Microchip Atmospheric Pressure Ionization-Mass

Spectrometry

By

Pekka Östman

Division of Pharmaceutical Chemistry

Faculty of Pharmacy

University of Helsinki

Finland

Academic dissertation

To be presented with the permission of the Faculty of Pharmacy of the

University of Helsinki, for public criticism in Auditorium 1041 of Viikki

Biocenter on March 2nd, 2007, at 12 o’clock noon

Helsinki 2007

2

Supervisors:

Professor Risto Kostiainen Division of Pharmaceutical Chemistry Faculty of Pharmacy University of Helsinki Finland Professor Tapio Kotiaho Environmental Chemistry and Environmental Analytical Chemistry Laboratory of Analytical Chemistry Department of Chemistry Faculty of Science University of Helsinki Finland

Reviewers: Professor Jonas Bergquist Analytical Chemistry & Neurochemistry Department of Physical & Analytical Chemistry Biomedical Centre Uppsala University Sweden Docent Seppo Auriola Department of Pharmaceutical Chemistry Faculty of Pharmacy University of Kuopio Finland

Opponent:

Professor Frants R. Lauritsen Analytical Environment Chemistry Department of Chemistry Faculty of Science University of Copenhagen Denmark

© Pekka Östman 2007 ISBN 978-952-10-3669-9 (printed version) ISSN 1795-7079 ISBN 978-952-10-3670-5 (PDF) http://ethesis.helsinki.fi

Helsinki University Printing House Helsinki 2007

Finland

3

Contents LIST OF ORIGINAL PUBLICATIONS ................................................................... 4 ABBREVIATIONS AND SYMBOLS ...................................................................... 5 ABSTRACT.............................................................................................................. 6 1. INTRODUCTION................................................................................................. 8 2. REVIEW OF THE LITERATURE...................................................................... 10

2.1 Lab-on-a-chip and microfluidics .................................................................... 10 2.1.1 Introduction ............................................................................................ 10 2.1.2 Materials................................................................................................. 12

2.2 Microfluidic devices with ESI-MS................................................................. 14 2.3 Low-flow rate APCI and APPI ...................................................................... 18 2.4 DIOS ............................................................................................................. 21

2.4.1 MALDI-MS............................................................................................ 21 2.4.2 DIOS-MS................................................................................................ 22 2.4.3 DIOS-MS applications ............................................................................24

3. AIMS OF THIS STUDY ..................................................................................... 25 4. MATERIALS AND METHODS......................................................................... 26

4.1 Chemicals, materials, and instrumentation ..................................................... 26 4.2 Instrumentation.............................................................................................. 30 4.3 Microchip APCI-MS ..................................................................................... 30

4.3.1 Microfabrication process of the microchip APCI..................................... 30 4.3.2 Microchip-MS interfacing setup.............................................................. 32 4.3.3 Characterization of the prototype microchip APCI-MS (I)....................... 34 4.3.4 GC/microchip APCI-MS (II)................................................................... 34 4.3.5 CapLC/microchip APCI-MS (III)............................................................ 35

4.4 AP-DIOS-MS................................................................................................ 36 5. RESULTS AND DISCUSSION .......................................................................... 38

5.1 Microchip APCI ............................................................................................ 38 5.1.1 Microchip characterization (I-III) ............................................................ 38 5.1.2 GC/microchip APCI-MS (II)................................................................... 42 5.1.3 CapLC/microchip APCI-MS (III)............................................................ 45

5.2 AP-DIOS-MS................................................................................................ 48 5.2.1 Effect of the PA of an analyte on ionization in AP-DIOS-MS (V) ........... 49 5.2.2 Suitability of AP-DIOS-MS for analysis of drugs (IV) ............................ 53

6. CONCLUSIONS................................................................................................. 54 ACKNOWLEDGEMENTS..................................................................................... 56 REFERENCES........................................................................................................ 58

4

LIST OF ORIGINAL PUBLICATIONS

This doctoral thesis is based on the following five articles, hereafter referred to by

their Roman numerals (I-V):

I Östman, P., Marttila, S. J., Kotiaho, T., Franssila, S., and Kostiainen R.

Microchip Atmospheric Pressure Chemical Ionization Mass Spectrometry.

Analytical Chemistry 2004, 76, 6659-6664.

II Östman, P., Luosujärvi, L., Haapala, M., Grigoras, K., Ketola, R. A., Kotiaho,

T., Franssila, S., and Kostiainen, R. Gas Chromatography - Microchip

Atmospheric Pressure Chemical Ionization - Mass Spectrometry. Analytical

Chemistry 2006, 78, 3027-3031.

III Östman, P., Jäntti, S., Grigoras, K., Saarela, V., Ketola, R. A., Franssila, S.,

Kotiaho, T., and Kostiainen R. Capillary Liquid Chromatography - Microchip

Atmospheric Pressure Chemical Ionization - Mass Spectrometry. Lab on a Chip

2006, 6, 948-953.

IV Huikko, K., Östman, P., Sauber, C., Mandel, F., Grigoras, K., Franssila, S.,

Kotiaho, T., and Kostiainen, R. Feasibility of Atmospheric Pressure

Desorption/Ionization on Silicon Mass Spectrometry in Analysis of Drugs.

Rapid Communications in Mass Spectrometry 2003, 17, 1339–1343.

V Östman, P., Pakarinen, J., Vainiotalo, P., Franssila, S., Kostiainen, R., and

Kotiaho, T. Minimum Proton Affinity for Efficient Ionization with Atmospheric

Pressure Desorption/Ionization on Silicon Mass Spectrometry. Rapid

Communications in Mass Spectrometry 2006, 20, 3669–3673.

5

ABBREVIATIONS AND SYMBOLS

AP atmospheric pressure

AP-DIOS atmospheric pressure desorption/ionization on silicon

APCI atmospheric pressure chemical ionization

API atmospheric pressure ionization

capLC capillary liquid chromatography

CE capillary electrophoresis

DIOS desorption/ionization on silicon

DRIE deep-reactive ion etching

ESI electrospray ionization

GC gas chromatography

HF hydrofluoric acid

i.d. internal diameter

LC liquid chromatography

LOD limit of detection

[M+H] + protonated molecule

MALDI matrix-assisted laser desorption/ionization

MEMS microelectromechanical systems

MS mass spectrometry

MS/MS tandem mass spectrometry

MW molecular weight

m/z mass-to-charge ratio

PA proton affinity

pSi porous silicon

RSD relative standard deviation

SEM scanning electron microscopy

S/N signal-to-noise

TEST testosterone

µ-TAS micro-total analysis systems

6

ABSTRACT

The atmospheric pressure ionization techniques in mass spectrometry (MS),

electrospray ionization (ESI), atmospheric pressure chemical ionization (APCI), and

atmospheric pressure photoionization (APPI), play key roles in the life sciences,

biopharmaceutical chemistry, and environmental research. ESI is the most commonly

used ionization method and the discovery of nano electrospray (nano ESI) and its

combination with nanoliquid chromatography (nanoLC) has established the modern

forms of proteomics, system biology, and metabolomics. With the pressing need for

analytical methods faster than nanoLC/MS, microfluidic ESI-MS systems for

combining rapid separations with high detection sensitivity have been intensively

studied. The complementary API techniques, APCI and APPI, have been overlooked

so far.

For the first time, a novel microchip heated nebulizer for atmospheric pressure

chemical ionization mass spectrometry (APCI-MS) was developed and evaluated. The

microchip included a capillary insertion channel, stopper, vaporizer channel, nozzle,

and nebulizer gas inlet fabricated on a silicon wafer, either anisotropic wet-etched or

deep-reactive ion-etched, and a platinum heater sputtered on a glass wafer. These two

wafers were joined by anodic bonding, creating a two-dimensional version of an

APCI source. The microfabricated stopper in the vaporizer channel ensured correct

positioning of the capillary, which led to reproducible insertion and minimization of

the dead volumes inside the microchip APCI. The etched nozzle in the microchip

formed a narrow sample plume that was ionized by an external corona needle, and the

ions formed were analyzed with a mass spectrometer. The nebulizer chip provided

flow rates down to 50 nl/min thus enabling, for the first time, the use of low-flow rate

separation techniques with APCI-MS.

The feasibility of the microchip APCI for quantitative work was investigated with gas

chromatography/mass spectrometry (GC/MS) and capillary liquid

chromatography/mass spectrometry (capLC/MS). For GC/MS work, a set of volatile

organic compounds and testosterone representing semivolatile compounds were used.

7

The spectra produced by microchip APCI showed intensive protonated molecules and

some fragmentation products as in classical chemical ionization for structure

elucidation. In quantitative analysis the GC/microchip APCI-MS showed good

linearity (regression coefficient (r2) = 0.9989) and repeatability (relative standard

deviation (RSD = 4.4%, n=6)). The limits of detection (LODs) with signal-to-noise

ratio (S/N) of 3 were between 0.5 and 2 µmol/l in the MS mode, using selected ion

monitoring and 0.05 µmol/l with MS/MS using multiple reaction monitoring. The

capLC/microchip APCI-MS work was done with selected neurosteroids. The

capLC/microchip APCI-MS provided quantitative repeatability (RSD, n=6, of peak

areas varied between 12.7% and 18.9%) and good linearity (r2 of area varied between

0.9919 and 0.9994). The LODs (S/N=3) in the MS/MS mode for the neurosteroids

were 20 - 1000 fmol (10 - 500 nmol/l). Comparison of the LODs showed that the

microchip APCI sensitivity in terms of concentration in the sample was comparable to

that of macro-APCI, but the mass flow sensitivity was about 100-200 times better

with microchip APCI. Rapid heat transfer allowed the use of optimized temperatures

for each compound during an LC run.

In another approach, the feasibility of atmospheric pressure desorption/ionization on

silicon-mass spectrometry (AP-DIOS-MS) for drug analysis was investigated. It was

observed that only compounds with relatively high proton affinity, above a threshold

value of 920-950 kJ/mol, are efficiently ionized under AP-DIOS conditions. The

LODs achieved in the MS mode with midazolam, propranolol, and Angiotensin II

were 80 fmol, 20 pmol, and 1 pmol respectively. In the MS/MS mode the LODs for

midazolam and propranolol were 10 fmol and 5 pmol, respectively. The good linearity

(r2 > 0.991), linear dynamic range of 3 orders of magnitude, and repeatability (RSD =

35%, n=7) showed that the method was suitable for quantitative analysis.

8

1. INTRODUCTION

With the increasing demand for faster and more cost-effective analytical methods,

miniaturization of analytical instruments is becoming ever more popular. The major

motivations of miniaturization are increased speed of analysis, lower sample and

reagent consumption, and reduced waste production. Due to the high surface-to-

volume ratio diffusion times are short, allowing fast reactions. These miniaturized

analytical instruments can be highly automated and with the small dimensions, many

systems can be constructed in parallel, providing increasing sample throughput.

Additional advantages include robustness, reliability, the potential for mass

production resulting in low production costs, and use of disposable microchips. The

feasibility of integration of different functional units such as reaction, separation, and

detection on a single microchip consisting of all features of a complete lab has led to

the concepts ‘Micro-Total Analysis Systems’ (µ-TAS) and ‘lab-on-a-chip’. Decreased

sample volumes require an adequate, highly sensitive detection method. The detection

methods used with µ-TAS have most often been optical methods (such as absorption

and fluorescence), electrochemical methods, and mass spectrometry (MS). Most

microfluidic applications so far have relied on fluorescence detection, but MS is

attracting increasing interest for its high specificity and sensitivity.

Modern liquid chromatography/mass spectrometry (LC/MS) is based on atmospheric

pressure ionization (API) techniques: electrospray ionization (ESI), atmospheric

pressure chemical ionization (APCI), and the recently introduced atmospheric

pressure photoionization (APPI). These ionization techniques play key roles in the life

sciences, biopharmaceutical chemistry, and environmental research. ESI is an

excellent technique for ionic and polar compounds, but only polar solvents can be

used and the ionization efficiency for neutral and nonpolar compounds may be poor.

Furthermore, ESI is not amenable to gaseous samples. In contrast, APCI and

especially APPI are better ionization techniques for less or non polar and neutral small

molecules, but they are not suitable for large biomolecules. APCI provides high

ionization efficiency for polar and neutral compounds, allows use of polar and

nonpolar solvents, and tolerates higher electrolyte concentrations. Additionally,

9

suppression of ionization by coeluting compounds is significantly less with APCI than

with ESI. At present ESI is the most common ionization method and to date, it has

been the only ion source available for low flow rate capillary- and nano liquid

chromaography (capLC and nanoLC) applications and microfluidics. The flow rates

used with commercial APCI and APPI ion sources are high, typically above 50

µl/min, making them incompatible for low-flow rate separations.

Desorption/ionization on silicon mass spectrometry (DIOS-MS) is a relatively new

matrix-assisted laser desorption/ionization (MALDI)-related technique. In DIOS,

chemically etched porous silicon (pSi) is used as a sample support and as a substrate

to assist desorption/ionization instead of the matrix compounds used in MALDI. The

fact that no addition of matrix is needed reduces the sample preparation time and

produces high-quality mass spectra that are essentially free of the background peaks at

the low-molecular weight ranges (MW < 800) encountered in MALDI-MS. DIOS is a

relatively soft ionization method and therefore efficient ionization with little

fragmentation of the sample molecule is typically observed. Atmospheric pressure

(AP)-DIOS-MS differs mainly from conventional vacuum DIOS-MS in that the laser

DI steps are conducted under AP, as are the ESI and APCI techniques. AP-DIOS

provides easier interfacing to existing API instruments, including tandem instruments

such as triple quadrupoles and ion traps, and easier sample transfer than with vacuum

DIOS instruments.

In this work the feasibility of two miniaturized API techniques for MS were

evaluated. First, a novel microchip heated nebulizer for APCI-MS was developed and

evaluated (I ). Based on this prototype a new version was designed for analytical work.

Both anisotropic wet etching (II ) and deep-reactive ion etching (DRIE) (III )

fabrication methods were employed in production. The performance of the microchip

as an interface for API-MS for qualititative and quantitative work was evaluated with

gas chromatography (GC) (II ) and capLC (III ). Next, the feasibility of AP-DIOS-MS

for drug analysis was investigated (IV ) and the role of proton affinity (PA) in the

ionization process AP-DIOS-MS was systematically studied (V).

10

2. REVIEW OF THE LITERATURE

2.1 Lab-on-a-chip and microfluidics

2.1.1 Introduction

The current trend in analytical chemistry is miniaturization of analytical instruments,

using microfabrication processes to improve sensitivity and speed of analysis.

Microfabrication is based on the fabrication processes originally developed for micro-

electromechanical systems (MEMS) within the microelectronics industry.

The first analytical microfluidic chip introduced (1978) was a GC etched on a 5-cm

silicon wafer [1]. However, it was the introduction of the µ-TAS concept by Manz et

al. in 1990 [2] that resulted in a tremendous increase in the research and development

of microfluidic chips. The µ-TAS, or ‘lab-on-a-chip’, is a system with integrated

modules that performs all steps in an analysis: sampling, sample pretreatment, sample

transport, chemical reactions, separation, and detection. Microfluidics is used for

precise control of liquid (or gas) in the lab-on-a-chip system. The benefits of

microfluidics include improved performance by increased speeds and higher

throughputs of sample, parallel reactions, fast response, portability of instruments,

highly automated analyses, reduced manufacturing and operating costs, and reduced

resource consumption and waste production. The applications range from clinical

application [3,4], cell analysis and manipulation [5], DNA [6-8], combinatorial

chemistry [9], drug discovery [10,11], high throughput analysis [12], bioanalytical

and pharmaceutical applications [6,13], sizing and quantitation of proteins [6,14] and

peptides [6], and environmental analysis [15-17].

