Skip to Main Content
Skip Nav Destination

The development of miniaturized analytical or chemical processing systems for both biological and chemical applications is a fast growing field because such systems enable the performance of a series of successive operations at scales which are not easily handled by human experimenters. A key challenge arising from this continuous system miniaturization towards the micrometer scale, or even smaller, lies in the ability to sensitively detect low molecular concentrations in reduced sample volumes. Additionally, such analytical systems must be coupled to microfluidic devices with minimal loss of analytes and information. The last issue is the scalability of the detection technique, as the detection is performed on small sample sizes. The ideal technique for microfluidic detection would therefore present an enhanced sensitivity upon downscaling. The dream of the users in the (bio)chemical field would be a fully integrated and portable device that includes (micro)systems for sample handling, preparation and detection. Conventional detection systems are still bulky instruments, resulting in the paradox of coupling a smaller and smaller analytical or processing device to room-sized instrumentation for the detection.

On-chip detection firstly relied on optical techniques, such as ultraviolet (UV) absorbance, fluorescence or laser-induced fluorescence (LIF).1,2  The latter technique in particular has a sensitivity in the (sub)micromolar range which is suitable for microfluidic applications. Besides optical techniques, electrical-based techniques are also widely used for on-chip detection due to their sensitivity, e.g. detection based on conductivity,3  electrochemistry,4  electrochemiluminescence,5 etc. The main advantage of these techniques is that they are fully integrated on the microdevice via the introduction of electrodes; they do not rely on the use of complex and bulky instrumentation as is the case for optical techniques. More exotic techniques are also used in combination with microfluidics, such as nuclear magnetic resonance (NMR)6  and Raman spectroscopy.7  These techniques are less popular but are currently developing at a rapid rate. Since the late 1990s mass spectrometry (MS) has also been used for the detection stage for microfluidic processing systems;8  this combination is particularly striking if one considers the size of a mass spectrometer compared to that of a microchip! MS has rapidly replaced other techniques due to its very high sensitivity and other advantages such as a high selectivity compared to optical-based techniques, for instance. Consequently, it turned out that MS analysis could also benefit from the use of microfluidic systems for sample preparation prior to analysis. As a consequence, the field of microfluidics and MS has been rapidly growing with the appearance of dedicated products within the last decade.

In the first part of this introductory chapter, we briefly introduce the technique of mass spectrometry as well as two ionization methods, namely ESI (electrospray ionization) and MALDI (matrix-assisted laser desorption ionization), commonly used for the analysis of biological/biochemical samples or for organic chemistry purposes. The second part highlights the advantages brought by the miniaturization and coupling of microfabricated devices to MS, and how this marriage benefits both on-chip detection and the MS analysis. The third part of this chapter focuses on the different approaches adopted for coupling microfabricated systems to ESI-MS or MALDI-MS and on the miniaturization of the mass spectrometer itself. In the final part different fields of applications of miniaturization for MS analysis are presented. Moreover, the different technological developments and applications that are treated in greater detail in separate chapters in this book about miniaturization and mass spectrometry are reviewed.

Mass spectrometry is an analysis technique that detects substances as a function of their molecular weight, or, more precisely, that detects substances as ions as a function of their mass-to-charge ratio (m/z). The analysis starts with the ionization of the molecules, which are subsequently separated in an analyzer according to their size (m/z ratio) before they reach the detector. A mass spectrum is composed of a series of peaks at given m/z values, indicating the presence of ionic species characterized by these mass-to-charge ratio values.

The key part of the connection between microfabricated/microfluidic devices and a mass spectrometer is the ionization of the analyte, as molecules are introduced as ions for the analysis. Subsequently, they must be ionized on the chip or at the outlet of the chip to be detected. Ionization is achieved using many different techniques, depending on the molecule properties. However, we will focus here on two ionization techniques, ESI and MALDI. These two techniques prevail nowadays in the field of MS analysis and they consist of the two main ionization methods used in combination with microfluidic analysis, although two other techniques, atmospheric pressure chemical ionization (APCI) and atmospheric pressure photoionization (APPI), have also recently been reported for on-chip detection.9,10 

MALDI and ESI ionization techniques are known as soft ionization techniques: they are suitable for the ionization and the analysis of large molecules (MW > 1 kDa) such as polymers, proteins, peptides, nucleic acids and poly- or oligosaccharides without fragmenting them. During the ionization process, the molecules acquire enough energy to be transferred to the gas phase, but the amount of energy remains low enough not to induce any fragmentation of rearrangement of the molecules. MALDI and ESI are the two mostly used ionization techniques because of their compatibility with the analysis of large molecules, and especially biomolecules, and their routine sensitivity in the low picomole down to the femtomole range. This popularity of ESI and MALDI has mainly been caused by the explosion of analytical needs in the fields of biology and biochemistry. The discovery of these two soft ionization techniques in the mid-1980s is strongly linked to the prosperity of MS, and the recent growth of proteomics analysis marks in particular the golden age of the techniques of ESI-MS and MALDI-MS. As a consequence, the discovery of both ESI and MALDI was rewarded in 2002 with the Nobel prize to their inventors, Fenn and Tanaka, respectively.

