Sample preparation for size analysis of diesel soot particles using field-flow fractionation

Citation
Ws. Kim et al., Sample preparation for size analysis of diesel soot particles using field-flow fractionation, J LIQ CHR R, 24(13), 2001, pp. 1935-1951
Citations number
26
Categorie Soggetti
Chemistry & Analysis","Spectroscopy /Instrumentation/Analytical Sciences
Journal title
JOURNAL OF LIQUID CHROMATOGRAPHY & RELATED TECHNOLOGIES
ISSN journal
10826076 → ACNP
Volume
24
Issue
13
Year of publication
2001
Pages
1935 - 1951
Database
ISI
SICI code
1082-6076(2001)24:13<1935:SPFSAO>2.0.ZU;2-1
Abstract
Sample treatment procedure, and the experimental parameters that need to be considered in size analysis of diesel soot particles using field-flow frac tionation (FFF), are discussed. A probe-type sonication of 10 min in water containing 0.05% of Triton X-100 provides well-dispersed suspension of the soot particles. Sample treatment is found to be important for reduction in particle-particle interaction and, thus, for preparation of well-dispersed soot suspension. FFF elution profile of untreated suspension has a long tailing that does no t return to the baseline until the external field is removed. The scanning electron micrographs show that the untreated suspension contains aggregated particles. Extraction of chemicals is also found to be important for prepa ration of well-dispersed suspension, The suspensions prepared without the e xtraction step, have broader and larger size distributions than those prepa red with the extraction step, due to the presence of soluble organic fracti ons (SOF) that cause particle-aggregation. FFF provides an excellent repeat ability in size analysis of soot particles. The standard deviation of the m ean diameter measured by FIFFF was 3.6%. Both Flow FFF (FIFFF) and sedimentation FFF (SdFFF) are tested for size ana lysis of soot particles. SdFFF provides higher size-selectivity than FIFFF and is less problematic in terms of the particle-channel interaction. Unlik e in SdFFF, FIFFF analysis does not require the density information and, th us, the fractogram can be directly transformed to the size distribution. Bu t FIFFF provides lower size-selectivity and has narrower dynamic range than SdFFF.