not limited like it is the case in SEC where the separation can only be in the range of the default calibration curve which is defined by the given column-combination.

In addition FFF can be used together with various detector combinations. The most important setup is multi angle light scattering (MALS) together with concentration sensitive devices such as RI, UV or IR, which enables to detect the molar mass as well as the radius of gyration simultaneously [6]. Other combinations, such as for example with dynamic light scattering (DLS) [7], viscometers [8] or spectrometers like ICP-MS [9] are easily realizable.

Application of the separation techniques

The AF4 technique can be used for characterization of both, polymers and particles. The separation depends on the hydrodynamic radius of the species which is a very universal character of the most samples. The fractogram in Fig. 1 demonstrates the wide separation range and the variability of the mobile phase and the possible analytes. A separation with high resolution can easily be performed over a huge size range within a relatively short analysis time.

The performance of the AF4 compared with conventional methods can easily be demonstrated by the example of the separation of e.g. high molecular weight or highly branched polymers. In Fig. 2 two examples of the SEC and AF4 separation of synthetic polymer material are presented. For both samples the high molar mass is combined with an increased degree of branching which leads to important application properties such as high stability and low viscosity of the melt or solution.

Fig. 2: Separation of various synthetic polymers with AF4 and SEC. The radii and molar masses were determined by MALS detection.

The comparison of SEC and AF4 separation impressively shows the magnitude of the FFF principle and the variety of additional information such as the correct molar mass or the complete branching information from AF4. On the contrary, a correct evaluation of the SEC-MALS-data is not possible due to the high amount of shear degradation and the high partial retention of large macromolecules. The AF4 can be conducted at temperatures ranging from 5 to 220°C. Thus, the method can be applied for the analysis of the most polymeric materials.

Important commodity polymers, like e.g. high molar mass polyolefins are particularly shear sensitive and often pose significant amounts of branching which lead to various problems in the SEC. In Fig. 2 the separation of LDPE with AF4 and SEC is compared and the additional information from AF4 are pointed out. Also rubbers can be fully analyzed by AF4, while the SEC provides more or less insufficient information. In the analysis of the rubber sample which is shown in Fig. 2, a huge amount of the high molar mass species was not found by SEC, while AF4 shows the true molar masses and branching information. Since the samples are not filtered by a column during AF4 analysis, a high molar mass shoulder gets visible in the Light scattering signal of the rubber in Fig. 3. This shoulder represents the gel content of the sample. The low slope of the Rg-M-relationship indicates highly branched or cross linked material with very compact coil structures in solution. Thus AF4 often offers a new perspective on many interesting samples, which were claimed to be already well understood in the past.

In a similar way like for synthetic polymers also natural material such as proteins, starches and alginates can be separated. In addition AF4 is also well suited for characterization of higher biological structures such as micelles or viruses as well as for dense particulate samples like e.g. carbon nanotubes, colloids or toner particles.

In addition to the advantages of the shear-free und universal separation which is realized by the adjustable cross-flow gradient, AF4 offers also other possibilities for the optimization of the sample analysis. A focusing of the sample at the beginning of the separation process allows e.g. the enrichment of highly diluted solutions. The injected sample is focused in a narrow zone at the channel entrance by a second inlet stream (focus flow) which counteracts the carrier flow. As a benefit additional band broadening which can be a consequence of the longitudinal diffusion of the analyte is repressed by the focusing even in case of excessive long injection times. Finally the signal-to-noise ratio of the later detected signals is increased. A second possibility for signal improvement is provided by the slot outlet technology. Here, the upper solvent layer is removed by an additional pump. This layer contains no sample and as a result the dilution inside the channel is decreased which improves the signal heights.

In addition to size-separation the Thermal FFF (TF3) provides also a separation according to chemical composition. An example for this selectivity is shown in Fig. 3 on the hand of the separation of polystyrene (PS) and polymethylmethacrylate (PMMA) with the same hydrodynamic radius.

Fig. 3: Separation of two narrow distributed polymer standards of PS and PMMA - insufficient separation in the SEC due to the similar hydrodynamic volumes and successful separation with Thermal FFF caused by the different thermal diffusion coefficients of both materials.

As it is visible in Fig. 3, no separation of the two different standards with the same hydrodynamic radius can be archived by SEC. The slope of the molar mass in the elugram is a result of the different refractive indices of the two co-eluting species in THF. However, the Thermal FFF shows a good separation into two separate peaks. PMMA elutes later due to its higher thermal diffusion ability. The improved separation allows an individual analysis of the molecular weights of both samples with different dn/dc (the refractive index increment dn/dc considers the change of the refractive index of the analyte solution with the concentration, which is important for the determination of the molar mass from light scattering data). The constant molar mass values of both peaks from TF3 confirm a correct separation of the narrow disperse polymer standards.

A special kind of FFF which is especially suited for particle analysis is the Centrifugal FFF (CF 3). The rotation of the round channel provides a centrifugal field which enables to separate according to size and density. The high selectivity of the centrifugal field for very dense material leads to improved resolution with this method in many cases. In Fig. 4 the separation of gold rods, of different size and shape, with AF4 and CF3 is shown

Fig. 4: Separation of gold nano particles of different shape and size with AF4 and CF3. In addition the efficiency of the FFF separation was confirmed by electron microscopy.

Conclusions

It was demonstrated that the FFF technology represents a universal applicable separation method with high flexibility. It offers excellent performance even for the separation of very large particles or ultra high molar mass polymers. In addition, the special properties of FFF enable to separate samples which have not been analyzable with other methods until now. The variability of the separation and features such as focus or slot outlet offer new opportunities for the optimization of the sample analysis. Due to the huge flexibility of FFF against different solvents and sample materials allow to perform an optimized separation in case of the most analytical problems. Due to its numerous benefits and the high separation stability of the modern FFF-systems it is likely that the FFF technology will become one of the most commonly used analytical separation techniques in the near future.

Literature

[1]     M. Parth, N. Aust, K. Lederer, Int. J. Polym. Anal. Charact. 8, 2003, 175.

[2]     N. Aust, J. Biochem. Biophys. Meth. 6, 2003, 323.

[3]     M. D. Zammit, Polymer 39, 1998, 5789.

[4]     F. A. Messaud, R. D. Sanderson, J. R. Runyon, T. Otte, Prog. Polym. Sci. 34, 2009, 351.

[5]     S. K. Ratanathanawongs, J. C. Giddings, Anal. Chem. 64, 1992, 6.

[6]     T. Otte, R. Brüll, T. Macko, H. Pasch, J.  Chrom. A 1217, 2010, 722.

[7]     K. D. Caldwell, J. Li, J.-T. Li, D. G. Dalgleish, J. Chrom. A 604, 1992, 63.

[8]     E. P. C. Mes, H. de Jonge, T. Klein, R. Welz, D. T. Gillespie, J. Chrom. A 1154, 2007, 319.

[9]     M. Hasselöv, B. Lyven, C. Haraldson, W. Sirinawin, Anal. Chem. 71, 1999, 3497.

More about Postnova Analytics
Your browser is not current. Microsoft Internet Explorer 6.0 does not support some functions on Chemie.DE