Do We Really Need Chromatography?
The Role of Chromatography and Different Detection Techniques in the Context of Mass Spectrometry
- Fig. 1: UHPLC-TOF chromatogram of a pesticide mixture. Every analyte which has been automatically detected by the software is highlighted by a peak marker (in grey). The black curve represents the total ion chromatogram. The other colours represent the extracted ion chromatograms.
- Fig. 2: Multiple Reaction Monitoring chromatogram of 50 compounds on a Q Trap 3200 system using Eksigent ExpressLC-Ultra micro liquid chromatography and a nanobore monolithic column (Merck Nano Chromolith RP18 150 x 0.1 mm). Gradient: from 1 % to 99 % acetonitrile within 15 minutes. Flow rate: 5 µl/min.
- Fig. 3: Overlay of 11 gradients from 100 consecutive gradient runs for calculation of the gradient dwell volume. Every tenth gradient has been overlaid.
- Fig. 4: Comparison between different dwell volumes for conventional UHPLC (50 µl) and micro-LC (1 µl).
- Fig. 5: Online SPE-LC-MS/MS chromatogram for the determination of atenolol, metoprolol, bisoprolol and carbamazepine using Spark Holland Symbiosis Pico and AB Sciex Q Trap 3200. (c = 1 ng/L, enrichment volume: 10 mL, SPE cartridge: Spark HySphere C18HD, HPLC column: Phenomenex 50 x 2.1 mm Kinetex 2.6 µm C18 100 Å).
- Fig. 6: Schematic illustration of an online Raman detector on the basis of a liquid core waveguide.
Although the capabilities of modern mass spectrometers steadily increase, so does the requirement to measure plenty of analytes in complex matrices. In this respect, the question can be regarded as mere rhetoric, because matrix effects can lead to strong ion suppression rendering the analysis useless .
The Multi-method Paradox
While some years ago the first LC-MS multi methods contained not more than ten compounds, today's methods often cover more than 100 analytes which have to be quantified in a single run . The question arises about the role of chromatography in multi-target methods? This is highlighted by the section of a chromatogram depicted in figure 1. Here, only a small elution window is shown. The chromatogram was recorded by using UHPLC-TOF hyphenation. As can be seen, more than 40 compounds have been detected with an average peak width of about 10 seconds. It is simply not possible to chromatographically resolve all analytes in the mixture by changing the selectivity of the phase system. In this case, only multidimensional chromatographic techniques might be of use because of their theoretically higher peak capacity, e. g. heart cut LC-LC or comprehensive LCxLC .
Time-of-flight versus Triple Quadrupole Instruments
While TOF analysers are often used for screening experiments or non-target analyses, triple quadrupole instruments can be considered the workhorses for quantification in the lower ng/l range in target analyses . When applying triple quadrupole MS detection, the multiple reaction monitoring (MRM) mode is usually used and at least two specific mass transitions are monitored per analyte. If a method includes 50 analytes, this means that 100 MRM transitions have to be monitored. Today's instrumentation offers the possibility of scheduling the measurement of single mass transitions according to the elution time of the compounds. If this option is not available, all mass transitions have to be monitored during the complete chromatographic run. The limitation of this approach is shown in figure 2. Here, 50 analytes have been measured simultaneously on a triple quadrupole instrument using only one mass transition per analyte.
While at higher concentrations, the peak width is about 10 seconds, it decreases to a few seconds for low concentrations, so that only a few data points per peak are obtained. By adjusting the dwell time of the mass spectrometer to the lowest technical limit, only the noise is increased (data not shown). Therefore, only systems which allow for a scheduled monitoring of MRM transitions and which have a very fast duty time can be used for analysing a huge number of analytes in a single run. An alternative approach is to prolong the gradient time in order to decrease the number of peaks eluting per unit time to acquire at least 10 data points per peak. However, this will have a negative impact on the analysis time and decrease the sample throughput.
