The role of different phenomena in surface-activated chemical ionization (SACI) performance

JOURNAL OF MASS SPECTROMETRYJ. Mass Spectrom. 2005; 40: 1550–1557Published online in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/jms.913

The role of different phenomena in surface-activatedchemical ionization (SACI) performance†

Simone Cristoni,1 Luigi Rossi Bernardi,1 Federico Guidugli,2 Michela Tubaro3 andPietro Traldi3∗

1 Universita degli Studi di Milano – CISI, via Fratelli Cervi 93, 20090 Segrate, Milano, Italy2 Thermo Electron, via Rivoltana, Rodano, Milano, Italy3 CNR-ISTM, Corso Stati Uniti 4, 35100 Padova, Italy

Received 2 July 2005; Accepted 28 July 2005

In previous studies, the production of ions in an APCI source without any corona discharge was observed,and the intensity of the ion signals showed significant increases on placing a metallic surface at 45°inside an orthogonal ion source. This method was named surface-activated chemical ionization (SACI).The present study was performed to investigate the mechanisms of ion production with or without thepresence of the metallic surface, by varying instrumental parameters and the geometrical configuration.Approximate calculations show that, in the absence of corona discharge and of any additional surfaces,ions cannot be produced by collisional phenomena, because of their low kinetic energy, in the 10−2 to 10−3

eV range. Two alternative possibilities have been considered: the first takes into account that ions mayoriginate by collision of neutral clusters of polar solvent molecules with the APCI source surfaces throughclusterelectric effect. The second takes into account that the water dissociation constant kw is temperaturedependent, passing from 10−14.1669 at 20 °C to 10−12.4318 at 90 °C. It means that the [H+] varies from 8.3 × 10−8

to 6.1 × 10−7 M going from 20 to 90 °C. Hence, at the high temperatures experimented in the APCI vaporizer,H+ becomes available in solution in molar quantities analogous to those of analyte, and the protonation ofthe analyte itself can consequently occur. The activation of further ionization processes in the presence ofthe metallic surface can be reasonably attributed to interactions between gas-phase analyte molecules andsolvent molecules adsorbed on the surface. Experiments performed with a thin layer of deuterated glycerolon the surface led to unequivocal results, i.e. the production of [M + D]+ ions of the analyte. Copyright 2005 John Wiley & Sons, Ltd.

KEYWORDS: surface-activated chemical ionization; surface-activated ionization

INTRODUCTION

Surface-activated chemical ionization (SACI) was recentlyproposed as an effective approach, as an alternative to atmo-spheric pressure chemical ionization (APCI) and electrospray(ESI) in the mass spectrometric analysis of biologically rele-vant molecules.1

A SACI experiment is particularly simple to realize. In aconventional APCI ion source, the corona discharge needleis substituted by a metallic surface, placed at a potentialof a few hundred volts. The sample solution is vaporizedin the usual way by the APCI nebulizer operating at atemperature in the range 350 �400 °C. Even if, in principle, noionizing conditions are present (i.e. electrons from the coronadischarge are completely suppressed and the vaporizingconditions are far from those typical of the thermosprayapproach), the production of ionized molecular species (i.e.[M C H]C, [M C Na]C, [M C K]C) is observed in high yields.

ŁCorrespondence to: Pietro Traldi, CNR-ISTM, Corso Stati Uniti 4,35100 Padova, Italy. E-mail: [email protected]†Paper presented at the 23rd Informal Meeting on MassSpectrometry, Fiera di Primiero, Italy, 15–19 May 2005.

The method has been successfully applied to a seriesof analytical problems.2 – 4 For example, the qualitative andquantitative analyses of amphetamines in urine was easilyobtained in SACI conditions both by direct infusion andby LC–MS, exhibiting a detection limit equal to or betterthan that usually obtained by ESI/MS2; it is remarkable thatthe same samples analyzed under conventional (discharge-on) APCI conditions did not lead to any significant data.2

Analogous results have been obtained for the determinationof levels of morphine, cocaine and their metabolites inbiological fluids.3 An interesting application of SACI inthe medical field was the qualitative and quantitativedetermination of 21-deoxycortisol (21-DF), an endogenousmetabolite marker for the diagnosis of congenital adrenalhyperplasia (CAH) autosomal recessive disorder.4 Also,in that case, SACI exhibited high sensitivity, allowingthe determination of 21-DF levels in plasma in the range2–50 ng/ml, sufficient for an unequivocal diagnosis of thedisease. Further investigations that exploit SACI in theanalysis of different substrates5,6 are in progress, and thepreliminary results confirm the validity of this instrumentalapproach.