The typical microchip is a planar silicon, glass, or polymer device with an overall size

ranging from millimeters to centimeters and an incorporated network of

interconnected microchannels. The most popular fluid propulsion technique has been

electroosmotic flow. There are several advantages in using electroosmotic flow, such

as simple liquid control, simple chip fabrication, flat flow profile provided by the

electroosmotic flow, and fast analysis [13]. Some examples include on-chip capillary

11

electrophoresis (CE) separation of rhodamine B and di-chlorofluorescein in 0.8 ms

[18] and a baseline separation of 19 amino acids in 165 s [19]. Separation was also

performed with miniaturized LC [20-24]. There is also a commercially available

‘HPLC-Chip’ from Agilent. A novel concept in fluid movement is the fully automated

compact disc format microfluidic system utilizing centrifugal force for fluid

movement, which was developed for sample preparation prior to MALDI-MS analysis

[25-27], and is commercially available from Gyros. The same centrifugal force

concept has also been applied for multiple enzymatic assays [28].

The complexity of bioanalytical samples usually requires sample pretreatment.

Sample pretreatment has been realized in microfluids either by conventional methods,

such as solid-phase extraction, or utilizing the new phenomena offered by scaled-

down devices, such as laminar fluid diffusion interfacing. In laminar fluid diffusion

interfacing two or more streams flow in parallel in a microfluidic chip with no mixing

of the fluids other than by diffusion of particles across the diffusion interface. The

laminar fluid diffusion interface was demonstrated in the separation of small and large

molecules [29], and was successfully applied to blood sample preparation [30] and

on-line desalting of macromolecule solutions [31]. On-chip solid-phase extraction was

done with monolithic porous polymers [32,33], C-18 modified channels [34],

octadecylsilane-coated silica beads [35,36] and octadecylsilane-packed columns [37],

and water-soluble poly(acrylamide-co-alkylacrylamides) [38]. Membrane-based

sample pretreatment can include desalting, elution of analytes from the membrane,

and direct analysis of small drugs, peptides, and proteins [39], concentration of

proteins using a porous silica membrane between adjacent microchannels [40], and

miniaturized supported liquid membrane extraction for sample enrichment [41].

The small sample volumes used in microfluidics require adequate, high-sensitivity

detection techniques. In most microfluidic applications so far, detection has relied on

optical techniques, especially on laser-induced fluorescence for its excellent

sensitivity. Furthermore, the sensitivity is greatly enhanced by reducing the size of the

detection volume, because the background signal that is generated by impurities in the

sample scales linearly with the size of the detection volume [42]. However,

derivatization schemes are needed with nonfluorescent compounds. An alternative to

laser-induced fluorescence is electrochemical detection. Electrochemical detectors can

12

be relatively easily integrated into a microfluidic chip with standard MEMS processes

without loss of sensitivity [43]. In addition, many compounds can be detected without

derivatization. Other detectors such as chemiluminescence detector [44,45], refractive

index detector [46-47], micromachined photoacoustic detector [48], and ultraviolet

(UV) detector [49] have also been used, showing the versatility of detectors that can

be used in µ-TAS. However, for high-throughput sample identification and structure

elucidation, chip-based MS detection is the method of choice. MS is highly sensitive

and is a selective detector suitable for handling the reduced sample volumes of

microfluidics. The capability for performing high-speed separations with microchips

is fully matched by high-throughput MS instrumentation [50]. Furthermore, MS is

able to provide the molecular weight and structural information of an analyte. So far

interfacing of microfluidic devices with MS has been done using either ESI or

MALDI interfaces, of which ESI is more widely used [50-54]. The complementary

API techniques, APCI and APPI, were just recently introduced for interfacing

microfluidic devices with MS [I, 55].

2.1.2 Materials

Silicon and Glass

The fabrication of microfluidic devices was initially undertaken using

micromachining technologies employing silicon and glass [2,56,57]. These include

several different fabrication techniques that selectively add (e.g. sputtering of metal,

oxide growth, or resist spinning) or remove material (etching) from planar substrates.

Once a pattern has been produced in a resist material by lithography, it can be

transferred to the substrate by means of micromachining techniques.

The advantages of silicon are the good mechanical rigidity of microdevices, versatility

of the structural geometry obtainable, and well-developed surface micromachining

technology [58]. The well-developed micromachining technology enables easy

integration of different functionalities, e.g. electrical and thermal functionalities, into

the microdevices. Unfortunately, silicon is a semiconductor, and thick insulating

layers, such as of thermal oxide or nitride, need to be fabricated for high-voltage

applications. Glass is an insulator with a high breakdown voltage, and is optically

13

transparent to UV/visible light. The thermal expansion coefficient for many glasses is

in the same order as silicon, making glass a good substrate for bonding with silicon.

The etching methods for silicon are classified under four categories: wet anisotropic,

wet isotropic, dry anisotropic, and dry isotropic [59-61]. The etching rates for wet

etching are approximately 1 µm/min and for dry etching 1-10 µm/min. In anisotropic

wet etching the structures are determined by the different etching rates of the silicon

crystal planes, while isotropic etching occurs at the same rate in all directions. In

plasma (dry) etching the shape of the channels can be varied from isotropic to vertical

by varying the etching conditions.

On the microfabrication standpoint, glass is difficult to etch and has poor dimensional

control. Glass etching is usually done with isotropic wet etching, limiting the channel

geometry to the elliptical [58]. Other fabrication methods include electrochemical

discharge drilling [62], ultrasonic drilling [63], laser drilling [64], and DRIE [65-68].

The DRIE etching rates for glasses vary between 250 nm/min and 1.2 µm/min,

depending on the chemical composition of the glass [65-68].

Polymers

The advantages of polymers are the low cost of polymer materials, easier fabrication

compared with silicon and glass, freedom in geometrical designs, and the wide range

of material selection [69]. The fabrication processes can be divided into direct

fabrication and replication techniques [57,69]. In direct methods, such as laser

ablation, reactive ion etching, lithography, layering, and mechanical milling, polymers

are structured individually to form system features [57]. Direct methods are often

complex and slow. Replication methods, such as injection molding, hot-embossing,

compression molding, casting, and soft-lithography, on the other hand, involve the use

of a microfabricated molding tool which can then be used to replicate identical

polymer microstructures [57,69]. The replication method enables massproduction of

devices at low cost. A wide variety of polymers have been used, including

poly(methyl methacrylate) (PMMA), polycarbonate (PC), cycloolefin copolymer

(COC), polyimide, polystyrene (PS), poly(dimethylsiloxane) (PDMS) [8,57,69,70],

and SU-8 [70-72].

14

2.2 Microfluidic devices with ESI-MS

2.2.1 Electrospray ionization

Formation of gas-phase ions by ESI was demonstrated by Dole et al. [73] and the first

interfacing of ESI with MS was done by Yamashita and Fenn [74-75]. ESI is a

method in which a liquid is dispersed into small charged droplets by applying a high

electric potential between the tip of a thin capillary and a counter electrode, here the

MS. The ionization process can be divided into three steps: formation of charged

droplets at the tip of the capillary, evaporation of solvent from the droplets and

formation of gas-phase ions. The discovery of nanoelectrospray (nano ESI) and its

combination with nanoLC extended bioanalysis to attomole-sensitivity [76]. An

molecular weight accuracy of 0.01% could be obtained for proteins by applying a

signal-averaging method to the multiple charged ions formed in the ESI process [77].

At present ESI is the most common ionization method used in API-MS. ESI is an

excellent technique for ionic and polar compounds ranging from small molecules to

proteins and peptides, but the ionization efficiency for neutral and nonpolar

compounds may be poor. Other disadvantages include that only polar solvents can be

used and volatile buffers are preferable to minimize contamination. For neutral, less

polar, and ionic compounds APCI or APPI is often more suitable.

2.2.2 Microchip-ESI-MS

For microfluidics, ESI is an ideal ionization source because the flow rates in

microfabricated devices are of the order of 0-300 nl/min [50], thus being similar to the

flow rates of nano ESI sources [78]. The first designs, introduced by Xue et al. [79]

and Ramsey and Ramsey [80], for glass chips were those in which the electrospray

was initiated directly from the edge of the microchip. To date, many different chip-

based ESI-MS interfaces from different materials have been fabricated, as shown in

Table 1. The interfacing techniques can be characterized as two types: interfaces that

spray directly from an exposed channel at the side of a chip (on-chip spraying), and

those that use an external emitter attached to the microchip for spraying (off-chip

spraying) [51-53].

15

Table 1. Microchip-ESI-MS applications, microchip material used and ESI interface type.

Analytes Chip material Interfacing type Reference

Proteins Glass On-chip [79]

Tetrabutylammonium iodide Glass On-chip [80]

Proteins Glass with external ESI needle Off-chip [81]

Proteins Glass with external fused silica capillary nozzle Off-chip [82]

Proteins Polycarbonate (PC) with external ESI needle Off-chip [83]

Proteolytic digests Glass with external ESI needle Off-chip [84]Glass with fused silica nozzle

Peptides Glass On-chip [85]Glass with external ESI needle Off-chip

Small molecules Glass with external fused silica nozzle Off-chip [86]

Peptides and proteins Gglass with external ESI needle Off-chip [87]

Peptides and proteins PDMS (polydimethylsiloxane) with external ESI needle Off-chip [88]

Peptides and proteins Silicon with parylene emitters Off-chip [89]

Peptides Glass with external ESI needle Off-chip [90]

Peptides and proteins Glass with external ESI needle Off-chip [91]

Proteins Silicon nozzle [92]

Small molecules, proteins Glass with external ESI needle Off-chip [93]

Proteins PC with PC emitter Off-chip [94]

Proteins Glass On-chip [95]Glass with external ESI needle Off-chip

Proteins PMMA (poly(methyl methacrylate)) On-chip [96]

Carnitines Glass with external ESI needle Off-chip [97]

Drug molecules Glass with external ESI needle Off-chip [98]

Proteolytic digests Glass with external ESI needle Off-chip [99]

Small molecules Polymer Zeonor 1020 with external ESI needle Off-chip [100]

Protein digest Glass with external ESI needle Off-chip [101]

Proteins PET (poly(ethylene terephthalate)) On-chip [102]

Peptides PDMS with PDMS emitter Off-chip [103]

Small molecules and proteins Parylene C nozzle [104]

Peptides PDMS with external ESI needle Off-chip [105]

Peptides PC with emitter Off-chip [106]PMMC (poly(methyl methacrylate)) with emitter

Peptides SU-8 nozzle [107]

Drug molecules PDMS On-chip [108]

Protein digest Silicon nozzle [109]

Peptides Glassy carbon with external ESI needle Off-chip [110]

Proteins SU-8 nozzle [111]

Peptides PDMS with external ESI needle Off-chip [112]

Peptides PDMS with external ESI needle Off-chip [113]

Drug molecules SU-8 nozzle [114]

Proteins SU-8 with SU-8 emitter Off-chip [115]

Peptides PPM (porous polymer monolith) emitter [116]

Peptides PDMS with external ESI needle Off-chip [117]

Peptides PPM emitter On-chip [118]Proteins Micro ion spray source Off-chip [119]Methanol-water PC emitter array On-chip [120]

On-chip spraying

The on-chip spraying design is attractive due to its simplicity. The outlet can be

formed simply by dicing the chip and it is also free from deadvolume, since no

16

external spraying capillary or needle is needed for spraying. However, in the first

studies the solution exiting the channel resulted in problems by spreading on the

hydrophilic glass surface, which led to the formation of large liquid droplets (volume

of about 12 nl [80]). This wetting prevented the formation of a well-focused electric

field essential for the generation of stable electrospray [79]. In addition, such large

liquid droplets caused excessive band-broadening and sample dilution [84]. For

microchip separations combined with on-chip ESI-MS, in which peak volumes are

typically below 5 nl, any separation would be lost in the too large dead volume of the

electrospray cone [50]. Wetting was minimized by derivatizing the edge of the

glass[79,80], by pneumatically assisting the spray [85], and by fabricating the

microchip from a hydrophobic polymer [108]. Alternatively, on-chip spraying devices

can be fabricated using silicon [92,109] or various polymers such as parylene

[89,104], PC [120], PMMA [118], PDMS [103], and SU-8 [107]. Traditionally,

silicon has allowed very precise and small dimensions for the emitter (15 µm),

resulting in reliable and reproducible electrospray with signal stability and intensity

that are comparable to those obtained using a pulled capillary of similar dimensions

[92]. On the other hand, the polymers used are moldable to any shape and due to their

inherent hydrophobic character, they are suitable for use as ESI emitters with no

further modification procedures needed [54]. In addition, disposable microchip ESI-

MS devices can be made from polymers at low cost. However, the resistance of

different polymers to organic solvents will need to be explored in more detail.

Off-chip spraying

In off-chip spraying, a conventional fused-silica capillary [81,82], an ESI tip [86-

87,91] or embedded spraying capillary [105] is attached to the outlet of the

microchannel. Since a sharp electrospray tip or nozzle is used good ESI conditions are

more readily achieved, resulting in small, well-defined droplets that enhance both

sensitivity and resolution [121]. The ESI capillary can also be removed and replaced

if clogged, preserving the microfluidic device [84]. In addition, this design generally

results in performances comparable to those found for microcolumn separations [50].

However, precise low dead volume alignment of the ESI tip with the separation

channel [86] is essential for maintaining separation efficiency by preventing band-

broadening in these microchip designs [50]. An alternative design, a miniaturized

pneumatic nebulizer coupled with a subatmospheric liquid junction ESI interface can

17

be used [85,91,97,98,100,119]. In the liquid junction interface, a suitable solvent that

is continuously delivered with the aid of an external pump to the liquid junction

interface ensures the transport of the sample through the pneumatically assisted ESI

needle. Proper adjustment of the capillary ends and dimensions in the interface

ensures stable sample delivery to the spray and mixing of the CE buffer with a spray

solution from the liquid junction [85].

Both on-chip CE and on-chip LC separations have been used with microchip ESI-MS.

Zhang et al. recorded separation performance similar to that of standard CE

(separation efficiency >300 000 plates/meter) with on-chip CE separations of peptides

and tryptic digest on a glass chip followed by ESI-MS with a liquid junction-based

ESI interface [85]. Li et al. used a nanospray emitter on a glass chip for the on-chip

CE combined with ESI-MS of synthetic peptide mixtures [90]. Peptide separations

were conducted in less than 90 s with peak widths of approximately 2 s with low

nanomolar detection limits for various peptides and for in-gel digests of proteins. As

an estimation, less than 25 ng of original protein was used in the analysis. Deng et al.

used a glass chip for on-chip CE coupled with a miniature microsprayer via a

microliquid junction at the exit of the CE capillary channel for on-line ESI detection

of carnitine and selected acylcarnitines [97]. Carnitine and three acylcarnitines were

separated in less than 48 s with a measured CE separation efficiency of 2860 plates

(peak width at half-height method). An on-chip ESI-MS device [92], a benchtop

LC/MS, and fraction collection have recently been integrated into a commercial

analytical platform by Advion. Alternatively, it can be used as an infusion ESI-MS

interface. Another LC-based separation device, a microfluidic LC with ESI-MS

detection, was demonstrated [20,22]. The microfluidic device was comprised of a

laser-ablated enrichment column and a reversed-phase separation channel, integrated

rotary valve, and a nano ESI emitter embedded together in polyimide layers. The

device was connected to an external gradient capillary pump, and a microwell plate

autosampler for sample loading and mobile phase delivery. The system provided good

reproducibility of retention time and peak intensity with relative standard deviation

(RSD) values of less than 0.5% and 9.1%, respectively. Sensitivity measurements on

protein digests spiked into rat plasma provided a detection limit of 1-5 fmol. It is

commercially available from Agilent Technologies.

