HPLC/NMR at the nanoliter scale Chromatography coupled with spectroscopy is a powerful analytical tool for fast structural elucidation of complex mixtures. Of the possible combinations, GC/MS and HPLC/UV are now well established techniques, and HPLC/MS is well on its way to becoming a standard method. HPLC/NMR, however, is still in its infancy, although the instrumentation is commercially available and is being used to solve demanding analytical problems. Does it make sense to go another step and couple micro-HPLC (using packed capillaries) with NMR? For Klaus Albert and his co-workers at the Institute of Organic Chemistry at the University of Tubingen (Germany), the answer is definitely "Yes". With this approach they determined the structure of chemically labile kitols in a fully deuterated mobile phase {Anal. .hem. 1997,69, 1421-25). According to Albert, ,his waa the first successful application of micro-HPLC/ NMR to a real analytical problem that could not be solved by other means. As a detector, NMR is less mass sensitive than other spectroscopic techniques; typical sensitivities with a 600-MHz NMR instrument running a standard FT experiment are in the nanogram range. The structural information, however, is unique because it reveals how atoms are llnked within a molecule. Unfortunately, it is an expensive analysis, requiring deuterated solvents to elute the analytes. Thus an important stimulus for developing microHPLC is to lower the costs of using these high-priced solvents. The main challenge of HPLC/NMR lies in desisrnint? 3. detector cell says Albert. Since his first attempts Albert's group has continuously improved the setup but "there is still a lot to do " he
Still in development: schematic of the HPLC/NMR.
needs a full day to adjust the magnetic field. Therefore the flow cell is left untouched in the spectrometer as long as a series of experiments is running. Albert hopes that future improvements in the probe head design will make shimming easier. With access to five NMR instruments, including spectrometers for imaging and solid-state NMR, the Tubingen group is looking at other applications of NMR detection. A 400-MHz instrument has been coupled with gel permeation chromatography. In these experiments, dichloromethane is the typical eluent, sample concentrations are high, and flow rates are low. Therefore, a 400-MHz instrument is sufficiently sensitive for these studies. Packed-column HPLC has also been run on the 400-MHz instrument. Mobile phases used for the reversed-phase HPLC include CH3CN/D20, CH:jOH/ D 2 0, or 1H-acetone/D20. At present, the interaction of analytes with the C lg chains of the stationary phase are being studied; results of the HPLC/NMR experiments are correlating well with molecular modeling calculations. Albert's group has also combined SFC with NMR. The relaxation times are long (a consequence of the lower viscosity of supercritical fluids), and integration of the NMR signals is not possible. Albert hopes to overcome these problems by using stationary phases with immobilized free radicals to reduce the relaxation time and thereby enhance the signal, an idea proposed by Harry Dora of the Virginia Polytechnic Institute and State University. Of course C02 as the mobile phase gives no signal in 'H NMR! Albert's research also includes the use of HPLC coupled with UV, NMR, and MS to screen South American plants for compounds of pharmaceutical interest. "We are not running out of work," Albert says and smiles. Veronika R. Meyer
matches the coil length and creates a cell with a volume of 200 nL This design needs further improvement, observes Albert, because the present cell is surrounded by a 1-mm air gap, which lowers sensitivity. The chromatographic resolution is poor because of the relatively large volume of the detector cell and, to a lesser extent, the long transfer line. However, this drawback is not significant because the NMR structural information allows overlapping peaks to be clearly distinguished, says Albert. For a 2D NMR experiment, the sample can be kept within the detection cell by stopping the flow during the necessary interval. The kitol separations were run with a flow rate of only 3 uL/min, resulting in extremely low solvent consumption, which allowed Albert's group to use expensive, fully deuterated solvents, such as acetonitrile-d3. This arrangement has two advantages. First, there is no need to suppress any NMR signals coming from the eluentt In the case of the kitol separation, it was necessary to evaluate the full range of the spectrum and see all couplings. Second, although HPLC solvents are "UV pure", Albert finds that signals from impurities appear in the NMR spectrum For the kitol separations, the HPLC pump was placed approximately 2 m from along with the expected sola 600-MHz spectrometer. The HPLC capil- vent peaks. On the other lary—250 um in diameter, 15 cm long, and hand when NMR solvents are used as HPLC solvents packed with a 3-um C18 phase—was the impurity peaks are gone placed directly below and outside the from the NMR spectrum NMR magnet field. This placement allowed easy exchange of the microcolumn Because the flow cell is without the need to re-shim the magnetic not spun, shimming of the field. A 40-cm-long transfer rapillary with magnetic field needs to be a 50-um i.d. takes the eluate from the liqdone with great care. Lii uid chromatograph to the detection cell Hong Tseng, who is working located between the Helmholtz coils. on her master's degree, has what Albert calls "a golden The detection flow cell is fabricated from a glass capillary with a 180-um i.d. and hand" when it comes to mag- Setting up the system, (left to right) net shimming, but even she measures 8 to 9 mm in length, which G. Schloatterbeck.
Tseng, Albert, and
Analytical Chemistry News & Features, August 1, 1997 4 5 9 A