More recent technological developments in the field of MS have strengthened this prosperity and the potential of MS as an analysis tool for complex samples, “real-world” biological samples as well as complex synthetic mixtures. For instance, coupling MS to a separation technique such as capillary electrophoresis (CE) or liquid chromatography (LC) enables one to reduce the sample complexity and to successively analyze the species present therein. Also, tandem (MS2), or even MSn, helps to elucidate the structure of substances and the sequence of biopolymers such as peptides.

These points, the potential of MS and the recent developments in the field of MS account for the increasing popularity of (ESI- and MALDI)-MS and explains why MS has naturally been associated with the recent miniaturization trend and the recent appearance of microfabricated and microfluidic systems.

Electrospray ionization was first reported by Fenn in 1984, and further developed in 1988.11  This technique relies on the generation of a spray from a liquid upon application of a high voltage. Typically, the sample to be analyzed is introduced in a liquid phase in a capillary. A strong electric field is created (high voltage of several kilovolts) between the liquid and a counter electrode (i.e. inlet of the mass spectrometer) placed some centimeters in front of the capillary. Upon application of the electric field, the liquid breaks into a gas of highly charged droplets whose size depends on different parameters such as the capillary inner diameter, the flow rate of the liquid and the applied voltage. During transport, the droplets evolve to the ultimate stage of ions in the gas phase. The solvent (typically, an organic solvent such as methanol, ethanol or acetonitrile) evaporates, and the droplets thereby shrink until they reach the so-called Rayleigh limit where surface tension exactly compensates electrical forces, as expressed by

Equation 1.1

where q is the droplet charge, r its radius, γ the surface tension and ε0 the permittivity of vacuum. Beyond this limit, the equilibrium is lost and droplets break into smaller droplets that undergo the same process. This cycle proceeds until the stage of ions in the gas phase is reached, and these ions in the gas phase enter the mass spectrometer to be analyzed.

The technique of ESI gives rise to multi-charged species as long as there are several protonation sites on the analytes. A typical ESI mass spectrum presents a group of peaks for one analyte species, corresponding to the different charge states with a Gaussian distribution. One advantage of this is that the analyzer and detector of the mass spectrometer work with a reduced mass range as ions are analyzed and detected as a function of m/z and not of their molecular weight. Mass spectra are of course more difficult to interpret compared to MALDI mass spectra, although many deconvolution techniques are now routinely available to transform raw data into reconstructed spectra presenting peaks for mono-charged species. A limitation of the technique of ESI, especially compared to MALDI, is the low analysis throughput and its tedious and cumbersome preparation with the manual introduction of the sample in the capillary source.

There are now two commonly defined regimes for electrospray analysis. These regimes are distinguished by the inner diameter of the capillary source, the liquid flow rate and the applied ionization voltage, and as already mentioned above these three parameters dictate the size of the generated droplets. These two regimes are

  • classical ESI with a capillary inner diameter of ca 100 μm, a flow rate of 1–20 μL min−1 and ionization voltage of 3–4 kV; and

  • nanoESI with a capillary inner diameter of ca 1–10 μm, a flow rate below 1 μL min−1 and ionization voltage of 1 kV.

The miniaturization of the technique started in 1994 with the description of microESI by the group of Caprioli.12,13  A further step towards miniaturization was reported by Wilm and Mann in the 1990s with the development of nanoESI.14  The miniaturization of the technique is driven by the improvement it offers. ESI is sensitive only to the concentration of species and not to the total amount of analytes; as a consequence, its miniaturization does not affect the sensitivity of the analysis but rather enhances it, as explained below.

Working with smaller sources appears to be an advantage as it leads to the formation of smaller droplets, and this in turn gives a number of improvements. The increased surface-to-volume ratio of the droplets brings two main advantages: it favors evaporation phenomena allowing for using a higher relative amount of water in solution; and it yields an increased surface charge density, which promotes coulombic fission of the droplets. The evaporation–fission cycle becomes more efficient and gives more ions in the gas phase, and this leads all together to an increase of the ionization yield, from 10−9 for classical ESI up to 10−4 for nanoESI. This is illustrated by Equation (1.2) that gives the charge concentration as a function of the droplet size (radius): the smaller the droplet, the higher the charge concentration, and the higher the ionization probability for a molecule in the droplet:

Equation 1.2

where q/V represents the charge volume concentration in a droplet, r the radius of a droplet and γ the surface tension. Moreover, this increased charge concentration limits ion suppression phenomena; the competition between molecules to acquire a charge is lower, and as droplets are smaller there are also obviously fewer analytes per droplet. One particular consequence is a higher tolerance of the analysis to the presence of salts or other solution contamination; this is of great interest for the technique of ESI where the presence of salts usually fully hinders the process of ionization.

Additionally, miniaturization of the sources also leads to a decrease of the sample flow rate and the use of a lower ionization voltage. Ionization conditions are smoother, and the ionization source can be placed closer to the mass spectrometer inlet. Consequently, not only more ions are formed, but also more ions enter the mass spectrometer for their analysis.

The enhancement in performance brought about by nanoESI compared to classical ESI is crucial for some fields of applications, such as bioanalysis, and especially proteomics. Miniaturizing ESI provides increased sensitivity of the analysis and enables its application for the analysis of complex samples with a wide range of analyte concentrations. When working with nanoESI the probability of ionizing molecules is higher, and this is of great importance for the analysis of complex real-world samples where some compounds are present as traces.

