ChromSword

As previously described, Eq. 6 contains two constants characteristic of the system and the sample, feo and S, which can be determined by two chromatographic mns differing only in tc. These two values allow to calculate log fe using Eq. 4. However, because there is no empirical solution, values of log few and S have to be computed by iteration. Such procedures are included in several commercially available LC software packages, such as Drylab (Rheodyne, CA, USA), Chromsword (Merck, Darmstadt, Germany), ACD/LC simulator (Advanced Chemical Development, Toronto, Canada) or Osiris (Datalys, Grenoble, Erance). This approach was comprehensively described and successfully applied for accurate log P determination of several solutes with diverse chemical structures [9, 12, 43, 50]. 

Optimization of the gradient separation could be expressed in the form of a mathematical algorithm. Usually from a couple of gradient runs of the same analyte mixture, this is sufficient to calculate empirical constants for the equation similar to equation (2-97). These algorithms are implemented in most of the optimization software, such as DryLab , ACDLabs , and ChromSword . 

Manual optimization of the gradient or the use of Drylab, Chromsword, ACD to determine the desired resolution/selectivity... 

Alternatively, the results from the gradient runs for each sample can be inputted into Drylab, ACD, or Chromsword for further optimization (see Sections 8.5.6.11). For the predicted experimental conditions (i.e., gradient slope, temperature, flow rate), if desired selectivity and resolution can be obtained, an experiment can be run for verification. The peak purity for the main analyte (MS and DAD detection) should be checked in the verification run. If the desired selectivity and/or the target analyte are not spectrally homogeneous, go to Step 6, Figure 8-37. 

In this chapter the third level of computer-assisted HPLC—the use of expert systems (like Drylab [1], AutoChrom [2], and ChromSword [3]) for effective method development— is discussed. 

There are two main approaches to the prediction of retention times based on chemical structures. Both use a training set of compounds to characterize the system prior to creation of a prediction expression. The hrst (used in ChromSword ) uses experimental retention times for a set of prescribed compounds to create an expression based on molar volume and energy of interaction with water [43] ... 

Disadvantages Error levels still remain a concern, particularly with gradient systems and ionizable compounds. Systematic studies have not been published to date, but average errors in k for gradient systems can approach 30%. Also, both ChromSword and LC Simulator require a reasonable training set of compounds in order to characterize a chromatographic method for a particular compound. 

Finally, structure-based predictive software is commercially available (such as CHROMDREAM, CHROMSWORD or ELUEX) for mobile phase optimisation in RPC. This software incorporates some features of the expert system, as it predicts the retention on the basis of the molecular structures of all sample components (which should be known) and the known behaviour of model compounds on various HPLC columns. No initial experimental runs are necessary as the retention data are calculated from the additive contributions of the individual structural elements to the retention, contained in the software databa.se and consequently optimum composition of the mobile phase is suggested. Such predictions are necessarily only approximate, do not take into account stereochemical and intramolecular interaction effects, and predicted separation conditions can be used rather as the recommendation for the initial experimental run in the subsequent optimisation procedure. 

Structure-based commercial optimization softwar (e.g., Chromdream, Chromsword, or Eluex) incorporate some features of the expert system, as the retention is predicted based on the additive contributions of thdS... 

ChromSword Auto Software incorporated into Waters AMDS. Useful tool for optimizing isocratic and gradient methods and for predicting optimum column configurations. Similar to DryLab and can also work 14,000... 

HiPac (53) from Phase Separation is another commercially available software package. In several aspects this software is similar to DryLab, but its most important feature is that it can estimate the optimum mobile-phase conditions for the separation of the mixture at hand or only a selected number of peaks. Recently, ChromSword (commercially available from Merck, Germany) was introduced (34). It uses a retention mc el based on solvophobic theory. The input for this package can be the structural formulas of the solutes, the combination of structural formulas and retention data from a single run, or retention data from two runs. Data from additional runs are incorporated into the model, and prediction accuracy below 3% can be achieved under these circumstances. 

ChromSword includes an optimization algorithm. It can, however, also be used as a prediction program ... 

Figure 3.24 ChromSword prediction for a last two peaks come cioser together, as in gradient from 45 to 80% methanol in 25 min the real measurement shown in Figure 3.22. based on the linear model. With one click The input data are those from Tabie 3.4. the quadratic function is selected and the...

Galushko, S. et ah, Chromsword Software for Automated and Computer-Assisted Development of HPLC Methods, in HPLC Made to... 