18

While the use of EOF and CE separations in microfluidics has several advantages,

such as its simplicity and the flat flow profile provided by the electroosmotic flow, the

reproducibility and robustness of the on-chip CE/ESI-MS microfluidic devices appear

challenging, and no commercially available analytical platform exists.

2.3 Low-flow rate APCI and APPI

In APCI [122,123] and in APPI [124,125], the ionization of gas-phase sample

molecules is initiated by electrons emitted by the corona discharge needle (Fig. 1) or

by a krypton discharge lamp emitting 10-eV photons, respectively. This is in contrast

to ESI, in which charge separation occurs in the liquid phase at the end of the

microchannel or capillary (section 2.2.1 Electrospray ionization).

H O3+H O3

+

O2

O2O2

O2

N2

H O2

N2

O2

MH+

S

S

SH+

O2

N2

M

SS

M

H O2 N2N2

M

H O2

MS

Corona dischargeneedle

M = analyte molecule= solvent moleculeS

+•

H O2+•

+•

+•

Figure 1. Schematic drawing of an atmospheric pressure chemical ionization.

For LC/MS and CE/MS using either APCI or APPI, the eluted compounds are

vaporized with the aid of nebulizer gas and heat. However, the gas-phase ionization

process also enables the use of GC for sample separation or direct gaseous sample

introduction.

APCI is an ionization technique that is capable of ionizing with high efficiency both

polar and ionic compounds in addition to neutral compounds [122,126]. Other

advantages of APCI over ESI include the use of both polar and nonpolar solvents and

the toleration of higher electrolyte concentrations. Furthermore, suppression of

19

ionization by the coeluting compounds is significantly less with APCI than with ESI

[127]. APPI has the same advantages as APCI, but is even more suitable for nonpolar

compounds [124]. Since the ionization occurs in the gas phase, only relatively small

and thermally stable compounds up to about 1000 amu can be analyzed.

LC/APCI-MS and LC/APPI-MS are widely used in environmental analysis, drug

discovery, metabolics, and bioanalytics [128-136 and references therein]. In addition,

APCI-MS without chromatographic separation was used in environmental analysis

[137-145] and real-time analysis of breath and volatile flavors [146-154]. GC/APCI-

MS was used successfully in environmental analysis in the 1980s [155-166]. In the

early 1990s Ravelsky et al. used GC/APPI-MS for a wide variety of small organic

molecules [167,168].

Commercial APCI and APPI ion sources have been designed primarily for

conventional LC at relatively high flow rates, typically above 50 µl/min, making them

incompatible for low-flow rate separations. To take full advantage of the benefits of

microfluidic devices (section 2.1.1 Introduction) and also the improved sensitivity

provided by cap- and nanoLC, a miniaturized ion source compatible with low flow

rates is essential.

Although some attempts have been made to facilitate APCI ion sources for flow rates

below 10 µl/min, these have been done either by custom-made sources or

modifications of commercial sources and they lack the concept of microchip-based

miniaturization. Tyrefors et al. constructed a custom-made interface for supercritical

fluid chromatography [169]. LODs (with S/N = 3) for anthracene and trilaurin were

about 1 ng and 10 pg, respectively. Repeatabilities (RSD%) of retention time and

relative peak height were 0.24% and 2.6%, respectively. The interface was later

modified by Nyholm et al. for high-temperature open-tubular LC to allow more

efficient vaporization of the liquid mobile phase prior to ionization [170]. The flow

rates of the mobile phase ranged between 0.1 and 1.6 µl/min. The mass flow

sensitivity of 7,8-benzoquinoline was 3 pg/s, corresponding to a concentration of 1

µM using a flow rate of 1 µl/min. Tanaka et al. modified a commercial APCI interface

to accommodate flow rates in the range of 1-10 µl/min for CE/APCI-MS [171]. The

20

modification was done by drilling a hole to the interface to accommodate a stainless

steel tube, providing coaxial sheath liquid flow, around the fused-silica sample

capillary. At a sheath liquid flow rate of 5 µl/min an occasionally unstable MS signal

was observed, presumably due to the breaking of electrical contact across the sheath

liquid, but increasing the sheath liquid flow rate up to 10 µl/min solved the problem.

Nilsson et al. compared CE with APPI-MS and ESI-MS using sheath liquid

interfacing in the analysis of small pharmaceuticals [172]. The flow rate of the sheath

liquid in CE/ESI-MS varied between 5 and 25 µl/min. The lower sheath liquid flow

rate gave a higher signal from the analytes but also resulted in increased background.

In the CE/APPI-MS experiment, the sheath liquid flow rates varied between 15 and

100 µl/min. No clear trends were observed for any of the analytes, supporting the

theory that APPI behaves as a mass-flow sensitive technique. Compared with ESI, the

APPI technique provided a cluster-free background, indicating that the APPI process

is less affected by nonvolatile salts in CE buffers and that a wider range of CE buffers

can be used in CE/APPI-MS analysis then in CE/ESI-MS. Kauppila et al.

demonstrated the use of the same nebulizer chip as used for microchip APCI (the

prototype microchip APCI) in dopant-assisted microchip APPI-MS [55]. Ionization in

the positive and negative ion modes was successfully achieved for naphthalenes and

the spectra were in general similar to those obtained with conventional APPI. The

micro-APPI was compatible for flow rates in the range of 0.05-5 µl/min and was most

efficient at 1-5 µl/min. A stable signal was demonstrated throughout a 5-h

measurement, which proved the excellent stability of microchip APPI.

21

2.4 DIOS

2.4.1 MALDI-MS

MALDI [173] is a soft, i.e. causing little fragmentation, MS ionization technique

widely used for large biomolecules (such as peptides, proteins, oligonucleotides, and

oligosaccharides) and synthetic polymers. In MALDI the sample is mixed with a

matrix, typically in the ratio of 1:1000, and spotted onto a target plate where the

mixture crystallizes. The three most commonly used matrices are 3,5-dimethoxy-4-

hydroxycinnamic acid (sinapinic acid), α-cyano-4-hydroxycinnamic acid (alpha-

cyano or alpha-matrix), and 2,5-dihydroxybenzoic acid (DHB). The crystallized

mixture is irritated with either an infrared (IR) or UV laser and the matrix transforms

the laser energy into excitation energy for the sample. The DI in MALDI is believed

to be a two-step process: primary ionization during or shortly after the laser pulse

followed by ion-molecule reaction in the expanding plume of desorbed material [174].

So far MALDI has been used only in a limited way for small molecules, because the

matrix causes background interference in the low-mass region (MW < 800). Recently,

it was shown that the analysis of small molecules with MALDI can be accomplished

by suppressing the matrix background at an appropriate analyte-to-matrix molar ratio

[175,176] or by using an ionic liquid matrix [177,178].

MALDI has also been combined with microfluidics. The groups of Laurell and

Marko-Varga developed a ‘microfabricated toolbox’ for protein identification, in

which the sample protein digest is spotted with a piezoelectric microdispenser into a

nanovial MALDI target plate and measured with MALDI-MS [179-183]. Liu et al.

demonstrated on-chip CE separation followed by off-line MALDI-MS detection of

proteins [184]. Mok et al. performed protein separation in a plastic chip placed on the

standard MALDI plate [185]. Brivio et al. physically incorporated a continuous-flow

microchip with integrated microdigestion reactor into the standard MALDI sample

plate of an MS instrument [186]. A high-density fully automated compact disc format

microfluidic system was developed for protein sample preparation prior to off-line

MALDI-MS analysis [26-27,187].

22

2.4.2 DIOS-MS

DIOS-MS is a relatively new MALDI-related technique introduced in 1999 by Wei et

al. [188]. In DIOS (Fig. 2) chemically etched pSi is used as a sample support and as a

substrate to assist ionization instead of the matrix compounds used in MALDI. The

fact that the addition of matrix is not needed reduces the sample preparation time and

produces high-quality mass spectra of low-MW (MW < 800) that are essentially free

of the background peaks encountered in MALDI [188-190]. A higher salt tolerance

was also suggested [189,191]. DIOS is a relatively soft ionization method and thus

typically results in efficient ionization with little fragmentation of the sample

molecule [192].

MS

Laser

pSi areas

Silicon substrate

Figure 2. Schematic drawing of desorption/ionization on silicon (DIOS).

The pore morphology and overall porosity greatly impact DIOS efficiency, providing

a suitable structure - a silicon 'skeleton' with up to 80% empty space [61,193] and

internal surface area up to 600 m2/cm3 [58] - for retention of analytes and solvent

molecules [189,190,194]. Pore sizes approximately 200 nm in depth and 50-100 nm in

diameter are best for a wide range of analytes [195]. In the case of UV-DIOS, the high

surface area and strong UV absorption of pSi promote energy transfer from the

substrate to the trapped analytes [189]. In the case of IR-DIOS, the IR laser excites the

23

vibrational groups of the solvent or the analyte itself, leading to desorption and

ionization of the surface solvent and analyte in the expanding plume [196,197].

The DI processes in DIOS have not yet been solved. It is currently believed that

surface roughness, not porosity, is the key element in DI. This was demonstrated with

various mechanically created roughness [196-198] or MEMS-created surface

structures, such as deposited nanostructured thin film [199], ordered silicon

nanocavity arrays [200], silicon nanowires [201], or various submicrometer structures

[202]. The formation of gas-phase ions is initiated on the surface of sharp crystal tips

that protrude out of the sample surface [195,201,203]. During laser irradiation, the

tips act as tiny antennas producing significant field enhancement in the vicinity of the

nanometer tip. Thus, the laser energy is efficiently focused onto a small area [201].

However, porosity plays an important role by creating a scaffold on which more

analytes and solvent molecules can be retained [200] (compared with rough surfaces)

and by resupplying the surface with analyte after a laser pulse [195].

Studies of the fundamentals of DIOS have shown that protonation is the favored

ionization process in DIOS [189,190,198,204]. Deprotonation [189,190,204] and

radical cation formation [203,205] were also observed. It was suggested that the

sources of protons in the positive-ion mode are the silicon hydride surface [189,204]

and residual solvents or contaminants on the surface [188,198]. Recently, Budimir et

al. [206] suggested that ionization in DIOS occurs in the gas phase. They observed

alkali-adduct homo- and heterotrimers with a defined statistical distribution and

concluded that the existence of this statistical distribution reflects a situation that

cannot exist in solution.

Significant improvements have been made to DIOS-MS through surface

modifications. Derivatization of the pSi can be used to make it more resistant to air

oxidation [190,204], ozone oxidation, and acid/base hydrolysis [207]. Derivatization

of the pSi can also serve to enhance DIOS-MS [194,207,208], provide a lower

background, and require less laser power for DI [204,207]. For example, very high

sensitivity, such as 800 ymol for des-Arg9-bradykinin, was attained using a silylated

pSi surface [207].

24

2.4.3 DIOS-MS applications

DIOS-MS has been widely applied for small molecule analysis such as drug

molecules [188,195,204,209-211], illicit drugs [212], and organic dye [203]. Peptides

and proteins have also been investigated, in more then half of the publications listed in

Table 2.

Table 2. DIOS-MS applications.

Analytes Note Reference

Small drug molecules, peptides Native and dodecyl-, ethyl-, phenyl-, and oxide-derivatized pSi [188]

Peptides, proteins Positive- and negative-ion mode [189]

Small molecules, peptides [190]

Exacrine tissue and single neuron [191]

Amino acids, peptides [192]

Small molecules, peptides Biotin-avidin-coated, silylated and oxidized pSi [194]

Forensics [195]

MALDI matrix compounds, peptides, proteins IR-DIOS-MS [196]

Peptides IR-DIOS-MS [197]

Organic dye Reduction of dye [203]

Small molecules Alkane-, alkene, and carboxylic acid-derivatized pSi [204]Positive- and negative-ion mode

Porphyrin derivatives [205]

Fatty acids Negative-ion mode [206]

Peptides Silylated oxidized pSi [207]

Proteins [208]

Small drug molecules, peptides AP-DIOS-MS [209]

Small drug molecules [210]

Small molecules, protein digest [211]

Amphetamines and fentanyls [212]

Proteins [213]

Proteins Perfluorinated surfacants [214]C8F17- and C18-derivatized pSi

Polypropyleneglycol mixtures [215]

Enzymatic reaction of the immobilized enzymes [216]

Enzyme-inhibition reactions [217]

Proteins [218]

Nucleic acid, carbohydrate and steroid analysis [219]

Pentose-borate complexes Silicon dioxide-derivatized pSi [220]

Catecholamines [221]

Cysteine sulfonic acid-containing peptides Positive- and negative-ion mode [222]

25

3. AIMS OF THIS STUDY

The overall aims of the study were to develop and evaluate new microchip-based

interfacing techniques for API-MS.

Specifically, the aims of the research were:

• to develop a miniaturized heated nebulizer for APCI-MS and evaluate its

performance as an interface for capLC/MS and GC/MS (I, II, III ).

• to test the suitability of DIOS-MS at atmospheric pressure for drug analysis,

evaluate its performance for quantitative analysis of drugs, and to clarify the

ionization mechanism for various low-molecular-weight compounds (IV, V ).

26

4. MATERIALS AND METHODS

The chemicals, samples, materials, instruments, analytical methods, and

microfabrication processes are briefly described in this section. The chemicals and

instruments are listed in tables, whereas the microfabrication processes and analytical

procedures are shortly described. More detailed descriptions can be found in the

original publications (I-V ).

4.1 Chemicals, materials, and instrumentation

The reference standards, chemicals, and materials used in this study are listed in Table

3. The purpose of each item is briefly noted.

27

Table 3. Standard compounds, chemicals, and materials used in the study.