As ESI works on a continuous flow of liquid, it has quickly been coupled to LC or other liquid-phase separation techniques as an alternative to optical detection.15  Mass spectrometry gives more information on the eluted compound, and the resulting hyphenated technique enables one to decrease the complexity of samples before their analysis by MS. High performance liquid chromatography (HPLC) is coupled to conventional ESI-MS while nanoLC is connected to nanoESI-MS for a better match in the flow-rate values.

Matrix-assisted laser desorption ionization has simultaneously been developed by Karas and Hillenkamp16  in Germany and by Tanaka et al.17  in Japan in 1985. With this technique molecules are ionized via laser irradiation of the sample and with the help of other small organic molecules, called the matrix. The matrix strongly absorbs the light of the laser and transfers it together with a charge, mostly a proton, to the analytes. Thereby, analytes reach the gas phase as ions that are ultimately analyzed by the mass spectrometer.

A MALDI-MS analysis proceeds as follows. A solution of matrix is mixed with the sample to be analyzed. A droplet of the resulting mixture is deposited on a plate and allowed to dry; the evaporation of the solvents leads to the co-crystallization of the analytes with the matrix. Once placed under vacuum, the spots of crystal are irradiated with the laser; the laser energy is absorbed by the molecules of the matrix, and subsequently transferred to the analytes that are desorbed from the surface. Simultaneously, the analytes capture a proton from the matrix molecules and become ions. Analytes finally reach the state of ions in the gas phase after desolvation of the matrix molecules.

As for ESI, the MALDI technique depends neither on the properties of the analytes nor on the absorption properties of the molecule as energy transfer proceeds via molecules of matrix. Besides, the mass or the size of the molecules does not influence the ionization and desorption process so that it can be applied to any molecule. Contrary to ESI, analyzed species are mono-charged with MALDI; this simplifies the interpretation of mass spectra but this imposes working with a detector covering a wider mass (m/z) range.

The laser wavelength can be either in the UV or infrared range, with the former being most commonly used. Therefore, matrix molecules are aromatic compounds that can absorb UV light and present a carboxylic acid moiety for the protonation of the analytes. Matrix molecules are simultaneously detected together with the analytes, and give peaks in the low mass range, i.e. below m/z 600. Consequently, MALDI-MS analysis is often limited to compounds with a molecular weight above m/z  600. Compared to ESI, the MALDI technique has a higher tolerance to the contamination of samples, and especially to the presence of salts that strongly hinders analyte ionization in ESI-MS. On-target sample cleaning is widely used in case of a high level of contamination. Another advantage of MALDI compared to ESI is the higher throughput of the analysis and its possible automation.

While ESI works on continuous flows of samples, MALDI analysis is performed on droplets or discrete amounts of liquid. Yet, recent developments have aimed at coupling a separation step relying on liquid chromatography to MALDI-MS analysis. For that purpose, the liquid eluted from the chromatography column is deposited in a continuous and automated way on a MALDI target.18,19  Other improvements concern the MALDI target to alleviate the use of a matrix. The matrix can be covalently attached on the target plate surface to avoid its desorption together with the analytes,20  or targets based on porous silicon21  are used that does not require the addition of a matrix and that gives enhanced analysis sensitivity. Lastly, an obvious improvement for the process of ionization comes from on-target concentration of the analytes, as will be discussed later. An easy way to obtain sample concentration is to confine it on a smaller surface area. Subsequently, a major breakthrough recently was obtained with the appearance of smart targets consisting of a uniform hydrophobic area patterned with small hydrophilic spots where the sample is deposited.22 

Coupling a microfluidic or microfabricated system to MS appears to be fruitful for both the detection on the chip (microchip point of view) and for the MS analysis (MS point of view). On the one hand, MS is a powerful technique for on-chip detection due to its sensitivity and the amount of information it provides on the sample; on the other hand, by using microfluidics prior to the MS analysis, new opportunities for the field of MS are created as it provides better MS capabilities compared to conventional sample preparation techniques.

A first interest of MS when used in combination with microfabricated structures, or at the outlet of microfluidic devices, is the match in the volume of liquid handled. A typical MS analysis requires less than 1 μL of liquid, for ESI-MS as well as for MALDI-MS techniques. When working with a continuous flow of liquid and ESI-MS, the MS performance is even more enhanced for flow rates down to 50–100 nL min−1: the lower the flow rate, the better the MS analysis. This flow-rate range corresponds to flow-rate values observed in microfluidic devices. Consequently, the technique of MS is easily scalable and exhibits an enhanced response when the sample size is decreased. This is not the case for instance for other detection techniques, such as UV absorbance or amperometry: these two techniques require large detection area or volume, which is the opposite of the quest of microfluidics. This first advantage of MS compared to other technique goes together with its high sensitivity.