Structure-based commercial optimization software (e.g., Chromdream, Chromsword, or Eluex) incorporate... 

Automation can generally lead to the elimination of errors and to time saving. Meanwhile, fuUy automated computer-aided method development and semi-automated optimization in HPLC have reached a remarkable level of maturity and sophistication. Through several real examples, Lloyd R. Snyder (Chapter 4.1) and Sergey Galushko (Chapter 4.2) describe the possibilities offered by the software packages DryLab and ChromSword , respectively. Michael Pf er (Chapter 4.3) compares the two software concepts from the point of view of the user, and presents a new software tool that also incorporates automatic colurrm selection. 

A number of experimental optimization programs are available, including DryLab 2000 (www.laesources.com), ACD/LC Simulator 8.0 (www.acdlabs.com), and ChromSword (www.chromsword.com). 

Optimal isocratic and/or gradient conditions are predicted by computer optimization software by retention time and peak width modeling, e.g., with DryLab (LC Resources, BASi Northwest Laboratory Services, Walnut Creek, CA, USA), LC Simulator (Advanced Chemistry Development, Toronto, Canada), or ChromSword (VWR International, Darmstadt, Germany). 

Canada), ChromSmart MS (Intelligent Laboratory Solutions Inc., Naperville, IL, USA), and ChromSword Auto (VWR International, Darmstadt, Germany). It is envisaged that there will be an increase in the number and flexibility of these systems in the coming years. Inexperienced analysts may then operate the systems in a totally automated fashion starting from the input of the chemical structure of the analyte of interest. For more demanding samples, the more experienced analyst will have access to a fully flexible system, which would allow multiple points of entry and exit from the method development platform. 

ChromSword Software for Automated and Computer-Assisted Development of HPLC Methods... 

ChromSword supports both off-line interactive procedures for method optimization and on-line unattended method development. 

ChromSword for computer-assisted HPLC method development was developed between 1990 and 1995 as an extension of ChromDream HPLC method development software [1]. In 1999, the first version for automatic HPLC optimization was developed and launched by S. Galushko in collaboration with Merck KGaA (Darmstadt, Germany). As a result of cooperation with VWR International Scientific Instruments, Darmstadt, Germany, Hitachi High Technologies... 

America, San Jose, CA, USA, and Scientific Software Inc., Pleasanton, CA, USA, versions for Hitachi LaChrom and LaChromElite hardware and EZChromElite software were launched. Partnership with Agilent Technologies resulted in the creation of a powerful version that supports Agilent 1100 LC and LC/MS systems with six columns and twelve solvent selectors. Further systems supported by ChromSword Auto are Waters Alliance LC systems with Millenium and Empower softwares. ChromSword Auto also supports 2-8 column and 2-16 solvent switching valves for all HPLC systems described. 

ChromSword works with different retention models. The retention model is a type of experiment-based mathematical expression that describes the relationship between the retention of a compound and its properties, as well as the conditions appertaining to the chromatographic experiments. The determination of retention models that adequately describe the effect of chromatographic conditions on the retention of compounds in a sample is very much the focal point in method development software. In this case, on the basis of only a few experiments, the... 

ChromSword supports three approaches for the determinahon of retenhon models in reversed-phase HPLC ... 

ChromSword supports the optimization of separation for polynomial models up to a power of six. Thus, the most complex retention-concentration effects can be described and the separation optimized. AU polynomial models predict the retention of solutes rather precisely in the interpolation region of those concentrations studied. These models are less reliable in the extrapolation region. For example, if experiments were performed with 40% and 50% of organic solvent in a mobile phase, one can expect rather good prediction of retention and separation in the region between these concentrahons and less accuracy in the regions of 30-35% and 50-55%. Extrapolahon to wider hmits very often leads to substantial deviations between predicted and experimental data. 

ChromSword can work with substantial masses of data. One sample file can contain up to 100 compounds, including structural formulae and data for up to 20 runs in the each module. 

Both approaches are applied rather successfully in practice and can be automated, although automation of the intelligent approach is substantially more difficult. ChromSword supports both approaches and several types of problems can be solved automatically. Typically, these are ... 

Fig. 1. Forced stress test. Results of automatic optimization of the separation of an active pharmaceutical ingredient (API) impurities and degradation products are present at concentration levels of 0.01-0.7% relative to the API. Software ChromSword 3.1, ChemStation 10.2. Agilent 1100 HPLC system. Column Inertsil ODS-3, 5 pm, 15 cm x 4.6 mm.

---