Standard/chemical/material Manufacturer/supplier Note Publication

Midazolam Sigma-Aldrich (Germany) Reference standard IV

Hoffman-La Roche (Switzerland)

V

Propranolol Sigma-Aldrich Reference standard I,IV, V

Testosterone Sigma-Aldrich / Fluka (Switzerland) Reference standard I,IV, V

D3-Testosterone Sigma-Aldrich Reference standard IV

1-(Methylamine)-naphthalen Sigma-Aldrich Reference standard IV, V

2-Naphthyl acetic acid Sigma-Aldrich Reference standard IV

2-Naphtoic acid Sigma-Aldrich Reference standard IV

Ketoprofen Sigma-Aldrich Reference standard IV

Paracetamol Sigma-Aldrich Reference standard IV

Paracetamol glucuronide Sigma-Aldrich Reference standard IV

Angiotensin II Sigma-Aldrich Reference standard IV

1,4-Naphtaquinone Fluka Reference standard IV

9-Aminoacridine May & Baker (England) Reference standard V

Benzo[h]quinoline Sigma-Aldrich Reference standard V

Verapamil Sigma-Aldrich Reference standard I,V

4-Acridinol Fluka Reference standard V

Fluorescein Sigma-Aldrich Reference standard V

9-Acridinecarboxylic acid Fluka Reference standard V

Dopamine Fluka Reference standard I,V

Anthracene Fluka Reference standard V

9-methyl-9-anthracenecarboxylate Sigma-Aldrich Reference standard V

Acenaphthene Sigma-Aldrich Reference standard V

4-Bromo-1H-Pyrazole Sigma-Aldrich Reference standard V

4-Chloro-Benzenamine Sigma-Aldrich Reference standard V

9H-Carbazole B.D.H. (England) Reference standard V

2-Naphthalenecarboxylic acid Sigma-Aldrich Reference standard V

Naphthalene Merck (Germany) Reference standard V

2-Naphthalenol Sigma-Aldrich Reference standard V

Acridine Sigma-Aldrich Reference standard I

28

Table 3. (Continued)

Quinoline Acros (Belgium) Reference standard I

Benzaldehyde Sigma-Aldrich Reference standard II

2-Acetylnaphthalene Sigma-Aldrich Reference standard II

Anisole Fluka Reference standard II

Acetoacetone Merck Reference standard II

Dehydroisoandrosterone Sigma-Aldrich Reference standard III

Pregnenolone Sigma-Aldrich Reference standard III

Testosterone Sigma-Aldrich Reference standard III

Progesterone Sigma-Aldrich Reference standard III

Ammonia Merck Reagent I-III

Peroxide Merck Reagent I-III

Hydrogen chloride Merck Reagent I-III

Tetramethyl ammonium hydroxide Honeywell (Germany) Reagent I

Buffered hydrofluoric acid Merck Reagent I-III

Phosphoric acid Merck Reagent I

Acetic acid Merck/VWR Int. (Finland) Reagent I

Potassium hydroxide Merck Reagent II

2-Propanol Rathburn (UK) Reagent / solvent I,II,III,IV

AZ 4562 Clariant AZ, AG (Germany) Photoresist II, III

AZ 5214 AZ Electronic Materials, GmbH (Germany) Photoresist III

Chromium Testbourne Ltd (England) Reagent II,III

Platinum Kultakeskus Oy (Finland) Heater material II,III

Aluminum Testbourne Ltd (England) Heater material I

Hydrofluoric acid Merck/VWR Int. Reagent IV, V

Chlorotrimethylsilane Sigma-Aldrich Reagent III

Methanol J.T.Baker (Holland) Solvent I-V

Water Millipore (USA) Solvent I-V

Ethanol Primalco (Finland) Solvent IV, V

Acetone J.T.Baker Solvent I,II,III,V

Hexane Merck Solvent II

Acetonitrile Rathburn Solvent III

29

Table 3. (Continued)

N2 Whatman 75-720 generator (USA) Nebulizer/curtain /drying/collision gas

I,II,III,IV

N2 Woikoski (Finland) Nebulizer gas I

He Woikoski Carrier gas II

SF6 AGA (Finland) Etchant III

C4F8 AGA Etchant III

O2 AGA Etchant III

Ar AGA Reagent II,III

Air Atlas Copco (Belgium) Auxillary gas I,III

Pyrex 7740 glass Corning (USA) Substrate material I-III

Silicon wafers (100), <0.025 ohm cm resistivity

Okmetic (Finland) Substrate material I

Silicon wafers (100), >500 ohm cm resistivity

Okmetic Substrate material II,III

Epoxy Loctite (Finland) Epoxy glue I

High-temperature epoxy Cotronics (USA) Epoxy glue II,III

Nanoport Upchurch Scientific (USA) Fluidic connectors I-III

PEEK tubing, i.d. 50 µm, o.d. n µm Upchurch Scientific

Sample inlet capillary I

PEEK tubing, i.d. 510 µm, o.d. n µm Upchurch Scientific

Nebulizer gas inlet tubing I-III

FactorFour VF-5ms Varian (USA) GC column II

Deactivated fused-silica capillary, i.d. 0.15 mm, o.d. 0.22 mm SGE (USA) Sample inlet

capillary II

Capillary Column Butt Connector Supelco (USA) Column connector II

SymmetryShield RP18 (0.32 mm, 100 mm, 3.5 µm)a

Waters (USA) LC column III

XTerra MS C-18 (2.1 mm, 100 mm, 3.5 µm)a Waters LC column III

Deactivated fused-silica capillary, i.d. 50 µm, o.d. 220 µm SGE

Sample inlet capillary III

NanoTight Fitting Upchurch Scientific Column tubing fitting III

Darning needle Entaco Limited (England) Corona discharge needle

I-III

a (i.d., length, particle size)

30

4.2 Instrumentation

The instrumentation used in this study is listed in Table 4.

Table 4. Instrumentation used in the experimental work.

Instrumentation Model / type Manufacturer Publication

Mass spectrometer PE Sciex API-300 PE Sciex (Canada) I-III1100 LC/MSD Trap Agilent Technologies (USA) IVQ-Tof micro Waters Micromass (UK) V

Ion source APCI PE Sciex I-IIIAP-MALDI Mass Tech Inc. (USA) IV-V

Liquid chromatograph HP1050 Hewlett-Packard GmbH (Fed. Rep. Of Germany) I, III1100 Series Capillary LC system Agilent Technologies (Germany) III

Gas chromatograph HP5890A gas chromatograph Hewlett-Packard (West Germany) II

Syringe pump Harvard PHD 2000 Advanced Syringe Pump Harvard Apparatus Inc. (USA) I,IIINanopump Upchurch Scientific (USA) III

Loop injector Rheodyne 7725 Rheodyne (USA) I

Eluent flow splitter Acurate AC-100-VAR LC Packings (Switzerland) I

Thermometer Fluke 54 Series II Fluke Corporation (USA) I-III

Power supply GPS-3030 Good Will Instruments Co. Ltd (Taiwan) I, II

EPS EP-6515 EPS Stromversorgung GmbH (Germany) II, IIIHP E3632A Agilent Technologies IV-V

Nitrogen generator Atlas Copco Wilrijk (Belgium) I-III

Aligner Electronic Visions AL6-2 Electronic Visions (USA) I-III

Etcher STS ASE STS ASE (England) III

4.3 Microchip APCI-MS

4.3.1 Microfabrication process of the microchip APCI

The microchip APCI consists of two wafers: a silicon wafer and a Pyrex glass wafer.

The silicon wafer consists of fluidic inlets, vaporizer channel, and a nozzle. The

integrated heater was fabricated on the Pyrex glass wafer.

The prototype (Fig. 3a) had three through-wafer inlets: one for the sample and two for

the nebulizer gas (I ). The fluidic network was fabricated by anisotropic wet etching in

a 25 wt-% tetramethyl ammonium hydroxide solution at 80 °C. The etch rate was

about 0.5 µm/min and was minimized by etching from both sides of the wafer

simultaneously. The widths of the nebulizer gas and liquid sample channels were 300

µm and 120 µm, respectively, the channel depth was 190 µm. A 300-nm aluminum

layer was sputtered onto the Pyrex glass wafer to create the heater electrode and

31

contact pads. The two wafers were joined by anodic bonding. The wafer stack was

heated to 350 °C and a voltage of 300 V was applied for 15 min. The fluidic

connectors were glued onto the microchip with epoxy.

The final microchip APCI (Fig. 3b) was designed for quantitative work. This was

established by modifying the chip layout to enable simpler sample delivery while

minimizing dead volumes. For introduction of the sample inlet capillary, an inlet

channel was designed from the edge between the contact pads into the vaporizer

channel. A stopper was designed in the middle of the vaporizer channel to bind the

sample inlet capillary (II,III ). This setup required only one through-wafer inlet for the

nebulizer gas, situated behind the stopper. The fluidic network was fabricated using

anisotropic etching in a potassium hydroxide-isopropanol mixture (KOH: 20 wt-%,

IPA: 10% wt-%) at 82 °C (II ) or DRIE with carbon hexafluoride (CF6) (III ). In

anisotropic etching the capillary insertion channel width and depth were 520 µm and

230 µm, respectively. The vaporizer channel width was 300 µm and depth 240 µm.

With DRIE the capillary insertion channel width and depth were 250 µm and 230 µm,

respectively. The vaporizer channel width and depth were 800 µm and 240 µm,

respectively. In the heater and contact pad fabrication, a chromium adhesion

promotion layer (10 nm) was sputtered first, followed by a 100-nm-thick platinum

layer (II,III ). Platinum was sputtered in two steps to reduce thermal cracking. The

resistance of the heater was about 250 Ω. The silicon and Pyrex glass wafers were

joined together by anodic bonding. The wafer stack was heated to 320 °C, and a

voltage of 500 V was applied for 15 min, with maximum current reaching 15 mA. The

fluidic connector for the nebulizer gas and the sample inlet capillary were glued onto

the microchip with epoxy.

sample inlet

nebulizer gas inlets

stopper

nozzle

contact pads

sample inletcapillary

heater

vaporizer channel

a) b)

nebulizer gasinlet

Figure 3. Schematic drawing of the a) prototype and b) final microchip APCI.

32

4.3.2 Microchip-MS interfacing setup

The microchip was mounted on the x,y,z-stage, which allowed easy optimization of

the position of the microchip in front of the MS inlet for the prototype (I ). The

optimal distance between the chip edge and the curtain plate was 0.5-1.0 cm (Fig. 4a).

The final microchip APCI (II,III ) was positioned 1 cm away from the MS orifice in

an orthogonal position (Figs. 4b and 4c). The orthogonal position gave much lower

background, especially in MS measurements. The external corona discharge needle

was a knitting needle .

The nebulizer gas was introduced into the microchip APCI via a 510 µm interal

diameter (i.d.) polyetheretherketone (PEEK™) tubing. The back pressure of the

nebulizer gas was optimized for each sample introduction method and ranged from

0.15 bar to 0.5 bar. The temperature of the microchip APCI was adjusted with an

external power supply and monitored with a commercial thermometer. Temperature

readings were taken at the silicon surface above the vaporizer channel and next to the

inlets.

In the GC/microchip APCI-MS work the metal transfer line from a commercial

GC/MS apparatus was extended with a self-made resistor wire heater that

encapsulated the sample inlet capillary. The metal transfer line, the self-made resistor

wire heater, and the microchip APCI were adjusted to an optimal temperature of 280

°C. For the human urine sample measurement the temperature was raised to 300 °C.

33

0.5 cm

1.0 cmMS

sample inlet

microchip APCI

c)

Figure 4. Microchip APCI-MS interfacing with a) prototype (I) and b) final microchip

APCI (II,III) (here showing the GC interface). c) Schematic drawing of the

orthogonal position used in GC/microchip APCI-MS and capLC/microchip APCI-MS

(II,III).

34

4.3.3 Characterization of the prototype microchip APCI-MS (I)

The performance of the microchip APCI was demonstrated in the analysis of six

compounds. The compounds were ionized in the positive- and negative-ion modes

and detected using selective ion monitoring (SIM) of the protonated molecule

([M+H] +), tandem MS (MS/MS) measurements using multiple reaction monitoring

(MRM) of two product ions, or collecting full-scan mass spectra over a range of m/z

100-500. The ion source and MS parameters were optimized for individual analytes.

The analytes were introduced into the microchip APCI by direct infusion via 50-µm

i.d. PEEK™ tubing. For flow rates higher than 1 µl/min, an LC with an autosampler

combined with a splitter was used. For flow rates lower than 1 µl/min, a microsyringe

pump combined with a loop injector was used. The analyte concentrations ranged

from1 nmol/l to 100 µmol/l. The solvent systems were based on pure methanol or

with mixtures of water-methanol (either 20/80 or 80/20 v/v) + 0.1% acetic acid.

Performance of the prototype microchip APCI with respect to performance of the

heater, signal current dependence on flow rate, stability of the ion current,

repeatability, and linearity was evaluated. The mass spectra and LODs of the analytes

obtained using microchip APCI and commercial APCI were compared.

4.3.4 GC/microchip APCI-MS (II)

The suitability of the microchip APCI-MS combined with GC was tested with a set of

relatively labile volatile organic compounds: acetoacetone, anisole, benzaldehyde, and

2-acetylnaphthalene. Positive ion APCI-MS and APCI-MS/MS methods were

optimized with regard to ion source parameters and fragmentation using direct

infusion. The position of the microchip APCI was optimized with respect to the MS

orifice, the external corona discharge needle position, and the nebulizer gas flow rate

by repeated injections of benzaldehyde and monitoring the signal of the [M+H]+. The

compounds were detected by SIM of the [M+H]+, MRM of two product ions, and

collecting full-scan mass spectra over a range of m/z 30-230.

35

The temperature program was optimized for the separation of the four analytes. A 15

m x 0.25 mm i.d. 5% phenyl-95% dimethylpolysiloxane column was used for

separation and helium as the carrier gas. The samples in hexane were injected

manually with splitless injection. The performance for quantitative work was

evaluated by determining the LODs, linearity of response, and repeatability.

The GC/microchip APCI-MS technique was applied for detection of underivatized

testosterone (TEST) from human urine sample. Sample pretreatment of human urine

sample consisted of enzymatic hydrolysis of steroid glucuronites followed by liquid

extraction at basic pH [223]. Prior to measurement the ethanol solvent was evaporated

to dryness and changed to hexane.

4.3.5 CapLC/microchip APCI-MS (III)

Four selected neurosteroids (dehydroisoandrosterone (DHEA), TEST, progesterone

(PROG), and pregnenolone (PREG)) were used as test compounds. Optimization of

the position of the microchip APCI with respect to the MS orifice was done by direct

infusion of TEST and monitoring the signal of the [M+H]+. The positive ion APCI-

MS/MS method was optimized with regard to the ion source parameters, using direct

infusion.

The capLC separation method was based on the method of Jäntti et al. [224]. The

capLC separations were carried out on a reverse-phase C18-capillary column (0.32

mm i.d. x 100 mm, particle size 3.5 µm) with a flow rate of 10 µl/min and a

water/methanol/acetonitrile-based gradient. A reference LC/APCI-MS method, with

the same gradient composition as the capLC/APCI-MS method and a reverse-phase

C18-column (2.1 mm i.d. x 100 mm, particle size 3.5 µm), was optimized with respect

to sensitivity and selectivity.

The performance of the microchip APCI for capLC/APCI-MS was investigated with

respect to LODs, linearity, and repeatability. The performance of the

36

capLC/microchip APCI-MS was compared with LC/APCI-MS with respect to the

LODs.

The surface of the vaporizer channel of the microchip APCI was deactivated by

simultaneous introduction of undiluted chlorotrimethylsilane with a flow rate of 50

µl/min through the nebulizer inlet and acetonitrile with a flow rate of 500 nl/min

through the sample inlet capillary with the heating power of the integrated heater set

to 1.2 W. Acetonitrile was introduced to prevent clogging of the sample inlet

capillary. After derivatization the microchip was washed with acetonitrile.

4.4 AP-DIOS-MS

The pSi spots were made by electrochemical etching of n-type Si wafers in a 1:1 (v/v)

solution of hydrofluoric acid (HF) and ethanol with simultaneous frontside and

backside illumination. The porous areas were defined with a stencil mask containing

an array of round holes, 1 mm in diameter. After the etching step the pSi sample

plates were rinsed and stored in ethanol. An additional HF treatment was performed

prior to usage to remove the oxide layer. The pSi sample plates were attached to an in-

house-modified AP-MALDI target plate with double-sided conductive tape.

The DIOS-MS measurements were done in both the positive- and negative-ion modes.

The MS instruments were operated with a pulsed nitrogen laser at 337 nm. The laser

pulse energy was either adjusted with an attenuator (IV) or fixed at a defined

attenuator value (V). Sample volumes of 0.4 µl (in units of 0.2 µl) were applied on the

pSi sample areas.