Mass spectrometry is a fairly universal technique: it is a label-free detection technique and can be applied to any molecule as long as it can be ionized (i.e. that it presents a protonation site for instance). Optical techniques (e.g. UV absorbance and fluorescence-based techniques) require that the analytes present given properties to be detected. They must absorb or emit in the UV or in the fluorescence range. If they do not, they must be coupled to an external moiety (aromatic group, fluorophore) that presents such absorption/emission properties, and this imposes an additional derivatization of the analytes before their detection, but after their separation for instance on a microfluidic device. On the contrary, MS is widely used without any further derivatization step of the analytes for a large variety of species with different sizes: small organic compounds, inorganic compounds such as macromolecular complexes, any kind of biological samples, from metabolites up to large native proteins via oligosaccharides, nucleic acids and peptides. Lastly, MS enables the simultaneous analysis of different compounds, and consequently the analysis of complex samples. The technique presents a large dynamic range for the analysis and covers a wide range of concentrations, also in a single complex sample without any compound separation.

An obvious advantage of MS compared to other techniques is the amount of information a single MS analysis can provide about a sample. Optical techniques give a binary response regardless of the analyzed compound; absorption/emission or not. It is therefore difficult to identify the analytes as these techniques do not yield any further information on the nature or the structure of the analytes. Mass spectrometry does. Firstly, analytes are detected as a function of their molecular weight and the precision of the mass determination lies in the region of 10−4 amu. From this basic information, the size of a molecule can be deduced, and the molecule can be identified. The precision of the molecular weight of the species is particularly crucial for biological analysis, and the characterization and identification of large compounds such as peptides and proteins. Secondly, MS analysis is nowadays routinely performed in the tandem (MS2) or even MSn mode, where molecules are analyzed 2 or n times, respectively. For instance, for a MS2 analysis, molecules are analyzed a first time to provide their molecular weight. Thereafter, they are individually selected, fragmented and the resulting fragments analyzed anew. The characterization of the resulting fragments provides structural information on the analyzed species. Tandem MS is widely used for instance in the fields of proteomics, glycomics, metabolomics or organic chemistry. For the two former fields of applications, fragmenting the analytes enables one to derive their sequences. In particular, for proteomics analysis, the peptide sequences are used for database mining and the identification (name, nature and origin) of proteins present in the analyzed samples. The fragmentation of smaller compounds such as small synthetic organic molecules or metabolites gives the precise structure of the compounds, and for instance the localization of substituted groups. NMR analysis provides the same type of structural information, but the latter requires a much larger amount of material (several milligrams) for a single analysis.

Lastly, MS is compatible with high-throughput analysis. Each analysis, for both ESI-MS and MALDI-MS, is fast and lasts for less than a minute. As already mentioned, MALDI-MS analysis can easily be automated with dedicated software so that hundreds of samples can be analyzed without human intervention. ESI-MS lends itself less to high-throughput analysis as sample preparation and introduction in the ionization capillary are more tedious and time consuming. Still, automation is fully conceivable when coupled to LC for instance.

The coupling to microfabricated structures or a microfluidic-based preparation of samples also benefits the MS analysis. Replacing conventional ionization tools by microfabricated structures brings a tremendous enhancement of the analysis quality. As discussed above, both ESI-MS and MALDI-MS can exhibit enhanced performance when using smaller ionization sources for introducing the samples or smaller spots for sample deposition, respectively. In both cases, the smaller the structures, the higher the ionization yield and the less the amount of sample required and consumed. Moreover, using microfabricated structures guarantees their reproducibility, and thereby the reproducibility of the analysis. NanoESI in particular suffers from the poor quality and a lack of reproducibility of the ionization sources whose fabrication relies on heating and pulling technique to realize apertures in the range of a few micrometers. For ESI-MS, the use of microfabrication techniques reliably produces small sources with decreased characteristic dimensions, leading to enhanced ionization phenomena. Moreover, the electrical contact for applying the ionization high voltage can be fabricated using robust microfabrication processes and integrated into the sources providing better and smoother conditions for high-voltage application. Similarly, microfabricated targets for MALDI-MS analysis would present smaller patterns for sample deposition, leading thereby to on-target sample concentration.

The chip format is fully compatible with the use of robots and the automation of the analysis, leading to high-throughput analysis or screening. The connection to a robotic interface is already routinely done for MALDI-MS analysis, and is especially interesting for ESI-MS for which the preparation of the samples is tedious and time consuming, and requires a skilled operator for the introduction of the sample in the ionization source. In that case, the sample preparation step represents a limitation for high-throughput analysis. This is no longer required if a microfluidic system is coupled to a robotic and automated interface. MALDI-MS analysis benefits also from the use of an automated and robotic interface for sample deposition on a target. Lastly, microfluidic systems lend themselves also to analysis multiplexing; as they are fabricated using microtechnology techniques, each single microchip can include several independent analysis microsystems (“horizontal integration”). This possible integration on a microchip is another step towards high-throughput analysis especially if the different systems “work” simultaneously.

