COMPARISONS OF METHODS

In short, micrometeorological methods generally meet the nondivergence criterion more frequently than do surface sampling methods satisfy representativeness for the surrounding area. 

There are two important reasons for comparing the performance of different particle sizing techniques/instruments for a particular apphcation. First, it is necessary to establish their reliability or, in the case of certain instruments, to verify that they are capable of yielding substantially the same particle size or PSD. The degree of success depends not only on the techniques in question but also on the nature of the system to which they are applied. Second, it may become clear that the use of two (or more) different measurement techniques to characterize a given kind of material may yield additional information that might not be obtainable from a single analysis method and which therefore would otherwise be overlooked. 

In the polymer field, by the natiue of the production process, it is obviously important to be able to determine the end-point of the reaction. It is therefore necessary to choose a particle sizing technique that is appropriate and effective, one that can yield reliable and acciuate measures of the mean diameter and standard deviation of colloidal, mostly-submicron emulsions. In production environments it is particularly important that the size analysis instrument be fast, easy to use and able to reveal reproducibly the trend of particle growth diuing the polymerization reaction, so the operator can change production parameters based on these real-time results. 

The ensemble techniques of LD (Section 8.4.1.1) and DLS (Section 8.4.1.2) are capable of providing the required information in a relative short time, and therefore they are often used in both production environments and research facilities. Both the LD and DLS techniques typically yield enough information to determine whether the desired end-point was achieved, both during the reaction and after. However, DLS is often chosen over LD because the answers that it produces quickly and reliably - the intensity-weighted mean diameter and standard deviation (typically using the method of cumulants) - require no knowledge of particle parameters. By contrast, LD-based instruments require that both 

The CHDF technique (Section 8.4.2.2) is capable of providing very accurate size distribution information for dilute polymer emulsions containing more than one peak in the PSD. However, it cannot be used to analyze soft polymers, due to particle retention problems. This technique is used more frequently in research/QC labs than in production environments, because it requires significant operator expertise and experience. 

By comparison, the HOST centrifugation technique (Section 8.4.2.1) is much more operator friendly and therefore is attractive for both R D and production applications. It can measure relatively concentrated emulsions, which may be an important advantage in cases where an emulsion becomes coUoidaUy imstable upon dilution. A potentially significant disadvantage of this technique is that one measurement may require several hours of centrifugation, making it less useful in production situations where fast decisions are required. On the other hand, in R D environments it can prove to be very attractive, since multiple samples (e.g. 12) can be analyzed at the same time. 

To ensure a constant-flux layer one can simply move the measurement height closer to the surface. For the eddy correlation method, however, the response time of the instrument must be faster as the measurement height approaches the surface, because high-frequency turbulent eddies then contribute proportionally more to the concentration fluxes than at higher levels. On the other hand, fluxes measured very close to the surface may be less representative of those over the entire area for which the measurement is intended. For the gradient method, the requirement that z/zq 1 (based on the requirements of similarity 

A complicating factor in dry deposition measurements is the presence of sources of the depositing substance in the footprint of the measurement. Whereas the flux of SO2 is nearly always unidirectional (downward) and the surface is a sink for SO2, gases such as H2S, NH i, and NO, may have surface sources. It may be possible in some cases to specify a surface emission rate. For NO2, the situation appears to be even more complex than just adding a surface emission rate to the resistance model. As much as 50% of the NO2 initially removed at the surface can reappear as NO as a result of surface emissions (Meyers and Baldocchi, 1988). 

The assumptions of similarity weaken as one moves from flat and uniform surfaces into hilly terrain and associated natural surface covers (Doran et al., 1989). Fluxes are likely to change substantially over rather short distances (1 km or less), and it may be extremely difficult to establish a representative flux value for more extended regions from measurements at a single site. Some measurements have been carried out over sloping terrain, and methods have been developed to take into account the effects of such slopes in the determination of flux values (McMillen, 1988). However, there is presently no generally useful method for establishing flux values over rough terrain. 

The dimensionless concentration gradient that is a function of atmospheric stability, 

FIGURE 19.6 Relationship between characteristic times of vertical turbulent diffusion tj) and a number of atmospheric reactions tc) for a layer of thickness Az = 10 m and different thermal stability classes (Kramm et al., 1993). For example, the HNO3—NH3—NH4NO3 equilibrium (reaction 2) has a reaction timescale comparable to that of turbulent diffusion under unstable and neutral conditions. 

Chemical analysis M Requires an atom or group present in small proportions best for low molecular weights. 

Vapor-pressure lowering Applicable to molecular weights of up to 10.  

Diffusion For molecular weights up to 10. Gives minimal molecular weights based on spherical shape. 

Sedimentation For molecular weights up to 5 X lO . Result is independent of shape. 

Viscosity Best used in association with other methods, 

---

WR Pearson. Comparison of methods for searching protein sequence databases. Protein Sci 4 1145-1160, 1995. 

Marti, H. F. and R. A. Waller. An Exploratory Comparison of Methods for Combining Failure Rate Data from Different Data Sources. Report No. LA-7556-MS, Los Alamos Scientific Laboratory, 1987. 

Heywood, N. 1. and Cheng, D. C.-H. Trans Inst. Measurement and Control 6 (1984) 33. Comparison of methods for predicting head loss in turbulent pipe flow of non-Newtonian fluids. 

Table 5.22 Comparison of method performance for LC-ToF-MS and LC-MS-MS determination of Idoxifene...

