The Pros and Cons of Direct Injection

I’m a big advocate of proper sample preparation – after all, it’s an essential step in the analytical process. Typically, this important process enriches target analytes and enables sample clean-up, which eliminates matrix effects. You cannot correct errors that occur in this step even with the most sophisticated analysis. However, over the last few years there has been tremendous progress in mass spectrometry (MS) instrumentation combined with liquid chromatography (LC) that has, to some extent, rendered sample enrichment obsolete, when no significant signal suppression by matrix components occurs. In fact, this is often the case with very clean matrices, such as aqueous samples.

Successful determination of organic compounds down to the low ng/L range has been described, often in combination with injecting large water volumes (above 10 percent of the void volume of the analytical column), which focuses the target analytes on the stationary phase, allowing them to elute only after initiating the elution gradient. The approach clearly requires some retardation of target analytes under the initial eluent conditions, otherwise, we face two problems:

• Insufficient focusing will lead to broad peaks. For instance, in the past decade, the focus of water analysis beyond regular monitoring of priority pollutants has shifted to more polar contaminants and transformation products, so achieving sufficient retardation may be critical.
• Co-elution of most of the interfering polar matrix may suppress or unpredictably enhance the signal. Even in clean water samples, organic and inorganic matrix components are present in concentrations typically orders of magnitude higher than the target analytes.

So, what are the possible solutions? Stationary phases that enable purely aqueous initial conditions with an immediate solvent gradient can, in many cases, lead to satisfactory peak width and symmetry, as is seen for multiple pesticide metabolites. If on-column focusing is insufficient, consider using a very small pre-column filled with highly retentive material, such as porous graphitic carbon. And adding isotopically labeled internal standards before injection is another well-established approach that makes it possible to correct matrix effects on signal intensity (provided they are identical for both target and internal standard). For more complex matrices containing a lot of organic matter, it is best to check the sample by recording matrix effect profiles using constant post-column infusion of the target and the IS.

There are some drawbacks to the above approach, however. For example, for analyzing emerging contaminants and in particular transformation products, the corresponding isotopically-labeled standards may not be available commercially. Also, for multicomponent methods, many standards are cost prohibitive. Finally, low sensitivity caused by matrix suppression may hamper detection in the required concentration range – though it is possible to overcome this problem with a post-column additive, which is a well-established approach in other application areas of LC-MS/MS but rarely used in water analysis. Post-column addition of ammonia solution in the µmol range may, for example, enhance signal intensity for many compounds bearing an amide or amino group and at the same time compensate for differences in response due to the natural matrix.

If, however, the use of internal standards isn’t possible, standard addition is the most appropriate choice. Unfortunately, scientists often avoid this approach because it requires time-consuming multiple analyses of each sample. However, in water analysis, where in many samples target analytes are not detected above the requested reporting limit, it could be less problematic if you use short columns to reduce analysis time and if you automate the entire process. Also, chromatographic software can decide after an initial run of the sample whether quantification using standard addition is necessary at all. I’d like to ask vendors to continue software development with this in mind.