One strong point of microfluidic systems is their high level of integration. This integration can be “horizontal” when a microfluidic chip presents several identical analytical systems in parallel, but also “vertical” when several subsequent steps of a single process can be performed on-line using a single system. Sample preparation prior to MS analysis usually proceeds in several steps of sample treatment (e.g. reaction and digestion), separation and especially purification (e.g. desalting) and concentration. In a conventional protocol these different steps are performed manually by an experimenter and off-line, and this implies extensive and time-consuming manipulation of the samples, and as a result increased risk of sample loss and contamination, and human error. For instance, preparation of proteomic samples proceeds as follows. Proteins are extracted from a crude biological matrix, the proteins are processed (denaturation, disulfide bond reduction, etc.), digested, and the resulting peptides separated and purified using LC techniques. Particularly, the last step of sample separation and purification limits the analysis throughput, as a liquid-based separation roughly lasts one hour, which is much longer than the acquisition time for a single mass spectrum. On the contrary, if carried out on a microfluidic device, the preparation of the samples is fully integrated and much faster, because of the reduced size of the microsystems and the absence of human intervention, than when using a conventional methodology and gives rise to few or none of the issues listed above. Besides the factor of time, microfluidic systems lend themselves better to the control of low flow rates that leads to better ESI-MS analysis conditions. Lastly, using a fully integrated microsystem appears to be a cheaper alternative to bulky and conventional devices (e.g. nanoLC columns).

The next step to get fully integrated analytical tools that perform a total sample analysis from sample preparation to sample detection is to miniaturize the detection instrumentation and to integrate it into the same miniaturized device. Such efforts have already been described and successfully achieved for optical detection techniques. In the field of MS, the miniaturization of the instrumentation (mass spectrometer) started much before the miniaturization trend for organic and biological applications. Such development is driven by the need to have portable detection systems for on-site analysis for various fields of applications, among which, for instance, are spatial or environmental analysis and homeland safety. The mass spectrometer itself is often not sufficient for performing such analysis, especially when dealing with complex mixtures. Consequently, coupling the MS analysis with a front-end sample preparation step appears to be necessary so as to purify, concentrate or even separate species present in the crude sample. Thereby low-abundance species can be detected from complex samples that present a high level of contamination.

Additionally, as already mentioned earlier in this chapter, coupling a miniaturized device to MS is a sort of paradox if one compares the difference in sizes of the two parts of the analysis setup: while the microsystem has a size in the millimeter range and is portable, the mass spectrometer can be up to several meters in size and some hundreds of kilograms in weight! As a consequence, the analysis is confined to dedicated laboratories. Coupling microfabricated systems to a miniaturized and portable mass spectrometer having a size comparable to that of a microsystem to give a fully integrated and portable analysis setup would also benefit, for instance, the fields of biological and medical analysis.

The combination of miniaturized devices with MS can be divided in several categories, first according to the ionization technique which is used (ESI vs. MALDI), and second depending whether the device is fluidic (microfluidic device) or static (microfabricated structure or device), as a large number of miniaturized devices for MALDI-MS analysis mainly consist of smart MALDI targets without any integrated microfluidics. Fluid handling is in that case carried out by either an external experimenter or an automated and robotic interface. We will distinguish here these different “categories”.

Electrospray ionization works on a continuous flow of liquid, wherever the sample is introduced in the mass spectrometer by infusion or by using a pumping system that controls the flow rate (e.g. connection to liquid-based chromatography). Therefore, coupling a microsystem appears to be easier and more natural for ESI-MS analysis than for MALDI-MS, as is the case for a LC-MS connection. The microfluidic chip can be seen in that case as a miniaturized and more integrated LC system that is connected to a mass spectrometer. However, in the case of a microsystem the optimization of the coupling represents a bigger challenge, especially to avoid any kind of dead volume, as for the smaller systems the dead volumes are larger!

Coupling microfluidic systems to ESI-MS has been achieved via three different technological approaches. In the first approach, the electrospray is generated at the outlet of the microchannel, directly from the edge of the chip. This approach was firstly reported by Ramsey and Ramsey in 1997.23  The ionization voltage was applied at the inlet reservoir of the chip. Nonetheless, this simple coupling approach was not seen to be very successful as the liquid had a tendency to spread around the channel outlet instead of forming a spray. Two improvements were subsequently introduced, a pneumatic assistance to force the formation of an electrospray and a hydrophobic coating on the side surface of the chip to alleviate any spreading phenomena.

The second alternative relies on the use of a conventional capillary ionization source which is connected at the outlet of a microchannel, the capillary being directly inserted into the outlet channel or connected to the chip via a transfer capillary.24  The use of a capillary enables one to focus the liquid and direct it into the mass spectrometer. Using this configuration the spray quality is roughly the same as for conventional ESI-MS analysis and several other issues arise from this type of coupling. Capillaries are inserted manually which prevents any mass production of integrated systems. Besides, the connection between the microfluidic channel and the capillary inserted therein must be optimized to minimize any dead volume.25  Lastly, the inserted capillary is glued on the chip to maintain it in place, and care must be taken that the glue does not dissolve in the sample, which would induce background noise on the mass spectra.26 