Unsteady reaction data are often an excellent means for estimating physical parameters that would be difficult or impossible to elucidate from steady-state measurements. However, the associated problems in nonlinear optimization can be formidable. A recent review and comparison of methods is given by... 

A model problem. Comparison of methods. Further comparison of various iterative methods will be conducted by having recourse to the Dirichlet problem associated with Poisson s equation in the square 0 < < 1,... 

A further insight is that the best workflow depends on a combination of factors that can in many cases be expressed in closed mathematical form, allowing very rapid graphical feedback to users of what then becomes a visualization rather than a stochastic simulation tool. This particular approach is effective for simple binary comparisons of methods (e.g., use of in vitro alone vs. in silico as prefilter to in vitro). It can also be extended to evaluation of conditional sequencing for groups of compounds, using an extension of the sentinel approach [24]. 

Lam RL, Welch WJ. Comparison of methods based on diversity and similarity for molecule selection and the analysis of drug discovery data. Methods Mol Biol 2004 275 301-16. 

NUUTILA A M, KAMMioviRTA K and OKSMAN-CALDENTEY K-M (2002) Comparison of methods for the hydrolysis of flavonoids and phenolic acids from onion and spinach for HPLC-analysis, Food Chem, 76, 519-25. 

Waterhouse et al., A comparison of methods for quantifying oligomeric proanthocyanidins from grape. Am. J. Enol. Vitic., 51, 383, 2000. 

Uson, R., Laguna, A., Laguna, M., Uson, A. and Gimeno, M.C. (1988) Synthesis of pentahalophenylgold complexes of ylides a comparison of methods. Synthesis and Reactivity in Inorganic and Metal-Organic Chemistry, 18(1), 69-82. 

Kramer GN, Grobecker KH, Pauwels J (1993) Comparison of methods used for the preparation of biological CRMs. Fresenius J Anal Chem 352 125-130. 

Comparison of methods for calculating detection and quantification limits for analytical methods used for food... 

H. Rumpel, J. M. Pope 1993, (Chemical shift imaging in nudear magnetic resonance a comparison of methods), Cone. Magn. Reson. 5, 43. 

Traub SI, Johnson CE. Comparison of methods of estimating creatinine clearance in children. Am J Hosp Pharm 1980 37 195-201. 

W. H. Barr and S. Riegelman, Intestinal drug absorp- 28. tion and metabolism. I. Comparison of methods and models to study physiological factors in vitro and in... 

Reaction conditions reflux of a mixture of aromatic substrate, carboxylic acid, and Af-trifluoroacetic imidazolide in a molar ratio of 1 1.2 1.2 in trifluoroacetic acid. A comparison of methods A and B (see above) is presented in Table 14.7. 

WC Duncan-Hewitt, DJW Grant. True density and thermal expansivity of pharmaceutical solids—Comparison of methods of assessment of crystals. Int J Pharm 28 75-84, 1986. 

TITRATIONS FOR COMPARISON OF METHODS. The automated photometric and turbidimetric methods were compared using 30 cm3 samples of surfactant solution containing a nominal 20 mol SDBS to give an equivalence volume of 5 cm3. The effect of salinity on the titrations was studied using samples prepared containing sodium chloride concentrations of 0.0, 0.14, 0.70 and 1.46 wt%. The influence of the choice of filter (580 or 620 nm) was also investigated. 

COMPARISON OF METHODS. Both methods were fairly rapid with a typical analysis time of 5 minutes per sample as compared with 30 minutes per sample for the manual method ... 

A., Comparison of methods to calculate cyclosporine A bioavailability from consecutive oral and intravenous doses,/. Pharmacokinet. Biopharm. 

Knight, P. C., and Bridgwater, J., Comparison of Methods for Assessing Powder Attrition, Powder Tech., 44 99 (1985)... 

The Student s (W.S. Gossett) /-lest is useful for comparisons of the means and standard deviations of different analytical test methods. Descriptions of the theory and use of this statistic are readily available in standard statistical texts including those in the references [1-6]. Use of this test will indicate whether the differences between a set of measurement and the true (known) value for those measurements is statistically meaningful. For Table 36-1 a comparison of METHOD B test results for each of the locations is compared to the known spiked analyte value for each sample. This statistical test indicates that METHOD B results are lower than the known analyte values for Sample No. 5 (Lab 1 and Lab 2), and Sample No. 6 (Lab 1). METHOD B reported value is higher for Sample No. 6 (Lab 2). Average results for this test indicate that METHOD B may result in analytical values trending lower than actual values. 

For Table 36-2, a comparison of METHOD A results for each of the locations is made to the known spiked analyte value for each sample. This statistical test indicates that METHOD A results are lower than the known analyte values for Sample Nos. 4-6 for both Lab 1 and Lab 2. Average results for this test indicate that METHOD A is consistently lower than actual values. 

Table 36-1 Comparison of METHOD B test results to true value...

This set of articles presents the computational details and actual values for each of the statistical methods shown for collaborative tests. These methods include the use of precision and estimated accuracy comparisons, ANOVA tests, Student s t-testing, The Rank Test for Method Comparison, and the Efficient Comparison of Methods tests. From using these statistical tests the following conclusions can be derived ... 

Frank TS, Svoboda-Newman SM, Hsi ED. Comparison of methods for extracting DNA from formalin-fixed paraffin sections for nonisotopic PCR. Diagn. Mol. Pathol. 1996 5 220-224. 

A Comparison of Methods for Photospheric Abundance Determinations in K-Type Stars... 

---