Using Monolithic Columns and Micro-LC as Front End Instrument
For the application shown in figure 2, a micro-LC system has been used as the front end instrument with a nanobore monolithic column. The stationary phase plays a major role when accelerating the separation. Fully porous sub 2 µm particles are widely used for many years and have the advantage of a nearly horizontal slope of the C-term of the van-Deemter curve. Therefore, the flow rate can be increased until the maximum pressure of the system or the column will be reached. In some cases, very high pressures are already obtained at low flow rates if methanol is used as the organic modifier because of its substantially higher viscosity when compared to acetonitrile. Hence, core-shell stationary phases have in many cases replaced fully porous sub 2 µm particles. The lower diffusion path leads to comparable separation performance without the negative impact of high pressures. A third alternative is the use of monolithic columns. By their unique pore structure, extremely low pressures will be observed so that very high linear flow rates can be adjusted. Since the column inner diameter also exerts an influence on the backpressure, monolithic columns are very favorable when nanobore columns with an ID of 100 µm are used.
Reducing the ID of the column means that specially designed HPLC systems with an extraordinarily low extra-column volume must be used. In this respect, capillary- or micro liquid chromatographic systems are the first choice for these kinds of columns. Micro liquid chromatography is of renewed interest because it has many advantages over conventional HPLC. First of all, solvent consumption is brought to a minimum when the pumps deliver the mobile phase in the true µl/min range without flow splitting. Furthermore, the injection of 50 nl and less is technically feasible if there is only a limited amount of sample material. Unfortunately, there are long-standing notions against micro-LC which are extensively discussed and have prevented it from making a real breakthrough in many fields of application. Problems are regularly observed in terms of the retention time reproducibility because many systems are not capable of delivering a constant flow. In many cases, a flow splitter is used to reduce the flow rate. Moreover, solvent gradient reproducibility is considered a major problem because the flow and thus, the solvent mixing cannot be controlled precisely.
However, modern instruments deliver the mobile phase without flow splitting and the extra-column volume is tremendously reduced. These facts are highlighted by the overlay of 100 consecutive gradient runs in figure 3, where every tenth gradient has been overlaid. The relative standard deviation between these gradients was calculated to be 0.6 %, which can be regarded excellent.
The next issue concerns the gradient delay volume, which has been experimentally determined to be 0.93 µl. Figure 4 shows the gradient delay when a true micro liquid chromatographic system is used (black line) and compared to the most advanced conventional UHPLC-system with a dwell volume of about 50 µl (red line). Although most pumps of conventional UHPLC systems are also capable of delivering low flow rates, it is impossible to obtain very fast cycle times in the low µl/min range. The reason is that it takes 10 minutes(!) until the gradient reaches the head of the column when the flow rate is adjusted to 5 µl/min. In contrast, the gradient delay is only 11.2 seconds for the micro-LC-system.
Future Challenges of LC-MS in Environmental Analysis
The future challenges of LC-MS in environmental analyses will certainly include lowering the limit of quantification for some priority pollutants. In 2012 the European Commission extended the European water framework directive and proposed environmental quality standards (EQS) in the context of the water framework directive. Besides different emerging contaminants, also two estrogenic substances, 17a-ethynylestradiole (EE2) and 17b-estradiole (E2), were added to this list. Furthermore, environmental quality standards for the annual average in surface waters were proposed. According to the low activation threshold for hormones the values were set to 0.035 ng/l for EE2 and 0.4 ng/l for E2. This means that effective sample enrichment has to be used in order to fulfil this requirement. Here, systems including online large volume injection of a few millilitres with liquid chromatographic separation and mass spectrometric detection might be the key to success. Figure 5 shows a chromatogram of some priority pollutants by online large volume injection on an SPE cartridge. It is possible to measure a concentration of 1 ng/l directly without any additional offline SPE procedures. It must be admitted, however, that this was the injection of a pure standard without any matrix. For a real application of this method, the washing procedures have to be adapted to obtain a very clean extract which can then be transferred by the solvent gradient from the SPE cartridge to the HPLC column. Using additional offline SPE procedures might even enhance the limit of detection in case of ultra-low concentrations and when ionization efficiency of the target compounds is very poor.
New Hyphenation Strategies
This article closes with a perspective of new chromatographic and hyphenation strategies in combination with mass spectrometry. During the last 30 years a tremendous instrumental progress in liquid chromatography as well as mass spectrometry has been achieved. This means that alternative detection methods cannot catch up with this development in a few years. Nevertheless, there are interesting and promising technologies which one day may also be incorporated to a "hypernated" system, yielding additional information which cannot be extracted by mass spectrometry alone.