The third approach is to micromachine the ionization source as a sharp structure and to integrate it onto the microfluidic chip. This last alternative apparently does not suffer from any of the aforementioned issues. The ionization source is fabricated at the same time as the rest of the microsystem, and should exhibit better characteristics than commercial capillary-based sources (smaller dimensions and reproducibility). With this approach, the production of multiplexed microsystems with separate sources for any individual device to avoid any cross-contamination issue is conceivable and easier. This microfabrication-based route to produce integrated microsystems or only ionization sources has already given rise to several devices based on different designs and materials. Structures can be classified according to two main types of designs, whether the ionization source reproduces a capillary-type cylindrical structure27,28  or whether the ionization source is planar and placed horizontally at the end of the outlet microchannel.29  Microfabricated sources have been made from different materials such as silicon,28  and also from polymer-based materials such as polydimethylsiloxane (PDMS),29  poly(methyl methacrylate),30  SU-8,31  polyimides,32  poly(ethylene terephthalate),33  parylene,34  polycarbonate,27 etc. The choice of the material has two main impacts on the structure and the analysis quality: it should be compatible with the production of reduced-sized structures in the micrometer range and should not give rise to any contamination issue, via degradation or dissolution in the solvent used for the analysis for instance. Another improvement to microfabricated sources consists of integrating the electrode for application of the ionization high voltage;33  this has been reported to give smoother and enhanced ionization conditions. A number of these microfabricated structures, demonstrated in stand-alone conditions or integrated in a more complex microfluidic device, are described in this book, in the chapters presenting ESI-related microfluidic development.

MALDI ionization mostly takes place in vacuum in spite of the recent development of atmospheric pressure ionization sources. Therefore, coupling a microfluidic system to MALDI-MS implies working under vacuum on the chip or performing the MS analysis off-line, i.e. introducing the microchip in the MALDI-MS once the fluidic operations are finished. Consequently, the first miniaturized developments for MALDI-MS analysis only concerned the microfabrication of enhanced or smart MALDI targets that do not include any fluidic component.

Microsystems for MALDI-MS analysis can be divided into three groups: microfabricated MALDI target plates, microsystems for off-line preparation of samples and microsystems that integrate both the sample preparation steps and the MALDI targets, with an on-line or off-line analysis of the samples.

The first category comprises micromachined MALDI targets that aim at concentrating and eventually cleaning samples on site using passive structures. One limitation in MALDI analysis lies in the fact that samples spread on the surface of the target upon deposition, and this results in a decrease of analysis sensitivity. To circumvent this liquid spreading, MALDI targets can be patterned so as to decrease the MALDI spot size and thereby to yield local concentration of the analytes. Two main approaches have been reported. The first one consists of machining nanovials where samples are deposited. These vials have an inverted pyramid structure, at the bottom of which the sample is concentrated.35  Such systems were first realized in silicon,35  and more recently their realization has been demonstrated in a polymer material.36  The analysis time is also decreased as the experimenter does not need to search for the sample on a large spot of crystal. The second approach relies on a chemical patterning of the surface. Typically, the whole surface of the MALDI target is made hydrophobic (via appropriate coating, a monolayer or the use of a hydrophobic material) and small hydrophilic spots are created therein.22  Thereby, the sample is confined to the hydrophilic spots, and the smaller the spots, the higher the sensitivity enhancement. Both approaches give a sensitivity improvement down to the attomole range. Another major technological development in this direction is the implementation of the purification step on such a chemically patterned MALDI target.20  For that purpose, the hydrophilic areas of the MALDI target are functionalized with a given stationary phase aiming at retaining a specific class of compounds. These devices have in particular been developed by the company Ciphergen and are now commercially available under the name ProteinChip® devices.20 

The second category of microsystems for MALDI-MS analysis focuses on the sample preparation steps before deposition on a (conventional) MALDI target. MALDI-MS analysis is subsequently done off-line and not on the same device. Such microsystems comprise two parts, one for the sample treatment (digestion, purification, separation) and the other for sample dispensing onto the MALDI target. The first part consists of a microfluidic device working on a continuous flow of liquid. The second part aims, as before, at limiting sample spreading on the MALDI target; however, instead of structuring a MALDI target, a robotic interface is used to improve the deposition of the sample on a confined area on a (conventional) MALDI target. Sample dispensing has been described using various types of technologies: a spotting technology, piezoelectric actuation or by spraying the sample as a thin film. The spotting technology18  borrows much from the interface which is used for coupling LC to MALDI-MS, but works here for smaller amounts of liquids. Another technique uses a piezoelectric dispenser35  that enables continuous deposition of sample on a MALDI plate via sub-nanolitre droplets of liquid. As soon as the droplet reaches the surface, solvent evaporates so that the sample does not spread on the surface. This dispensing technique has been coupled to a microreactor for the tryptic digestion of proteins and to a nanovial-based MALDI plate.35  An original dispensing method developed by the group of Murray is based on a rotating ball placed at the outlet of a microfluidic system on which samples are separated;37,38  the rotating ball makes the interface between the zone at atmospheric pressure where samples are prepared and the zone under vacuum where MALDI ionization takes place. This technique is described in more detail in Chapter 10. The last reported technology to couple sample preparation steps with a continuous flow to MALDI-MS consists of spraying the sample into a uniform film onto the MALDI target surface.39  This technology provides more homogeneous sample crystals, whose size is dictated by the distance between the spraying head and the MALDI surface. Spots down to 170 μm diameter with 150 μm spacing have been realized using the spraying technique. Sample deposition has been demonstrated after digestion of protein samples inside the spraying device using immobilized trypsin. One major advantage of these devices is their ability to be multiplexed and to work in parallel without any risk of contamination from one sample to another.