The feasibility of usin AP-DIOS in the analysis of small molecules was investigated

with respect to LODs, reproducibility, and linearity (IV). The compounds were

ionized in the positive- and negative-ion modes and detected by collecting full-scan

mass spectra over a range of m/z 100 to 500 or full scan MS/MS spectra over a range

of m/z 100 - 380. In addition, the ionization processes in the positive-ion mode in AP-

DIOS were more systematically investigated, with 20 compounds and propranolol

serving as external standard (V). The relative intensity of each analyte was calculated

37

from the ratio of the absolute intensity of a 100-µmol/l analyte to the intensity of 100

µmol/l propranolol. The 20 compounds were compared by the relative intensity of the

analyte as a function of its calculated PA value. Hybrid density functional theory

(DFT) calculations were used for the determination of the PA values of the reference

standards using the Gaussian 98 program (Gaussian, Inc., USA). The optimized

geometries for the neutrals and the protonated forms were computed at the B3LYP/6-

31G(d) level.

38

5. RESULTS AND DISCUSSION

The main results of the studies are shortly described in this chapter. Further details on

the experimental conditions, analytical and microfabrication procedures, and the

results can be found in the original publications (I-V).

5.1 Microchip APCI

5.1.1 Microchip characterization (I-III)

The microchip APCI was designed as a stable and durable interface for microfluidics,

GC/APCI-MS, and capLC/APCI-MS. Silicon and glass were chosen as substrates

based on their well-established microfabrication processes and advantage in material

properties, e.g. their excellent durability and high-temperature resistance. In

commercial APCI the sample liquid, nebulizer gas, and auxiliary gas flow

concentrically in three tubes while the microchip APCI is a two-dimensional structure

in which the nebulizer gas flows co-axially relative to the sample flow. The fluidic

network was fabricated on the silicon wafer and the heater on the glass wafer.

a) b)

Figure 5. a) Prototype microchip APCI with fluidic inlets attached. Chip size of 18 x

29 mm was mainly determined by the size of the fluidic connectors. b) The final

microchip APCI (in this study) with the nebulizer gas inlet and sample inlet capillary

present. Size of the microchip is 18 x 10 mm.

39

A prototype (Fig. 5a) was first fabricated with anisotropic silicon wet etching with an

integrated aluminum heater (I). This proof-of-concept showed that the microchip

APCI was capable of flow rates down to 50 nl/min and long-term analysis, and that it

produced highly stable ion current. Two approaches for the final design were

investigated (Fig. 5b). In the first (II), anisotropic silicon wet etching was used for

fabrication of the microchip. The second approach (III) utilized DRIE for fabrication.

Anisotropic wet etchings of silicon are limited by the silicon crystal orientation,

leaving the sidewalls of the channels at a 54.7° angle, and due to the long etching

times corner compensation structures are also required (Fig 6a). DRIE, however, is

independent of the crystal orientation and has excellent dimensional control. DRIE

allowed more precise feature shapes and vertical sidewalls (Fig. 6b). Furthermore,

DRIE had a 10-fold faster etching rate than anisotropic wet etching.

Figure 6. Scanning electron microscope pictures showing the nozzle of the microchip

APCI fabricated by a) wet etching and b) DRIE.

Of the two heater materials, aluminum and platinum, platinum proved superior. With

the aluminum heater the maximum continuous operating temperature was estimated to

be 150 °C. With the platinum heater, temperatures have been tested up to the glass-

softening temperature of about 600 °C [225], but the maximum operational surface

temperature limit, about 300 °C, was set by the epoxy glue used in connecting the

Nanoport™-assemblies. The platinum heater thus provides stable and long-term

operation even at high temperatures. The higher maximum operational surface

temperature of the platinum heater, compared with the aluminum heater, also provides

adequate temperature analysis of semivolatile compounds, as demonstrated in the

40

successful detection of underivatized TEST using GC/MS (II) (Fig. 7). Positioning

the heater away from the vaporizer channel (I), on the other side of the Pyrex glass

wafer and on top of the vaporizer channel (II,III), ensured that the heater was no

longer exposed to liquids and gases in the vaporizer channel, which could degenerate

the heater. The temperature of the microchip APCI can be ramped from room

temperature to 300 °C in less than 1 min [225]. The fast heat transfer of the microchip

APCI was utilized in the chromatographic separation of four neurosteroids (III) with

optimized temperatures for each analyte.

2 4 6 8 10 12 14 16 18 20 22Time (min)

0

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

5500

6000

6500 11.38

10.42

Abs

. int

. (cp

s)

Figure 7. Total ion chromatogram of 1 µl of 1.12 µmol/l underivatized TEST (tR =

11.38 min) measured in the the MRM mode (precursor ion m/z 289.0, product ions

m/z 109.1 and m/z 97.1) from human urine sample (II). (The peak at tR = 10.42 min is

epitestosterone, an epimer of testosterone).

Imaging results from IR thermography measurements of the final design for both wet-

etched and DRIE nozzles show that the nozzle in the microchip forms a very narrow

sample plume (for wet-etched 4 mm and for DRIE 2.3 mm) [225]. The more confined

plume of the DRIE-etched nozzle is due to the more precise feature shapes produced

with DRIE. Since the initial ionization occurs in a very small volume near the needle

tip [226], it can be expected that a confined plume results in larger fractions of ionized

analytes than a wide plume, leading to improved sensitivity. This can be deduced

from the measured LOD of acridine with the microchip APCI and commercial APCI

(I). The volume, in which the initial ionization in corona discharge occurs can be

41

assumed to be similar for both APCI sources, since nearly identical corona discharge

needles are used with the same voltages. The sensitivity in terms of concentration was

comparable to that of commercial APCI but the mass flow sensitivity of the microchip

APCI was about 100-200 hundred times better (Table 5). This means that microchip

APCI can be used with significantly smaller sample volumes than macro-APCI,

without decreasing sensitivity in terms of the concentration in the sample.

Table 5. LOD of acridine with microchip APCI and commercial APCI (I).

Flow rate Limits of detection

[µl/min] Consentration Mass flow

[mol/l] [mol/s]

Microchip APCIa 6.2 5 * 10-9 500 * 10-18

Commercial APCI 6.2 75 * 10-9 14 * 10-15

Commercial APCI 1000 5 * 10-9 80 * 10-15

a prototype microchip APCI

Figure 8. a) Absolute intensity of [M+H]+ of acridine as a function of flow rate. The

injection volume was 1 µl and eluent was a water-methanol (20/80 v/v) solution. b)

Selected ion chromatogram of protonated acridine produced by direct infusion of

water-methanol (80/20 v/v) sample solution with flow rate of 1 µl/min. The

concentration of acridine in both measurements was 100 µmol/l.

The suitability of different flow rates was tested by monitoring the signal of the

[M+H] + of acridine at flow rates of 0.05-2.5 µl/min (I). The injection volume and

concentration were kept constant during the test. Since the intensity of the measured

42

ion current was directly proportional to the flow rate the test also confirmed that

microchip APCI-MS was a mass flow-dependent detector (Fig. 8a).

The durability of the prototype was evaluated by recording the ion current of the

[M+H] + of a sample compound, acridine, by direct infusion at constant flow rate (2.5

µl/min) (I). The prototype produced highly stable ion current and allowed continuous

operation for many hours with no problems (Fig. 8b). The highly stable performance

of the prototype was also reported by Kauppila et al. [55] and Franssila et al. [225].

5.1.2 GC/microchip APCI-MS (II)

The applicability of the GC/microchip APCI-MS for quantitative work was

investigated with four volatile organic compounds: acetoacetone, anisole,

benzaldehyde, and 2-acetylnaphthalene.

The optimized position of the corona discharge needle was in front of the orifice and

as close as possible to the MS orifice without disruptive discharge between the

discharge needle and MS. The position of the microchip APCI towards the corona

discharge needle was crucial for efficient ionization, implying the presence of a very

confined plume. When the position of the microchip APCI was varied only a few

millimeters, the signal strength in MS was extinguished altogether. The temperature

of the microchip APCI was adjusted to the same temperature as the commercial metal

GC transfer line and the optimum distance and nebulizer gas flow rate were 1 cm and

2.5 ml/sec, respectively.

The separation efficiency of the GC/microchip APCI-MS system was excellent, with

narrow and well-shaped peaks, typical of GC separations (Fig. 9). For example,

Ravindra et al. compared low-pressure GC/MS with conventional GC/MS in the

analysis of polycyclic aromatic hydrocarbons. They reported peak widths at half

height for low-pressure GC/MS and conventional GC/MS to be 0.025 min and 0.035

min, respectively (for fluoranthene) [227]. The peak half widths and peak asymmetry

factors of the compounds in our system were 0.02-0.04 min and 1.09-1.20,

43

respectively (Table 6). The dead volumes in the microchip APCI or adsorption to the

microchip walls did not contribute to peak broadening. In addition, no memory effects

were observed.

100

2 3 4 5 6 7 8 9 10 11 12 13Time (min)

4.46

12.57 2-acetylnapht alene

[M+H] m/z=171.2+

h

7.26

anisole

[M+H] m/z=109.2+

6.66

benzaldehyde

[M+H] m/z=107.2+

acetoacetone

[M+H] m/z=101.2+

norm

aliz

ed in

tens

ity

50

0

Figure 9. Overlay of selected ion chromatograms of the [M+H]+ of acetoacetone,

benzaldehyde, anisole, and 2-acetylnaphthalene. The injected sample volume was 1 µl

and the concentrations were 100 µmol/l.

Table 6. Retention times (tR), peak widths at half height (Wh), peak base widths (Wb),

theoretical plates (N), plates per length (N/L), and peak asymmetry (As) calculated

from selected ion chromatograms of the [M+H]+ of the compounds. The

concentrations were 100 µmol/l and injection volume was 1 µl.

Analyte tR Wh Wb N N/L As

[min] [min] [min]

Acetoacetone 4.45 0.03 0.284 118700 7900 1.09

Anisole 6.66 0.022 0.251 512400 34200 1.2

Benzaldehyde 7.24 0.039 0.134 191000 12700 1.2

2-Acetylnaphthalene12.6 0.03 0.504 964400 64300 0.97

44

In addition to the abundant [M+H]+, all the compounds also showed some abundant

fragment ions (Fig. 10). The extent of fragmentation was slightly increased when the

chip temperature was raised. This increased fragmentation of the [M+H]+ gives the

same kind of structural information as in classical medium-pressure CI for structure

elucidation. The exothermicity of the proton transfer reactions could be controlled by

altering the PA of the reactant ions towards the PA of the sample compounds. This

could be accomplished by changing either the nebulizer gas or the carrier gas.

Alternatively, a modifier substance could be introduced into the system alongside the

nebulizer gas.

Figure 10. APCI mass spectrum of a) acetoacetone, b) anisole, c) benzaldehyde, and

d) 2-acetylnaphthalene. The concentrations were 100 µmol/l.

The applicability of the GC/microchip APCI-MS for quantitative work was

investigated with the optimized separation method. LODs (S/N = 3) for anisole,

benzaldehyde, and 2-acetylnaphthalene measured with SIM and MRM of two product

ions were between 0.5-2 µmol/l and 0.05 µmol/l, respectively. The linearity of

response for anisole was measured using MRM of two product ions in the range of

0.2-100 µmol/l. Good linearity with a regression coefficient (r2) of area 0.9989 was

observed. The repeatability was evaluated from six anisole samples with

45

concentrations of 2 µmol/l and measured with MRM of two product ions. The RSD (n

= 6) of the peak area was found to be 4.4%, indicating good repeatability. The LODs

are too early to compare between commercial GC/EI-MS and GC/microchip APCI-

MS, since a relatively old Sciex 300-mass spectrometer was used. The comparison

may reflect more the sensitivity of mass spectrometer itself than ionization efficiency

of EI and microchip APCI.

The system was finally tested with a human urine sample, which was originally

prepared for LC/MS/MS analysis of steroids. The TEST concentration of the sample

was 1.12 µmol/l (corresponding to 48.66 ng/ml). The concentration of TEST is close

to the minimum required performance limit (MRPL; 10 ng/ml for most steroids) with

normal GC-EI-MS set by World Anti-Doping Agency (WADA). Prior to

GC/microchip APCI-MS/MS measurement the ethanol solvent was evaporated to

dryness and changed to hexane. A new separation method and the temperatures of the

metal transfer line, self-made resistor wire heater, and the microchip APCI were

optimized for TEST. The detection of underivatized TEST in a complex biological

matrix (Fig. 7) shows the potential of the GC/API-MS system in bioanalysis.

5.1.3 CapLC/microchip APCI-MS (III)

The feasibility of the microchip APCI for capLC/APCI-MS was demonstrated with

four selected neurosteroids: TEST, DHEA, PROG, and PREG. A separation method

facilitating the optimal temperature conditions for each neurosteroid was developed

and the optimized method was used to evaluated the capLC/microchip APCI-MS

performance for quantitative analysis. The LODs acquired with capLC/microchip

APCI-MS were compared with those LODs acquired with LC/APCI-MS with a

commercial APCI source.

Taking advantage of the interfacing experiments described in GC/microchip APCI-

MS chapter (5.1.2), a similar interfacing was setup: the corona discharge needle was

mounted in front of the orifice and optimization of the microchip APCI with respect

to the corona discharge needle was done with direct infusion of TEST by monitoring

46

the signal intensity of the [M+H]+. The nebulizer gas back pressure was optimized to

a value of 0.35 bar.

A gradient system of water (soluent A) and water/methanol/acetonitrile (1/3/6 v/v/v)

(soluent B) in which the proportion of soluent B increased from 20% to 100% (v/v)

within 12 min provided best separation and ionization of the compounds. The capLC

runs were performed at a flow rate of 10 µl/min, a limit set by the capLC back

pressure. The microchip APCI temperature optimization was done by repeated

injections with increasing heating power and monitoring the MRM signal intensity of

the individual ion pairs.

In contrast to the GC/microchip APCI-MS experiment, adsorption of sample

compounds to the microchip APCI was observed. To minimize adsorptions and

chromatographic peak broadening the surface of the vaporizer channel of the

microchip APCI was deactivated with chlorotrimethylsilane (Fig. 11).

Figure 11. Merged MRM ion chromatogram (m/z 289.1 → 97.2 and m/z 289.1 →

109.2) of TEST with (a) undeactivated microchip APCI and (b) deactivated microchip

APCI. The concentrations were 100 nmol/l.

The separation efficiency of capLC/microchip APCI-MS (Fig. 12) was comparable to

that of capLC/ESI-MS [228] with peak widths at half height of 0.1-0.12 min and peak

base widths of 0.3-0.4 min. While the estimated dead volume of the DRIE-etched

microchip APCI was 1.5 µl, it did not, however, contribute to peak broadening. Since

the nebulizer gas flow rate was about 2.5 ml/s the entire vaporizer channel was

flushed in less than 1 ms. For quantitative analysis the capLC/microchip APCI-MS

47

provides good linearity with r2 value of the area varied between 0.9919 and 0.9994

and adequate repeatability with RSD (n = 6) varied between 12.7% and 18.9% (Table

8). LODs with an S/N of 3 were between 10 and 500 nmol/l. In comparison to a

commercial APCI source with the same MS and equal injection volume, the LODs

were about 10 times higher (Table 7) with the microchip APCI.

Figure 12. Overlay of extracted ion chromatograms of a) TEST and PROG and b)

DHEA and PREG. The concentrations were 1 µmol/l for TEST, PREG, and PROG,

and 10 µmol/l for DHEA. The second peak in the extracted ion chromatogram of

TEST is from DHEA.

