The Moving-Belt Interface

The operation of the interface may be divided conveniently into four stages, as follows  

A uniform deposit of analyte(s) on the belt is required and it is possible to do this with a range of mobile phases and flow rates by a very careful balancing of the rate of solvent deposition, the speed at which the belt moves and the amount of heat supplied by the infrared evaporator. 

If the belt moves too quickly, in relation to the rate of deposition, sample will not be deposited on all parts of the belt. This results in the production of an uneven total-ion-current (TIC) trace and a distortion of the mass spectra obtained, with consequent problems in interpretation, particularly if library searching is employed. 

An erratic TIC trace is also obtained if the belt is moving too slowly but in these circumstances this is due to the formation of droplets rather than the spreading of mobile phase on the belt. An additional problem encountered when droplets are formed on the belt is that more heat is required to evaporate the solvent and with this comes the increased likelihood of decomposition of any thermally labile compounds that may be present. 

When optimum experimental conditions have been obtained, all of the mobile phase is removed before the analyte(s) are introduced into the mass spectrometer for ionization. As a consequence, with certain limitations, it is possible to choose the ionization method to be used to provide the analytical information required. This is in contrast to the other LC-MS interfaces which are confined to particular forms of ionization because of the way in which they work. The moving belt can therefore provide both electron and chemical ionization spectra, yielding both structural and molecular weight information. 

---

The first interface to be made available commercially was the moving-belt interface, shown schematically in Figure 4.1. 

Only around 10% of the reported uses of the moving-belt interface involved the use of FAB the vast majority, some 90%, have utilized El or Cl [2]. 

Reference has been made to the problems associated with the presence of highly involatile analytes. Many buffers used in HPLC are inorganic and thus involatile and these tend to compromise the use of the interface, in particular with respect to snagging of the belt in the tunnel seals. The problem of inorganic buffers is not one confined to the moving-belt interface and, unless post-column extraction is to be used, those developing HPLC methods for use with mass spectrometry are advised to utilize relatively volatile buffers, such as ammonium acetate, if at all possible. 

The direct-liquid-introduction (DLI) interface was made available commercially just after the moving-belt interface to which, as no company produced both types, it was an alternative. At this time, therefore, the commercial LC-MS interface used within a laboratory was dictated by the manufacturer of the mass spectrometer already in use unless a new instrument was being purchased solely for LC-MS applications. The development of LC-MS in the early 1980s was such that this was very rare and it was therefore unusual that a scientific evaluation was carried out to assess the ability of a type of interface to solve problems within a particular laboratory. 

From a practical point of view, the DLI, unlike the moving-belt interface, contains no moving parts and is therefore more reliable in operation if adequate precautions are taken to minimize the frequency of the pinhole blocking. In addition, it does not require heat either to remove the mobile phase or to vaporize the analyte into the source of the mass spectrometer. The DLI is, consequently, better for the analysis of thermally labile materials. 

The advantages and disadvantages of this type of interface, particnlarly in comparison to the moving-belt interface which was available at the same time, are listed below. This was one of the first LC-MS interfaces to be made commercially available and, although used in a number of laboratories, its development was halted premamrely by the introduction of the thermospray interface (as we shall see later). 

No heat is applied to the interface and it is therefore able to deal with thermally labile materials better than the moving-belt interface. 

The interface contains no moving parts and is cheap and simple to construct and operate and is inherently more rehable than the moving-belt interface. 

The full potential of LC-MS could not be exploited until it was possible to study involatile and thermally labile compounds for which electron and chemical ionization are not appropriate. A relatively small number of reports of the use of the moving-belt interface with fast-atom bombardment ionization for the study of these types of compound have appeared. 

It has been previously noted (see Section 4.2 above) that use of the moving-belt interface allows El spectra to be obtained from compounds that do not yield spectra when analysis is attempted using a conventional El probe. The same is true when the dynamic-EAB probe is used in that spectra may be obtained from compounds that do not yield spectra when a static-FAB probe is used. This has been attributed to the presence of the mobile phase. 

Arguably the ultimate LC-MS interface would be one that provides El spectra, i.e. a spectrum from which structural information can be extracted by using famihar methodology, and this was one of the great advantages of the moving-belt interface. There is, however, an incompatibility between the types of compound separated by HPLC and the way in which electron ionization is achieved and therefore such an interface has restricted capability, as previously discussed with respect to the moving-belt interface (see Section 4.2 above). 