The third and last group of microfluidic systems for MALDI analysis comprise integrated systems that include both steps of the analysis, i.e. sample preparation and the MALDI target. The advantage of these systems is that they are fully integrated and hardly need any external intervention for sample analysis. While for most of these devices sample preparation is carried out off-line (i.e. not in the mass spectrometer), one unique system works “under vacuum” and enables on-line sample treatment and analysis. In this 3rd category, the first microfluidic device was developed by Gyros AB and gave rise to a commercial product, the Gyrolab®. The Gyrolab® is dedicated to biological sample analysis using affinity chromatography, and particularly targets the field of proteomics applications.40  This device has the shape of a compact disk that includes a series of radial microfluidic networks with MALDI targets placed at the external edge. Sample preparation is performed off-line on a dedicated workstation, with a robot making the link between conventional 96-well plates and Gyrolab® devices that contain 96 independent analysis networks. Thereafter the device is cut and the MALDI target parts are placed in the mass spectrometer for the analysis. Liu et al.41  have developed an alternative system whose channels are not covered with a lid for accessibility to the laser. In those open microchannels, peptide and oligosaccharide samples to which a matrix solution has first been added are separated using CE and dried in place under vacuum. MALDI analysis is done on the crystals formed in the channels by scanning along the whole channel with the laser. EWOD (electrowetting on a dielectric) is particularly suitable for MALDI-MS analysis, where liquids are displaced as droplets, i.e. the format which is conventionally used for MALDI analysis. Consequently, several integrated microsystems for MALDI-MS analysis rely on this pumping principle that enables sample treatment and mixing of the sample with the matrix solution. Wheeler et al. have reported EWOD-based microfluidic systems for MALDI-MS analysis. In a first system,42  the principle of EWOD is exploited to displace a sample droplet, to mix it with a droplet of matrix solution and to place it on a Teflon-based MALDI target. They further improved their system by including a purification step of the sample;43  a droplet of sample is displaced onto a Teflon-coated site where analytes and contamination adsorb. This spot is washed with deionized water to remove contamination, and later matrix is added to the purified analytes. In both cases, the system is opened at the end of sample preparation and spots are dried under vacuum before the system is introduced in the mass spectrometer. Chapter 12 presents an alternative system based on EWOD which is dedicated to kinetics studies;44  its performance is illustrated here for the investigation of an enzymatic reaction. The functioning principle of this device relies on the successive mixing within a few milliseconds of liquid droplets: (i) between the enzyme and the substrate solutions, (ii) the reaction mixture and a quenching agent, and finally (iii) the addition of the matrix solution to the quenched reaction mixture. A last and unique integrated microsystem for MALDI-MS analysis enables the on-line preparation of the samples within the mass spectrometer. This system which is described in Chapter 11 benefits from the vacuum environment inside the ionization source of the MALDI mass spectrometer for fluid actuation. Detection is done via the irradiation of the sample, either in the outlet reservoir of the device45  or through a detection window added for this purpose in the outlet channel of the chip.46  This system has successfully been demonstrated for the investigation of both biochemical and chemical reactions.

The combination of miniaturization to MS analysis finds two main fields of applications: biological/biochemical analysis and organic chemistry. These are two important fields of applications of MS analysis. In both cases, by using microfabricated devices for MS analysis or by combining microfluidics to MS, novel capabilities are reached and new applications are found.

The main field of application of microfluidics-to-MS coupling is the field of biological and medical diagnostics, and especially proteomics or protein analysis. Microfluidic developments in this field of analysis are driven by the large number of advantages provided by the miniaturization of the analysis and the coupling of microfabricated devices to MS analysis. Most of these advantages are already listed in Section 1.2.2. These advantages can be divided in three categories: (i) the gain in analysis quality and sensitivity, (ii) performance improvement by the high level of integration of sample treatment and preparation and (iii) the suitability of such microfabricated structures for high-throughput analysis. Using microfabricated tools and structures bring an enhancement in the analysis sensitivity. For ESI-MS analysis, this is accounted for by the improved ionization efficiency which is reached by decreasing the size of the ionization source, while MALDI benefits from both the miniaturization of the spots of analytes and the use of a microfabricated interface for sample deposition. Moreover, the miniaturization of the analysis may also result in a greater tolerance to sample contamination. This is of major interest for specific subfields of biological analysis where compounds cannot be amplified as is the case for nucleic acid analysis. For metabolomics applications, analytes of interest may be present as traces in the samples and crude samples consist of complex biological matrix. Proteomics analysis suffers from the same issues: proteins present a large dynamic range and proteins of interests are mostly of low abundance. Besides that, for proteomics applications analysis resolution is also an issue to guarantee the accurate identification of proteins using database mining. This is now made possible with the development and improvement of FT-MS instruments. Sample treatment and preparation prior to MS analysis can be fully integrated on a microfluidic device limiting sample manipulation. This high level of integration results in faster analysis and higher sample quality due to decreased loss and contamination. Moreover, more steps can be integrated on a single chip, and for proteomics applications a single chip can be used for performing sample preparation from the protein digestion step to the ionization of a purified and separated peptide mixture. Lastly, due to the reduced sample preparation and analysis times by using miniaturized systems and the possibility to integrate a (large) number of independent devices on a single system, microchips are a potential tool for high-throughput and multiplexed analysis. This is all the more reinforced by the easy connection of microfluidic systems to an automated and robotic interface.