Table 7. Linearity of response and repeatability of the capLC/microchip APCI-MS

system, and LODs (S/N=3) measured both with capLC/microchip APCI-MS and

LC/APCI-MS with a commercial APCI source. Linearity of response was measured

using MRM in the range of 200 nmol/l to 20 µmol/l for TEST, PREG, and PROG,

and 2 µmol/l to 200 µmol/l for DHEA. Repeatability was measured in MRM mode

with 10 µmol/l DHEA and 1 µmol/1 TEST, PREG and PROG, and an injection

volume of 2 µl.

Analyte Linearity of responseRepeatability Limits of detection (S/N=3)

(r2) (STD%, n=6) Microchip APCI Commercial APCI

TEST 0.9994 12.70% 0.01 µmol/l (20 fmola) 0.2 µmol/l (400 fmola)

DHEA 0.9946 14.90% 0.5 µmol/l (1 pmola) 5 µmol/l (10 pmola)

PROG 0.9928 15.70% 0.02 µmol/l (40 fmola) 0.2 µmol/l (400 fmola)

PREG 0.9919 18.90% 0.15µmol/l (300 fmola) 2 µmol/l (4 pmola)a injected amount.

48

5.2 AP-DIOS-MS

With the introduction of the AP-MALDI interface [228] DIOS under AP could also be

carried out. The main advantage of AP-DIOS relative to conventional vacuum DIOS

is that delay between the sample plate introduction to the ion source and the

measurement is minimized, since the sample plates are not transferred to vacuum.

This easier sample handling in API-DIOS enables higher sample throughput. In

addition, the AP-MALDI interface provides easier interfacing to existing API

instruments, including tandem instruments such as triple quadrupoles and ion traps.

DIOS operated using AP techniques are attractive to analytical laboratories requiring

high sample throughput, because the analyses are extremely rapid and are easily

automated.

The suitability of DIOS under AP for the analysis of small molecules was investigated

with unmodified n-type pSi. The MS operating parameters were optimized to

maximize the abundance of the [M+H]+. With AP-DIOS the signal was very stable

and lasted for at least 10 min in the ion trap instrument and about 5 minutes in the

quadrupole/time-of-flight instrument. This is in contrast to our earlier results with

vacuum DIOS [204], in which the signal lasted for only tens of seconds. A similar

long-lasting signal was reported by Pihlainen et al. [212]. This is because the

compounds are desorbed by the laser pulses from the pSi spot to the gas phase more

easily in a vacuum than under AP. The long-lasting signal is advantageous; e.g. in

reaction monitoring (required that the signal is stable) or structure characterization,

allowing operation of the instrument in different modes (negative- , positive-ion

mode, MS, and MS/MS measurements) with long accumulation times.

The mass spectra produced by the different AP instruments were very similar and

comparable to the mass spectra produced by vacuum DIOS (Fig. 13). Most

compounds showed an abundant [M+H]+ and the mass spectra were essentially clean

of background peaks at low mass ranges.

49

Figure 13. Mass spectrum of midazolam with a) ion trap instrument, b)

quadrupole/time-of-flight instrument, and c) vacuum time-of-flight instrument. The

concentration of midazolam were a) 13 pmol, b) 400 pmol, and c) 300 pmol. Fig 13c

reprinted from [204], with permission of The Royal Society of Chemistry.

5.2.1 Effect of the PA of an analyte on ionization in AP-DIOS-MS (V)

To further investigate the effect of PA on ionization in AP-DIOS-MS a systematic

study was performed using unmodified n-type pSi [204] with a set of compounds with

fixed ionization, laser, and MS parameters. The compounds studied included various

aromatic molecules of different sizes with different functionalities to ensure

variability in PA values (Table 8). The amount of sample mounted (400 pmol) on the

50

pSi area was selected such that good-quality mass spectra were obtained for

compounds that ionized easily, but at the same time was low enough to ensure that

direct laser DI was unlikely. A reference standard, propranolol, was used as an

external standard. It was applied to spots different from analytes on the same pSi

sample plate and measured after each sample measurement. The relative intensity of

each analyte was calculated from the ratio of the absolute intensity of a 100-µmol/l

analyte to the intensity of a 100-µmol/l external standard. The 20 compounds were

compared using the relative intensity of the analyte as a function of its calculated PA

value (Table 8).

Since the molecules studied included many compounds with unknown PA, the ab

initio and hybrid density functional calculations were performed to define the values

(Table 8). The lowest energy structures and the preferred protonation sites were first

determined using the HF/3-21G level of theory. The final geometry optimizations and

energy calculations of the most stable protonated and neutral molecules were carried

out using hybrid density functional theory calculations at the B3LYP/6-31G(d) level

of theory.

51

Table 8. Literature and calculated thermochemical values of the compounds.

Relative Calculated Reported

intensity PAb PAc

Compound vs propranolola [kJ/mol] [kJ/mol] Detected as

9-Aminoacridine 4,6 1045 - [M+H] +

Benzo[h]quinoline 4,6 972 973d [M+H] +

Verapamil 2,3 980 - [M+H] +

Midazolam 1,8 1002 - [M+H] +

Propranolol 1,0 987 - [M+H] +

4-Acridinol 0,9 985 - [M+H] +

Fluorescein 0,6 1014 - [M+H] +

Testosterone 0,3 925 - [M+H] +

9-Acridinecarboxylic acid 0,2 985 - [M+H] +

1-(Methylamine)-naphthalene 0 951 - [M+H - NH3]+

Dopamine - 940 - -

Anthracene - 911 877 -

9-Methyl-9-anthracenecarboxylate 0 896 - [M] +

Acenaphthene - 888 852 -

4-Bromo-1H-Pyrazole - 880 - -

4-Chloroaniline - 874 874 -

9H-Carbazole 0 870 - [M+M+H - H2]+

2-Naphthalenecarboxylic acid - 856 - -

Naphthalene - 841 803 -

2-Naphthalenol - 763 - -

a Signal intensity of a compound compared with the signal intensity of propranololb B3LYP/6-31G(d)+ZPE c [229]d PA of acridine

All compounds with calculated PA higher than 950 kJ/mol were ionized well, and

most of the compounds with a PA above 970 kJ/mol produced a very clean mass

spectrum with an abundant [M+H]+ (Fig. 14). The lower-than-expected ionization

efficiency of fluorescein, based on the calculated PA (Table 9), may be due to the

tautomeric equilibrium of fluorescein [230]. Only one of the three tautomeric

structures of fluorescein has the lactone ring and the conjugated double bonds; in two

other structures the lactone ring opens to form monoanion and dianion structures. In a

water-methanol solution, 46% of the fluorescein exists in lactone form, with the rest

in monoanion (48%) and dianion (6%) forms, which explains the low ionization

efficiency of fluorescein [231]. Protonation was also favored for 9-acridinecarboxylic

acid (Fig. 14i) and for 1-(methylamine)-naphthalene (Fig. 14j), but their [M+H]+

fragment easily by the neutral loss of formic acid or ammonia, respectively. Only

TEST (Fig. 14h) clearly appeared as [M+H]+ from the compounds having a PA below

52

950 kJ/mol. Dopamine was not observed at all, despite its relatively high PA (940

kJ/mol). However, vacuum DIOS was used to detect dopamine in the positive-ion

mode [221]. 9H-carbazole most likely produced a protonated dimer. Either a fragment

from that dimer or a radical cation was also observed (Fig. 14k). Methyl-9-

anthracenecarboxylate was detected as a radical cation (Fig. 14l).

Figure 14. AP-DIOS-MS spectra of a) 9-aminoacridine, b) benzo[h]quinoline, c)

verapamil, d) midazolam, e) propranolol, f) 4-acridinol, g) fluorescein, h)

testosterone, i) 9-acridinecarboxylic acid, j) 1-(methylamine)-naphthalene, k) 9H-

carbazole, and l) methyl-9-anthracenecarboxylate. The concentrations were 100 µM,

with 0.4 µl deposited on the pSi plate. m) An AP-DIOS-MS spectrum of the

background from a pSi sample plate without sample addition.

53

The high-PA compounds appear to ‘suppress’ the background. For example 9-

aminoacridine (Fig. 14a), benzo[h]quinoline (Fig. 14b), midazolam (Fig. 14d), and

propranolol (Fig. 14e) all produced clean mass spectra, whereas compounds with

lower PA or acid character, e.g. TEST (Fig. 14h) and 4-acridinol (Fig. 14f), had a

significantly increased background. Similar background suppression of high-PA

compounds was observed in MALDI with low matrix-analyte ratios [232,233].

Compounds with a PA higher than 920-950 kJ/mol were ionized efficiently as

[M+H] + in AP-DIOS-MS on unmodified pSi. Observation of the radical cation for

methyl-9-anthracenecarboxylate and the possible radical cation of 9H-carbazole may

indicate that if the ionization energy and PA of a compound is low enough (below

about 7.5 eV [229] and about 900 kJ/mol, respectively), formation of the radical

cation is possible. Radical cations were previously shown with DIOS [203,205] as

well as with other surface-assisted laser DI methods [196,234,235].

5.2.2 Suitability of AP-DIOS-MS for analysis of drugs (IV)

The suitability of AP-DIOS-MS for quantitative work was investigated with respect to

LODs, linearity of response, and repeatability. The LODs with an S/N ratio of 3 for

midazolam and propranolol in the MS mode were 80 fmol and 20 pmol, respectively,

and in the MS/MS mode 10 fmol and 5 pmol, respectively. The LODs for midazolam

and propranolol in the MS mode correspond well with those obtained earlier with

vacuum DIOS using time-of-flight mass spectrometry [204]. For Angiotensin II the

LOD was 1 pmol in the MS mode. AP-DIOS-MS showed good linearity with

dynamic linear ranges of three orders of magnitude. The linearity of the response was

measured for midazolam with the external standard method and for TEST with the

internal standard method in the range of 0.1-160 pmol and 35-555 pmol, respectively.

An r2 value of 0.991 was obtained for midazolam and a value of 0.994 was obtained

for TEST. The use of propranolol as an internal standard was also tested in the

quantitation of midazolam, but it failed to give reasonable results because midazolam

with higher PA suppressed propranolol. The repeatability of the AP-DIOS-MS

method was measured using 20 pmol of midazolam injected onto different pSi spots.

The RSD of the intensity of the [M+H]+ in seven measurements was 35%.

54

6. CONCLUSIONS

New microchip-based interfacing techniques for API-MS were investigated. A novel

miniaturized heated nebulizer for APCI-MS, the microchip APCI, was developed and

characterized. The performance of the microchip APCI was evaluated as an interface

for capLC/MS and GC/MS. Next, the suitability of DIOS-MS under AP for drug

analysis was investigated and the performance of AP-DIOS-MS for quantitative work

was evaluated. In addition, the ionization mechanism in AP-DIOS was further

investigated.

The microchip APCI provided stable, long-term analysis with high sensitivity from

minimal sample volumes. Good separation efficiency, repeatability and linearity for

both the GC/microchip APCI-MS system and the capLC/microchip APCI-MS

systems were obtained. The mass flow sensitivity of the microchip APCI was about

100-200 times better then commercial APCI ion source making it directly compatible

with low-flow rate separation systems, microfluidic systems, as well as with nano-

and capLC, without loss in sensitivity. In fact, the microchip can be used as an

interface to any API-MS equipped with an APCI source without modifications to the

commercial source. Since ionization is initiated with an external corona discharge

needle the ionization process is independent of flow rate. This is in contrast to

microchip ESI-MS where efficient ionization is limited to a narrow flow rate range

determined by the channel and ESI tip geometry. In addition, the current applied to

the discharge needle does not interfere with possible CE separation. The microchip

APCI can be used as a separate interface as demonstrated, or it can be eventually

integrated into a µ-TAS microchip with integrated injection, fluid handling, and

separation.

The AP-DIOS-MS was very similar to vacuum DIOS-MS. The mass spectra were

essentially free of the background peaks at low-MW (< 800 amu). The systematic

study of the effect of the PA of a compound on ionization in AP-DIOS-MS on

unmodified pSi showed that only compounds with a PA higher than 920-950 kJ/mol

55

were ionized efficiently as [M+H]+, suggesting gas-phase ionization. AP-DIOS-MS is

suitable for quantitative work. The main advantage of AP-DIOS over vacuum DIOS is

that AP-MALDI provides easier interfacing with existing API instruments and easier

sample transfer than with vacuum MALDI instruments. Since the sample plate is not

transferred to vacuum AP-DIOS enables even higher sample throughput than

conventional vacuum DIOS. The observed value of 920-950 kJ/mol in PA clearly

excludes nearly all neutral and less polar small organic molecules from AP-DIOS-MS

analysis on unmodified pSi and empathizes the need for surface modifications. AP-

DIOS-MS appears to be best suited for very rapid qualitative screening of polar

compounds, such as in forensics or proteomics.

56

ACKNOWLEDGEMENTS

This study was carried out in the Division of Pharmaceutical Chemistry and in Viikki

Drug Discovery Technology Center (DDTC), Faculty of Pharmacy, University of

Helsinki, during the years 2002-2006.

I am most grateful to my supervisors prof. Risto Kostiainen and prof. Tapio Kotiaho

for this opportunity to work at the Division of Pharmaceutical Chemistry. I thank

them for introducing me to the world of microfluidics, for deeper understanding of

mass spectrometry, but most importantly for their guidance, invaluable advises and

constant support during these years.

I want to thank all co-authors for their contribution to the work. Special thanks go to

Dr. Sami Franssila, Dr. Kestas Grigoras, Dr. Karti Huikko, Dr. Raimo A. Ketola, Dr.

Friedrich Mandel, Seppo Marttila, Dr. Jaana Pakarinen, Ville Saarela, Dr. Christian

Sauber, and Prof. Pirjo Vainiotalo for their contribution to this work. Prof. Jonas

Bergquist and Docent Seppo Auriola are acknowledged for their careful review of this

manuscript and for their valuable comments.

I also wish to thank Markus, Tiina, Piude, Inkku, Tiina, Sirkku, Päivi, Niina, Laura,

Kati, Kata, Kirsi, Mikko and all the other colleagues at the Division of Pharmaceutical

Chemistry. I am deeply grateful to Sami, Kestas, Ville, Santeri and Seppo in the

Micro- and Nanofabrication Centre (Helsinki University of Technology), Marc in the

Protein Chemistry Unit (Helsinki Biomedicum, University of Helsinki), and the

people in Environmental Chemistry and Environmental Analytical Chemistry

(Laboratory of Analytical Chemistry, University of Helsinki) for collaboration. In

addition, I wish to thank Kai Kolari (VTT Center for Microelectronics) for helping

with the DRIE process.

57

Finally, I extend my warmest thank to my wife Paula for her patience and support

during this project. Sincere thanks are also to my parents and parents-in-law for all

their support.

Financial support from the Finnish National Technology Agency TEKES, the

Academy of Finland, the Jenny and Antti Wihuri Foundation, and CHEMSEM

graduate school as well as instrumental support from Agilent Technologies are

gratefully acknowledged.

58

REFERENCES

[1] Terry S.C. and Angell J.B. Des.Biomed. Appl. Solid State Chem. Sens., Workshop (1978), Meeting date 1977, 207.