One of these techniques is the online Raman spectroscopy. It seems weird to introduce Raman spectroscopy when talking about LC-MS hyphenation, because Raman spectroscopy is usually considered as an analytical tool which is unable to detect traces. In the last years, however, the development of very sensitive detection devices, in particular CCD-cameras in combination with new sample cell designs, lead to a strong expansion of detection limits. An example shows figure 6. In this online Raman setup a liquid core waveguide adopts the role of the sample cell. This device was developed in order to gather spectral information for the detection of steroids and other pharmaceutical substances in combination with high-temperature liquid chromatography and isotope ratio mass spectrometry . Although the detection limit for a range of substances is around 500 µg/l, further optimization will be carried out by considering Surface Enhanced Raman Scattering (SERS) for enhancing Raman signal intensities. If this project is successful, a cost effective alternative for many laboratories would be available for confirmation of target compounds which cannot be distinguished by their UV spectra. This might also be the case when MS detection fails, as in the case of isobaric interferences. Referring to the above mentioned steroids, ionization efficiency in mass spectrometry is very low and hence, Raman detection might be a real alternative in a few years for delivering structural information of these compounds in combination with UV and MS detection.
The mobile phase which has been used for online Raman detection consists of pure water, because the Raman signals of water less interfere with those of analytes. In order to increase the solvent strength of the mobile phase, high-temperature liquid chromatography has been applied. Meanwhile, this technology is well characterized and also fully compatible to MS detection . In the current project, new chromatographic strategies like a combined use of solvent and temperature gradient programming will be explored, since it is yet not clear to which effect the Raman scattering will be influenced by organic solvents.
Liquid chromatography hyphenated to mass spectrometry will clearly dominate other hyphenation methods when it comes to micropollutant identification and quantification. An interesting approach, which can be routinely used, is the implementation of micro-liquid chromatography and the use of monolithic stationary phases. In order to determine very low concentrations of new emerging contaminants, a combination of offline and online solid phase extraction procedures need to be included in the LC-MS workflow. Because of the high technical advance, alternative detection techniques like Raman detection will not be able to catch up with LC-MS in the next years, although its incorporation into a "hypernated" system might be a valuable tool in order to extract as much spectral information as possible from complex samples.
The IGF project number 17497 N is funded via the Industrial Research Associations Working Group (Arbeitsgemeinschaft industrieller Forschungsvereinigungen, AiF) as part of the German Federal Ministry of Economics and Technology's Industrial Joint Research and Development (Industrielle Gemeinschaftsforschung, IGF) programme, as established by a resolution of the German Bundestag.
Furthermore, we would like to thank Norbert Wenkel and Dr. Andreas Bruchmann from Axel Semrau, for the loan of the Spark Holland Symbiosis Pico system (online SPE-HPLC). We would also like to thank Dr. Karin Cabrera and Dr. Stephan Altmaier from Merck for the loan of monolithic columns.
 Stahnke H. et al.: Anal. Chem. 81, 2185 (2009)
 Vishwanath V. et al.: Anal. Bioanal. Chem 395, 1355 (2009)
 Eeltink S. et al.: J. Chromatogr. A 1216, 7368 (2009)
 Zedda M. and Zwiener C.: Anal. Bioanal. Chem. 403, 2493-2502 (2012)
 Jochmann M.A. et al.: GIT Labor-Fachzeitschrift 54, 182-185 (2010)
 Teutenberg T.: High-Temperature Liquid Chromatography - A User's Guide for Method Development, Royal Society of Chemistry, Cambridge, 2010
Dr. Thorsten Teutenberg, Head of Department of Research Analysis
Juri Leonhardt, PhD student, Department of Research Analysis
Terence Hetzel, PhD student, Department of Research Analysis
Sandy-Dominic Freihoff, Lab manager, Department of Research Analysis
Dr. Jochen Tuerk, Head of Department of Environmental Hygiene & Micropollutants
Institut für Energie- und Umwelttechnik e. V.
Prof. Dr. Hans Bettermann,
Head of the working group for liquid phase laser spectroscopy, Institute of Physical Chemistry
Bjoern Fischer, PhD student, Institute of Physical Chemistry
Dr. Thorsten Teutenberg
Institut für Energie- und Umwelttechnik e. V.
Tel.: +49 2065 418179
Fax: +49 2065 418211