The connection of microfluidics to mass spectrometry is also widely used for the investigation of biochemical and organic chemistry reactions. Chapters 9, 11 and 12 illustrate such couplings for both ESI-MS and MALDI-MS. The advantages of direct coupling between a microfluidic system on which a reaction is carried out and a mass spectrometer are manifold. Firstly, it should be emphasized that with a microfluidic system the reaction mixture is always more homogeneous than in conventional glassware. Therefore, on-line analysis using a microsystem is always done on a homogeneous medium. Consequently, MS analysis from a microsystem gives precise snapshots and information on the progression of a given reaction in real time, as long as the reaction is “stopped” before the reaction mixture is introduced in the mass spectrometer. For ESI-MS, this is easily achieved by simply coupling a microfluidic system or simply short capillary tubing to a mass spectrometer and by taking care to avoid any dead volume. For MALDI-MS, unless the reaction is performed within the mass spectrometer and analyzed on-line, care should be taken to quench the reaction at given time points. The real-time characterization of reactions makes possible novel investigation of reaction kinetics, as reported for instance in Chapter 12 for the study of enzymatic reactions. Moreover, these continuous and real-time snapshots also enable one to study and identify reaction intermediates as long as their lifetime is long enough for an MS analysis. A last and novel application of such microfluidics-to-MS coupling concerns high-throughput analysis and the screening for instance of libraries of compounds using multiplexed microfluidic systems. This novel analytical tool in this context appears to be complementary to combinatorial chemistry studies that give rise to huge amounts of compounds to be characterized and identified.

A brief history of mass spectrometer miniaturization is given in Chapter 13. Developments for the miniaturization of MS instrumentation have been especially driven until now by a few fields of application that require on-site analysis and portable analysis instruments. However, novel fields of analysis would much benefit from fully integrated and portable analysis tools as we already discussed in Section 1.2.3.

One main field of applications of miniaturized mass spectrometers is related to space exploration.47  In such a context, the instrumentation must fulfill a number of requirements, starting from its size, weight and power consumption and going to its resistance to harsh environmental conditions (microgravity, exposure to shock and vibrations, for instance) and including of course its analytical performance and ability to detect trace-level species in complex samples. In that field, miniaturized mass spectrometers are mainly developed for two types of analysis: determining the composition of planetary atmospheres and monitoring the air quality in aircraft during long-duration manned space missions, with especially the detection of volatile organic compounds (VOCs) in the air.48 

A second important application of miniaturized mass spectrometers which is currently growing rapidly in the USA is homeland safety and the prevention of civil (bio)terrorism. For such purposes, the miniaturized mass spectrometer can also include a system for wireless data transmission for remote site sensing.49 

More recently, handheld mass spectrometers have also been successfully used for the study of ion/molecule reactions50  as well as for biological analysis, and the detection of peptides, oligonucleotides and biological spores.51 

Recent achievements in the miniaturization of MS instrumentation gave rise to the first instruments with characteristic dimensions in the micrometer range52–54  as in the case of microsystems dedicated to analysis and chemical processing. Such achievements should lead in the near future to the development of a fully integrated setup including the steps of both sample preparation and detection.

In this book we endeavor to cover most of the topics highlighted in the introduction, and to present (i) technological developments achieved for the coupling of microfluidic systems to MS, both for the ESI and MALDI ionization techniques, (ii) relevant applications of microfluidics-to-MS couplings and (iii) research aiming at miniaturizing the mass spectrometer itself. We hope thereby to give a complete and comprehensive view to the reader of the “miniaturization and mass spectrometry” field with this collection of chapters.

The book is divided into three sections, dealing respectively with ESI-MS coupling, MALDI-MS analysis and the miniaturization of mass spectrometers. The ESI-MS section starts by illustrating early couplings of microfluidic systems to nanoESI-MS using a transfer capillary. Subsequent chapters describe the development of novel microfabricated sources for nanoESI-MS analysis, these sources being integrated on a microfluidic system or not (yet), using various and eventually new geometries, and using various materials (silicon, parylene, SU-8, PDMS, polyimide). Finally, the application of on-line nanoESI-MS analysis after a microfluidic step of sample handling is demonstrated in the fields of (i) proteomics with the digestion/separation/purification of samples performed on a microchip and analyzed on-line by ESI-MS and (ii) organic chemistry and biochemical analysis. The second section of the book reports on MALDI-MS analysis after microfluidics through three chapters reporting a novel microfluidic-to-MALDI target interface and two fully integrated microfluidic systems. The first chapter describes an original interface for coupling a microfluidic device for proteomic sample preparation to a MALDI target, for both off-line and on-line analysis. In the second chapter of the section, on-line monitoring of reactions carried out within a mass spectrometer and using a microfluidic chip with vacuum-driven actuation of liquid is discussed. Lastly, the functioning of an EWOD-based microfluidic system is illustrated for real-time kinetic studies of an enzymatic reaction. The last section of the book gives an example of a micro-mass spectrometer (microMALDI-TOF-MS) and its applications for homeland security and clinical diagnostics.

Close Modal

or Create an Account

Close Modal
Close Modal