[2] Manz A., Graber N., and Widmer H.M. Sens. Actuators B (1990) 244. [3] Liu Y., Garcia C.D., and Henry C.S. Analyst 2003, 128, 1002. [4] Dupuy A.M., Lehmann S., and Cristol J.P. Clin. Chem. Lab. Med. 2005 43,

1291. [5] Yi C., Li C.-W., Ji S., and Yang M. Anal. Chim. Acta 2006, 560, 1. [6] Dolnik V. and Liu S. J. Sep. Sci. 2005, 28, 1994. [7] Kelly R.T. and Woolley A.T. Anal. Chem. 2005, 77, 96A. [8] Sun Y and Kwok Y.C. Anal. Chim. Acta 2006, 556, 80. [9] Watts P. and Haswell S.J. Cur. Opp. Chem. Bio. 2003, 7, 380. [10] Dittrich P.S. and Manz A. Nature Reviews 2006, 5, 210. [11] Pihl J., Karlsson M., and Chiu D.T. DDT 2005, 10, 1377. [12] Laurell T., Nilsson J., and Marko-Varga G. Anal. Chem. 2005, 77, 264A. [13] Huikko K., Kostiainen R., Kotiaho T. E. J. Pharm. Scie. 2003, 20, 149. [14] Cooper J.W., Wang Y., and Lee C.S. Electrophoresis 2004, 25, 3913. [15] Marle L. and Greenway G.M. Trends Anal. Chem. 2005, 24, 795. [16] Ramos L., Ramos J.J., and Brinkman U.A.Th. Anal. Bioanal. Chem. 2005,

381, 119. [17] Chena G.,Lin Y., and Wang J. Talanta 2006, 68, 497. [18] Jakobson S.C., Culbertson C.T., Daler J.E., and Ramsey J.M. Anal. Chem.

1998, 70, 3476. [19] Culbertson C.T., Jacobson S.C., and Ramsey J.M. Anal. Chem. 2000, 72,

5814. [20] Yin H., Killeen K., Brennen R., Sobek D., Werlich M., and van de Goor T.

Anal. Chem. 2005, 77, 527.

59

[21] Shintani Y., Hirako K., Motokawa M., Iwano T., Zhou X., Takano Y., Furuno

M., Minakuchi H., and Ueda M. J. Chromatogr. A, 2005, 1073, 17. [22] Fortier M.-H., Bonneil E., Goodley P., and Thibault P. Anal. Chem. 2005, 77,

1631. [23] Lazar I.M. and Karger B.L. Anal. Chem. 2002, 74, 6259. [24] Lazar I.M., Trisiripisal P., and Sarvaiya H.A. Anal. Chem. 2006, 78,5513. [25] Ehrnström, R. Lab Chip 2002, 2, 26N. [26] Gustafsson M., Hirschberg D., Palmberg C., Jörnvall H., and Bergman T.

Anal. Chem. 2004, 76, 345. [27] Hirschberg D., Jägerbrink T., Samskog J., Gustafsson M., Ståhlberg M.,

Alvelius G., Husman B., Carlquist M., Jörnvall H., and Bergman T. Anal. Chem. 2004, 76, 5864.

[28] Duffy D.C., Gillis H.L., Lin J., Sheppard Jr. N.F., and Kellogg G.J. Anal.

Chem. 1999, 71, 4669–4678. [29] Weigl B.H., Bardell R.L., Kesler N., and Morris C.J. Fresenius J. Anal. Chem.

2001, 371, 97. [30] Jandik P., Weigl B.H., Kessler N., Cheng J., Morris C.J., Schulte T., and

Avdalovic N. J. Chromatogr. A 2002, 954, 33. [31] Wilson D.J. and Konermann L. Anal. Chem. 2005, 77, 6887. [32] Yu C., Davey M.H., Svec F., and Fréchet J.M.J. Anal. Chem. 2001, 73, 5088. [33] Bhattacharyya, A. and Klapperich, C. M. Anal. Chem. 2006, 78, 788. [34] Kutter J.P., Jacobson S.C., and Ramsey, J.M. J. Microcolumn Sep. 2000, 12,

93. [35] Oleschuk R.D., Shultz-Lockyear L.L., Ning Y.B., and Harrison D.J. Anal.

Chem. 2000, 72, 585. [36] Ramsey J.D. and Collins G.E. Anal. Chem. 2005, 77, 6664. [37] Jemere A.B., Oleschuk R.D., Ouchen F., Fajuyigbe F., and Harrison D.J.

Electrophoresis 2002, 23, 3537. [38] Chiesl T.N., Shi W., and Barron A.E. Anal. Chem. 2005, 77, 772. [39] Lion N., Gellon J.-O., Jensen H., and Girault H.H. J. Chromatogr. A 2003,

1003, 11.

60

[40] Foote R.S., Khandurina J., Jacobson S.C., and Ramsey J.M. Anal. Chem. 2005,

77, 57. [41] Wang X.Y., Saridara C., and Mitra S. Anal. Chim. Acta 2005, 543, 92. [42] Janasek D., Franzke J., Manz A. Nature 2006, 442, 374-380. [43] Vandaveer IV W.R., Pasas-Farmer S.A., Fischer D.J., Frankenfeld C.N., and

Lunte S.M. Electrophoresis 2004, 25, 3528. [44] Fiaccabrino G.C., de Rooij N.F., and Koudelka-Hep M. Anal. Chim. Acta

1998, 359, 263. [45] Arora A., Eijkel J.C.T., Morf W.E., and Manz A. Anal. Chem. 2001, 73, 3282. [46] Burggraf N., Krattiger B., de Mello A.J., de Rooij N.F., and Manz A. Analyst

1998, 123, 1443. [47] Kameoka J. and Craighead H.G. Sens. Actuators B 2001, 77, 632. [48] Firebaugh S.L., Jensen K.F., and Schmidt M.A. J. Appl. Phys. 2002, 92, 1555. [49] Ro K.W., Lim K., Shim B.C., and Hahn J.H. Anal. Chem. 2005, 77, 5160. [50] Lazar I.M., Grym J., and Foret F. Mass Spectrometry Reviews 2006, 25,573. [51] de Mello A.J. Lab Chip 2001, 1, 7N. [52] Auroux P.-A., Iossifidis D., Reyes D.R., and Manz A. Anal. Chem. 2002, 74,

2637. [53] Limbach P.A. and Meng Z. Analyst 2002, 127,693. [54] Sung W.-C., Makamba H., and Chen S.-H. Electrophoresis 2005, 26, 1783. [55] Kauppila T.J., Östman P., Marttila S., Ketola R.A., Kotiaho T., Franssila S.,

and Kostiainen R. Anal. Chem. 2004, 76, 6797. [56] Manz A., Harrison D.J., Verpoorte E.M.J., Fettinger J.C., Paulus A., Lüdi H.,

and Widmer H.M. J. Chromatogr. A 1992, 593, 253. [57] de Mello A.J. Lab Chip 2002, 2, 31N. [58] Fintschenko Y. and van den Berg A. J. Chromatogr. A 1998, 819, 3. [59] Madou M. Fundamentals of Microfabrication, CRC Press, New York, 1997. [60] Elwenspoek M.C. and Jansen H.V. Silicon Micromachining, Cambridge

University Press, Cambridge, 1998.

61

[61] Franssila S. Introduction to Microfabrication, John Wiley & Sons, Chichester,

2004. [62] Dario P., Carrozza M.C., Croce N., Montesi M.C., and Cocco M. J.

Micromech. Microeng. 1995, 5, 64. [63] Diepold T. and Obermeier E. J. Micromech. Microeng. 1996, 6, 29. [64] Anthony T. R. J. Appl. Phys. 1981, 52, 5340. [65] Li X., Abe T., and Esashi M. Sens. Actuators A 2001, 87, 139. [66] Li X., Abe T., Liu Y., and Esashi M. J. MEMS 2002, 11, 625. [67] Ichiki T., Sugiyama Y., Taura R., Koidesawa T., and Horiike Y. Thin Solid

Films 2003, 435, 62. [68] Park J.H., Lee N.-E., Lee J., Park J.S., and Park H.D. Microelectronic

Engineering 2005, 82, 119. [69] Becker H. and Gärtner C. Electrophoresis 2000, 21, 12. [70] Kirby B.J. and Hasselbrink E.F. Jr. Electrophoresis 2004, 25, 203. [71] Heuschkela M.O., Guérina L., Buissonb B., Bertrandb D., and Renauda P.

Sens. Actuators B 1998, 48, 356. [72] Tuomikoski S. and Franssila S. Sens. Actuators A 2005, 120, 408. [73] Dole M., Mack L.L., Hines R.L., Mobley R.C., Ferguson L.D., and Alice M.B.

J. Chem. Phys. 1968, 49, 2240. [74] Yamashita M. and Fenn J.B. J. Phys. Chem. 1984, 88, 4451. [75] Yamashita M. and Fenn J.B. J. Phys. Chem. 1984, 88, 4671. [76] Mann M., Meng C.K., and Fenn, J.B. Anal. Chem. 1989, 61, 1702. [77] Fenn J.B., Mann M., Meng C.K., Wong S.F., and Whitehouse C.M., Science

1989, 246, 64. [78] Wilm M. and Mann M. Anal. Chem. 1996, 68, 1. [79] Xue Q., Foret F., Dunayevskiy Y.M., Zavracky P.M., McGruer N.E., and

Karger B.L. Anal. Chem. 1997, 69, 426. [80] Ramsey R.S. and Ramsey J.M. Anal. Chem. 1997, 69, 1174. [81] Figeys D., Ning Y., and Aebersold R. Anal. Chem. 1997, 69, 3153.

62

[82] Figeys D., Gygi S.P., McKinnon G., and Aebersold R. Anal. Chem. 1998, 70,

3728. [83] Xiang F., Lin Y., Wen J., Matson D.W., and Smith R.D. Anal. Chem. 1999,

71, 1485. [84] Li J., Thibault P., Bings N.H., Skinner C.D., Wang C., Colyer C., and Harrison

J. Anal. Chem. 1999, 71, 3036. [85] Zhang B., Liu H., Karger B.L., and Foret F. Anal. Chem. 1999, 71, 3258. [86] Bings N.H., Wang C., Skinner C.D., Colyer C.L., Thibault P., and Harrison

D.J. Anal. Chem. 1999, 71, 3292. [87] Lazar I.M., Ramsey R.S., Sundberg S., and Ramsey J.M. Anal. Chem. 1999,

71, 3627. [88] Chan J.H., Timperman A.T., Qin D., and Aebersold R. Anal. Chem. 1999, 71,

4437. [89] Licklider L., Wang X.-Q., Desai A., Tai Y.-C., and Lee T.D. Anal. Chem.

2000, 72, 367. [90] Li J., Kelly J.F., Chernushevich I., Harrison D.J., and Thibault P. Anal. Chem.

2000, 72, 599. [91] Zhang B., Foret F., and Karger B.L. Anal. Chem. 2000, 72, 1015. [92] Schultz G.A., Corso T.N., Prosser S.J., and Zhang S. Anal. Chem. 2000, 72,

4058. [93] Pinto D.M., Ning Y., and Figeys D. Electrophoresis 2000, 21, 181. [94] Wen J., Lin Y., Xiang F., Matson D.W., Udseth H.R., and Smith R.D.

Electrophoresis 2000, 21, 191. [95] Lazar I.M., Ramsey R.S., Jacobson S.C., Foote R.S., and Ramsey J.M. J.

Chromatogr. A 2000, 892, 195. [96] Yuan C.-H. and Shiea J. Anal. Chem. 2001, 73, 1080. [97] Deng Y., Henion J., Li J., Thibault P., Wang C., and Harrison D.J. Anal.

Chem. 2001, 73, 639. [98] Deng Y., Zhang H., and Henion J. Anal. Chem. 2001, 73, 1432. [99] Lazar I.M., Ramsey R.S., and Ramsey J.M. Anal. Chem. 2001, 73, 1733.

63

[100] Kameoka J., Craighead H.G., Zhang H., and Henion J. Anal. Chem. 2001, 73, 1935.

[101] Zhang B., Foret F., and Karger B.L. Anal. Chem. 2001, 73, 2675. [102] Rohner T.C., Rossier J.S., and Girault H.H. Anal. Chem. 2001, 73, 5353. [103] Kim J.-S. and Knapp D.R. J. Am. Soc. Mass Spectrom. 2001, 12, 463. [104] Kameoka J., Orth R., Ilic B., Czaplewski D., Wachs T., and Craighead H.G.

Anal. Chem. 2002, 74, 5897. [105] Chiou C.-H., Lee G.-B., Hsu H.-T., Chen P.-W., and Liao P.-C. Sens.

Actuators B 2002, 86, 280. [106] Svedberg M., Pettersson A., Nilsson S., Bergquist J., Nyholm L., Nikolajeff

F., and Markides K. Anal. Chem. 2003, 75, 3934. [107] Le Gac S., Arscott S., and Rolando C. Electrophoresis 2003, 24, 3640. [108] Huikko K., Östman P., Grigoras K., Tuomikoski S., Tiainen V.M., Soininen

A., Puolanne K., Manz A., Franssila S., Kostiainen R., and Kotiaho T. Lab Chip 2003, 3, 67.

[109] Sjödahl J.,Melin J., Griss P., Emmer Å., Stemme G., and Roeraade J. Rapid

Commun. Mass Spectrom. 2003, 17, 337. [110] Ssenyange S., Taylor J., Harrison D.J., and McDermott M.T. Anal. Chem.

2004, 76, 2393. [111] Arscott S., Le Gac S., Druon C., Tabourier P., and Rolando C. Sens. Actuators

B 2004, 98, 140. [112] Dahlin A.P., Bergström S.K., Andrén P.E., Markides K.E., and Bergquist J.

Anal. Chem. 2005, 77, 5356. [113] Thorslund S., Lindberg P., Andrén P.E., Nikolajeff F., and Bergquist J.

Electrophoresis 2005, 26, 4674. [114] Tuomikoski S., Sikanen T., Ketola R.A., Kostiainen R., Kotiaho T., and

Franssila S. Electrophoresis 2005, 26, 4691. [115] Carlier J., Arscott S., Thomy V., Camart J.-C., Cren-Olivé C., and Le Gac S. J.

Chromatogr. A 2005, 1071, 213. [116] Bedair M.F. and Oleschuk R.D. Anal. Chem. 2006, 78, 1130. [117] Lindberg P., Dahlin A.P., Bergström S.K., Thorslund S., Andrén P.E.,

Nikolajeff F., and Bergquist J. Electrophoresis 2006, 27, 2075.

64

[118] Koerner T. and Oleschuk R.D. Rapid Commun. Mass Spectrom. 2005, 19, 3279.

[119] Wachs T. and Henion J. Anal. Chem. 2001, 73, 632. [120] Tang K., Lin Y., Matson D.W., Kim T., and Smith R.D. Anal. Chem. 2001, 73,

1658. [121] Oleschuk R.D. and Harrison D.J. Trends Anal. Chem. 2000, 19, 379. [122] Horning E.C., Horning M.G., Carroll D.I., Dzidic I., Haegele K.D., and

Stillwell R.N. Anal. Chem. 1973, 45, 936. [123] Carroll D.I., Dzidic I., Stillwell N., Haegele K.D., and Horning E.C. Anal.

Chem. 1975, 47, 2369. [124] Robb D.B., Covey T.R., and Bruins A.P. Anal. Chem. 2000, 72, 3653. [125] Syage J.A., Evans M.D., and Hanold K.A. Am. Lab. 2000, 32, 24. [126] Sunner J., Ikonomou M.G., and Kebarle P. Anal. Chem. 1988, 60, 1300. [127] King R., Bonfiglio R., Fernandez-Metzler C., Miller-Stein C., and Olah T. J.