The range of compounds from which electron ionization spectra may be obtained using the particle-beam interface is, like the moving-belt interface, extended when compared to using more conventional methods of introduction, e.g. the solids probe, or via a GC. It is therefore not unusual for specffa obtained using this type of interface not to be found in commercial libraries of mass spectra. 

Spray deposition A method used to apply HPLC eluate in later versions of the moving-belt interface to provide a uniform layer of mobile phase on the belt and thus minimize the production of droplets. 

Various transport type interfaces, such as SFC-MB-MS and SFC-PB-MS, have been developed. The particle-beam interface eliminates most of the mobile phase using a two-stage momentum separator with the moving-belt interface, the column effluent is deposited on a belt, which is heated to evaporate the mobile phase. These interfaces allow the chromatograph and the mass spectrometer to operate independently. By depositing the analyte on a belt, the flow-rate and composition of the mobile phase can be altered without regard to a deterioration in the system s performance within practical limits. Both El and Cl spectra can be obtained. Moving-belt SFE-SFC-MS" has been described. 

Perhaps the most mechanically complex solution ever developed for uniting HPLC with mass spectrometry was the moving belt interface [54]. The heart of this system was a mechanically driven continuous belt (analogous to an escalator or moving walkway) to which the HPLC eluent was applied. The majority of the mobile phase was evaporated by a heat source (ideally hot enough to vaporize the solvents but not to... 

Coupling of liquid chromatography to mass spectrometry has not only led to a wide variety of interfaces, but also initiated the development of new ionization methods. [8-13,62] In retrospect, the moving belt interface seems rather a curiosity than a LC-interface. The LC effluent is deposited onto a metal wire or belt which is heated thereafter to desolvate the sample. Then, the belt traverses a region of... 

The moving wire interface was developed by Scott et al., and the moving belt interface by McFadden et al. ° This was the first commercial interface for LC/MS, introduced in 1977. In both of the techniques, the eluent is deposited onto a stainless-steel wire, or a plate usually made of polyimide (known as Kapton), followed by the removal of the solvent in vacuum. The residual solid analyte is vaporized into an ionization... 

The moving wire device has a number of major shortcomings. Due to the small surface area of the stainless-steel wire, such as available from a 0.1 mm diameter wire, the device can only accommodate about 10 pL/min eluent which results in poor sensitivity. The system is difficult to operate in a continuous mode. Modification of the moving wire approach has led to the invention of a continuous moving belt, which offers improved transfer efficiency and therefore higher sensitivity. The moving belt interface is capable of handling up to 1 mL/min of mobile phase. 

In the moving belt interface, effluent from the column is brought into contact with the belt, and the film of effluent is carried under an infra-red lamp evaporator which rapidly removes most of the solvent. The solute and residual solvent then pass through two vacuum locks, where remaining solvent is removed, and into the ion source of the mass spectrometer where the sample is flash vaporised. The belt then travels over a clean-up heater which removes any residual solute. 

El is performed in a high-vacuum ion source (typically < 10 Pa) intermolecular collisions are avoided in this way. As a result, El mass spectra are highly reproducible. Extensive collections of standardized El mass spectra are available [3-4], also for on-line computer evaluation. An important limitation of El is the necessity to present the analyte as a vaponr, which excludes the use of El in the study of nonvolatile and thermally labile compounds. El is widely applied in GC-MS [5]. In LC-MS, its applicability is hmited to the particle-beam interface and the moving-belt interface. 

A modified Pye Unicam moving-wire detector was described by Scott et al. [35] in 1974 to fit the vacuum requirements of a mass spectrometer (Figure 3.3). Part of the colunrn effluent is deposited on to a wire, which transports the liquid along a heating element to evaporate the solvents, and through a series of vacuum locks to the ion source where the analyte is thermally desorbed from the wire prior to the ionization. Ionization is independent of the LC system. Therefore, conventional El and Cl spectra can be obtained [35]. This approach was subsequently adapted in 1976 by MacFadden [36] into the moving-belt interface (Ch.4.4). 

The first commercial LC-MS interface, available in 1977, was the moving-belt interface, which was a modification by MacFadden et al. [36] of the moving-wire system described by Scott et al. [35], The moving-belt interface, discussed in Ch. 4.4, was capable of introducing up to 1 ml/min of mobile phase and achieving solvent-independent analyte ionization by El or CL A similar system was described by Millington et al. [72]. 

---