Am. Soc. Mass Spectrom. 2000, 11, 942. [128] Reemtsma T. J. Chromatogr. A 2003, 1000, 477. [129] Hayen H. and Karst U. J. Chromatogr. A 2003, 1000, 549. [130] Zwiener C. and Frimmel F.H. Anal. Bioanal. Chem. 2004, 378, 862. [131] Lee H. J. Liq. Chromatogr. & Rel. Technol. 2005, 28, 1161. [132] Chen J., Korfmacher W.A., and Hsieh Y. J. Chromatogr. B 2005, 820, 1. [133] Karst U. Anal. Bioanal. Chem. 2005, 382, 1744. [134] Sforza S., Dall’Asta C., and Marchelli R. Mass Spectrometry Reviews 2006,

25, 54. [135] Manini P., Andreoli R., and Mutti A. Toxicology Letters 2006, 162, 202. [136] Bos S.J., van Leeuwen S.M., and Karst U. Anal. Bioanal. Chem. 2006, 384,

85. [137] Shahin M.M. J. Chem. Phys. 1966, 45, 2600. [138] Siegel M.W. and Fite W.L. J. Phys. Chem. 1976, 80, 2871. [139] Kambara H., Mitsui Y., and Kanomata I. Anal. Chem. 1979, 51, 1447.

65

[140] Kambara H., Ogawa Y., Mitsui Y., and Kanomata I. Anal. Chem. 1980, 52,

1500. [141] Matsuoka S., Nakamura H., and Tamura T. J. Chem. Phys. 1981, 75, 681. [142] Mitsui Y., Kambara H., Kojima M., Tomita H., Katoh K., and Satoh K. Anal.

Chem. 1983, 55, 477. [143] Matsuoka S., Nakamura H., and Tamura T. J. Chem. Phys. 1983, 79, 825. [144] Foster K.L., Cadwell T.E., Benter T., Langer S., Hemminger J.C., and

Finlayson-Pitts B.J. Phys. Chem. Chem. Phys. 1999, 1, 5615. [145] Kinoue T., Asai S., Ishii Y., Ishikawa K., Fujii M., Nakano K., and Hasumi K.

Environ. Health Prev. Med. 2003, 5, 97. [146] Taylor A.J., Linforth R.S.T., Harvey B.A., and Blake A. Food Chemistry

2000, 71, 327. [147] Deibler K.D., Lavin E.H., Linforth R.S.T., Taylor A.J., and Acree T.E. J.

Agric. Food Chem. 2001, 49, 1388. [148] Hollowood T.A., Linforth R.S.T., and Taylor A. J. Chem. Senses. 2002, 27,

583. [149] Weel K.G.C., Boelrijk A.E.M., Alting A.C., Van Mil P.J.J.M, Burger J.J.,

Gruppen H., Voragen A.G.J., and Smit G. J. Agric. Food Chem. 2002, 50, 5149.

[150] Turner J.A., Linforth R.S.T., and Taylor A.J. J. Agric. Food Chem. 2002, 50,

5400. [151] Hodgson M., Linforth R.S.T., and Taylor A.J. J. Agric. Food Chem. 2003, 51,

5052. [152] Lethuaut L., Weel K.G.C., Boelrijk A.E.M., and Brossard C.D. J. Agric. Food

Chem. 2004, 52, 3478. [153] Weel K.G.C., Boelrijk A.E.M., Burger, J.J., Verschueren M., Gruppen H.,

Voragen A.G.J., and Smit G. J. Agric. Food Chem. 2004, 52, 6564. [154] Garratt L.C., Linforth R., Taylor A.J., Lowe K.C., Power J.B., and Davey

M.R. Plant Biotechnology Journal 2005, 3, 165. [155] Mitchum R.K., Moler G.F., and Korfmacher W.A. Anal. Chem. 1980, 52,

2278. [156] Mitchum R.K., Korfmacher W.A., and Moler G.F. Anal. Chem. 1982, 54, 719.

66

[157] Korfmacher W.A., Moler G.F., Delongchamp R.R., Mitchum R.K., and Harless R.L. Chemosphere 1984, 13, 669.

[158] Korfmacher W.A., Mitchum R.K., Hileman F.D., and Mazer T. Chemosphere

1983, 12, 1243. [159] Korfmacher W.A., Fu P.P., Chou M., and Mitchum R.K. HRC & CC 1984, 7,

41. [160] Korfmacher W.A. and Miller D.W. HRC & CC 1984, 7, 581. [161] Korfmacher W.A. amd Rushing L.G. HRC & CC 1986, 9, 293. [162] Korfmacher W.A., Rushing L.G., Engelbach R.J., Freeman J.P., Djuric Z.,

Fifer E.K., and Beland F.A. HRC & CC 1987, 10, 43. [163] Korfmacher W.A., Rushing L.G., Arey J., Zielinska B., and Pitts J.N., Jr. HRC

& CC 1987, 10, 641. [164] Engelbach R.J., Korfmacher W.A., and Rushing L.G. HRC & CC 1988, 11,

661. [165] Kinouchi T., Lopez M.A.T., Rushing L.G., Beland F.A., and Korfmacher

W.A. HRC 1990, 13, 281. [166] Korfmacher W.A., Rushing L.G., Siitonen P.H., Branscomb C.J., and Holder

C.L. HRC & CC 1987, 10, 332. [167] Revel'skii I.A., Yashin Y.S., Voznesenskii V.N., Kuroschkin V.K.,

Kostyanovskii R.G. Izv. Akad. Nauk SSSR, Ser. Khim. 1986, 9, 1987. [168] Revel'skii I.A., Yashin Y.S., Kuroschkin V.K., Kostyanovskii R.G.

Zavodskaya Laboratoriya 1991, 57, 1. [169] Tyrefors L.N., Moulder R.X., and Markides K.E. Anal. Chem. 1993, 65, 2835. [170] Nyholm L.M., Sjöberg P.J.R., and Markides K.E. J. Chromatogr. A 1996, 755,

153. [171] Tanaka Y., Otsuka K., and Terabe S. J. Pharm. Biomed. Anal. 2003, 30, 1889. [172] Nilsson S.L., Andersson C., Sjöberg P.J.R., Bylund D., Petersson P., Jörntén-

Karlsson M., and Markides K.E. Rapid Commun. Mass Spectrom. 2003, 17, 2267.

[173] Karas M., Bachmann D., and Hillenkamp F. Anal. Chem. 1985, 57, 2935. [174] Knochenmuss R. Analyst 2006, 131, 966.

67

[175] Knochenmuss R., Dubois F., Dale M., and Zenobi R. Rapid Commun. Mass Spectrom. 1996, 10, 871.

[176] Knochenmuss R., Karbach V., Wiesli U., Breuker K., and Zenobi R. Rapid

Commun. Mass Spectrom. 1998, 12, 529. [177] Santos L.S., Haddad R., Höehr N.F., Pilli R.A., and Eberlin M.N. Anal. Chem.

2004, 76, 2144. [178] Tholey A. and Heinzle E. Anal. Bioanal. Chem. 2006, 386, 24. [179] Önnerfjord P, Nilsson J, Wallman L, Laurell T, and Marko-Varga G. Anal.

Chem. 1998, 70, 4755. [180] Ekström S., Önnerfjord P., Nilsson J., Bengtsson M., Laurell T., and Marko-

Varga G. Anal. Chem. 2000, 72, 286–293. [181] Ekström S., Ericsson D., Önnerfjord P., Bengtsson M., Nilsson J., Marko-

Varga G., and Laurell T. Anal. Chem. 2001, 73, 214. [182] Marko-Varga G., Ekström S., Helldin G., Nilsson J., and Laurell T.

Electrophoresis 2001, 22, 3978. [183] Ekström S., Nilsson J., Helldin G., Laurell T., and Marko-Varga G.

Electrophoresis 2001, 22, 3984. [184] Liu J., Tseng K., Garcia B., Lebrilla C.B., Mukerjee E., Collins S., and Smith

R. Anal. Chem. 2001, 73, 2147. [185] Mok M.L., Hua L., Phua J.B., Wee M.K., and Sze N.S. Analyst 2004, 129,

109. [186] Brivio M., Fokkens R.H., Verboom W., Reinhoudt D.N., Tas N.R., Goedbloed

M., and van den Berg A. Anal. Chem. 2002, 74, 3972. [187] Ehrnström R. Lab Chip 2002, 2, 26N–30N. [188] Wei J., Buriak J.M., and Siuzdak G. Nature 1999, 399, 243. [189] Kruse R.A., Li X., Bohn P.W., and Sweedler J.V. Anal. Chem. 2001, 73, 3639. [190] Shen Z., Thomas J.J., Averbuj C., Broo K.M., Engerlhard M., Crowell J.E.,

Finn M.G., and Siuzdak G. Anal. Chem. 2001, 73, 612. [191] Kruse R.A., Rubakhin S.S., Romanova E.V., Bohn P.W., and Sweedler J.V. J.

Mass Spectrom. 2001, 36, 1317. [192] Go E.P., Shen Z., Harris K., and Siuzdak G. Anal. Chem. 2003, 75, 5475. [193] Canham, L.T. Appl. Phys. Lett. 1990, 57, 1046.

68

[194] Meng J-C., Siudzak G., and Finn M.G. Chem. Commun. 2004, 2108. [195] Thomas J.J., Shen Z., Blackledge R., and Siuzdak G. Anal. Chim. Acta 2001,

442, 183. [196] Rousell D.J., Dutta S.M., Little M.W., and Murray K.K. J. Mass Spectrom.

2004, 39,1182. [197] Bhattacharya, S.H., Raiford T.J., and Murray K.K. Anal. Chem. 2002, 74,

2228. [198] Alimpiev S., Nikiforov S., Karavanskii V., Minton T., and Sunner J. J Chem.

Phys. 2001, 115, 1891. [199] Cuiffi J.D., Hayes D.J., Fonash S.J., Brown K.N., and Jones A.D. Anal. Chem.

2001, 73, 1292. [200] Finkel, N.H., Prevo B.G., Velev O.D., and He, L. Anal. Chem. 2005, 77, 1088. [201] Go E.P., Apon J.V, Luo G., Saghatelian A., Daniels R.H., Sahi V., Dubrow R.,

Cravatt B.F., Vertes A., and Siuzdak G. Anal. Chem. 2005, 77, 1641. [202] Okuna S., Arakawa R., Okamoto K., Matsui Y., Seki S., Kozawa T., Tagawa

S., and Wada Y. Anal. Chem. 2005, 77, 5364. [203] Okuno S., Nakano M., Matsubayashi G., Arakawa R., and Wada Y. Rapid

Commun. Mass Spectrom. 2004, 18, 2811. [204] Tuomikoski S., Huikko K., Grigoras K., Östman P., Kostiainen R., Baumann

M., Abian J., Kotiaho T., and Franssila S. Lab Chip 2002, 2, 247. [205] Li Q., Winefordner J.D., and Powell DH. Proc. 53rd Annual Conference, Am.

Soc. for Mass Spectrom., San Antonio, Texas, June 5-9, 2005. [206] Budimir N., Blais J.-C., Fournier F., and Tabet J.-C. Rapid Commun. Mass

Spectrom. 2006, 20, 680. [207] Trauger S.A., Go E.P., Shen Z., Apon J. V., Compton B.J., Bouvier E.S.P.,

Finn M.G., and Siuzdak G. Anal. Chem. 2004, 76, 4484. [208] Thomas J.J., Shen Z., Crowell J.E., Finn M.G., and Siuzdak G. Proc. Natl.

Acad. Sci. 2001, 98, 4932. [209] Laiko V.V., Taranenko N.I., Berkout V.D., Musselman B.D., and Doroshenko

V.M. Rapid Commun. Mass Spectrom. 2002, 16, 1737. [210] Wall D.B., Finch J.W., and Cohen S.A. Rapid Commun. Mass Spectrom. 2004,

18, 1403.

69

[211] Go E.P., Prenni J.E., Wei J., Jones A., Hall S.C., Witkowska H.E., Shen Z., and Siuzdak G. Anal. Chem. 2003, 75, 2504.

[212] Pihlainen K., Grigoras K., Franssila S., Ketola R.,Kotiaho T., and Kostiainen

R. J. Mass Spectrom. 2005, 40, 539. [213] Lewis W.G., Shen Z., Finn M.G., and Siuzdak G. Int. J. Mass Spectrom. 2003,

226, 107. [214] Nordström A., Apon, J.V., Uritboonthai W., Go, E.P., and Siuzdak G. Anal.

Chem. 2006, 78, 272. [215] Okuno S., Wada Y., and Arakawa R. Int. J. Mass. Spectrom. 2005, 241, 43. [216] Xu S., Pan C., Hu L., Zhang Y., Guo Z., Li X., and Zou H. Electrophoresis

2004, 25, 3669. [217] Wall D.B., Finch J.W., and Cohen S.A. Rapid Commun. Mass Spectrom. 2004,

18, 1482. [218] Prenni J.E., Shen Z., Trauger S., Chen W., and Siuzdak G. Spectroscopy 2003

17, 693. [219] Compton B.J. and Siuzdak G. Spectroscopy 2003, 17, 699. [220] Li Q., Ricardo A., Benner S.A., Winefordner J.D., and Powell D.H. Anal.

Chem. 2005, 77, 4503 [221] Kraj A., Dylag T., Gorecka-Drzazga A., Bargiel S., Dziuban J., and Silberring

J. Acta Biochimica Polonica 2003, 50, 783. [222] Kinumi T., Shimomae Y., Arakawa R., Tatsu Y., Shigeri Y., Yumoto N., and

Niki E. J. Mass Spectrom. 2006, 41, 103. [223] Thevis M., Geyer H., Mareck U., and Schänzer W. J. Mass Spectrom. 2005,

40, 955. [224] Jäntti S., Uutela P., Airavaara M., Piepponen P., Kostiainen R., and Ketola R.

A. 21st Montreux Symposium on Liquid Chromatography/Mass Spectrometry, 10-12.11.2004, Montreux, Switzerland.

[225] Franssila S., Marttila S., Kolari K., Östman P., Kotiaho T., Kostiainen R.,

Lehtiniemi R., Fager C.-M., and Manninen J. J. MEMS 2006, 15, 1251-1259. [226] Chen J. and Davidson J.H. Plasma Chem. Plasma Process. 2002, 22, 495. [227] Ravindra K., Godoi A.F.L., Bencs L., and Van Grieken R. J. Chromatogr. A

2006 ,1114, 278.

70

[228] Doroshenko V.M., Laiko V.V., Taranenko N.I., Berkout V.D., and Lee H.S. Int. J. Mass Spectrom. 2002, 221, 39.

[229] National Institute of Standards and Technology. NIST chemistry webbook -

NIST standard reference database number 69, June 2005 release. ( Available at: http://www.webbook.nist.gov/chemistry/ ).

[230] Jang Y.H., Hwang S., and Chung D.S. Chemistry Letters 2001, 12, 1316. [231] Klonis N. and Sawyer W.H. Photochemistry and Photobiology 2000, 72, 179. [232] Knochenmuss R., Stortelder A., Breuker K., and Zenobi R. J. Mass Spectrom.

2000, 35, 1237. [233] Salo P.K., Salomies H., Harju K., Ketola R.A., Kotiaho T., Yli-Kauhaluoma

J., and Kostiainen R. J. Am. Soc. Mass Spectrom. 2005, 16, 906. [234] Dale M.J., Knochenmuss R., and Zenobi R. Anal. Chem. 1996, 68, 3321. [235] Chen Y.-C., Sheia J., and Sunner J. J Chromatogr. A 1998, 826, 77.