Copyright 2005 John Wiley & Sons, Ltd.

Surface-activated chemical ionization 1551

Along with these series of successful applications,further investigations were devoted to elucidate the possiblemechanism(s) responsible for the SACI performance. In aprevious study,1 the behavior of SACI in the analysis ofsome peptides was tested by varying the nebulizing gasflow, sample solution flow, pH of solutions and the metalsurface area. The data thus obtained seemed to suggest thatSACI is the result of several different ionizing events inwhich the metallic surface plays a fundamental role.

In the present work, we present and discuss the dataobtained by varying other operational parameters, mostlyrelated to the geometry of the metallic surface and thepotential applied to it, to further investigate the SACImechanism(s), as well as the identification of possiblemechanism(s) responsible for ion production in absence ofboth corona discharge and metallic surface.

EXPERIMENTAL

Chemicals21-Deoxycortisol, arginine and deuterated glycerol werepurchased from Sigma-Aldrich (Milan, Italy). P2 peptide(PHGGGWGQPHGGGWGQ, MW 1570.8 Da, a partialsequence of prion protein PrP) was synthesized in theMario Negri laboratory (Milan, Italy). Acetonitrile (CH3CN),methanol (CH3OH) and chloroform (CHCl3) were purchasedfrom JT Baker (Milan, Italy).

Mass spectrometryAll the measurements were made using a LCQ DECA XPPLUS (Thermo, San Jose, CA, USA) operating with a SACIionization source.1 Surfaces of different geometries werefabricated in stainless steel (Figs 1(a) and (b)). The DC powersupply for the surface potential was homebuilt and couldgenerate negative or positive potentials in the range 0–400 V.The sample solutions were injected by a syringe pump, withflow rates in the range 30 µl/min. The sheath gas flow waskept constant at 1.4 l/min.

The MS was operated in the positive ion mode.

RESULTS AND DISCUSSION

In the previous investigation of the effect of several instru-mental parameters on the efficiency of SACI,1 most attentionwas paid to the evaluation of vaporization parameters; thiswas to test the hypothesis that ion evaporation can play animportant role in the SACI mechanism. The data so obtainedpartially supported this hypothesis; by increasing the flowrates of either vaporizing gas (Fg) or solution (Fs) in theranges 0.6–2.5 l/min and 10–150 µl/min respectively, a rea-sonably linear relationship of Fs/Fg vs ion intensity wasobtained for a narrow range after which saturation phe-nomena were observed. The positive role of the surface wasproved by increasing its dimensions; in fact, a linear relation-ship between surface area and signal intensity was found byvarying the former from 1 to 4 cm2.

After that investigation,1 a series of questions naturallyarose that all led to the need to understand the real role of thesurface. Is it an active region for sample ionization, or does it

(a)

(b)

Figure 1. (a) Geometry of the planar surface and (b) geometryof the convex surface.

simply behave as an electrostatic mirror leading to better ionfocusing into the entrance capillary orifice of ions previouslyproduced? If the latter hypothesis is true, how are the ionsgenerated?

The behavior of the solution vaporizer of the API sourcecan be visualized as shown in Fig. 2(a). The solution isvaporized inside the capillary, and a high-density vapor isfirst generated. At the exit of the capillary, the solvent andsample vapors experience a negative pressure gradient thatleads first to an increase in their kinetic energy dependingon the gas expansion and then to its diffusion inside theAPI chamber. It is reasonable to assume that the solventmolecules that have a molecular mass lower than that of theanalyte undergo a preferential diffusion, while the analytemolecules have a greater tendency to maintain their originaldirection for purely inertial reasons.

If the ions preexist in the solution, the only effect observedwould be their transfer in the gas phase. This could be thecase for peptide ions, which, depending on the solution pH,can be in ionic form. In fact, in the previous investigation1

of SACI, a relationship between the [MH]C ion intensityand pH of the peptide solution was found. However, theformation of abundant [MH]C ions without the presence of acorona discharge has also been observed for molecules thatare much less likely to have been significantly preionizedin solution (e.g. steroids, amphetamines, morphine, cocaine,benzoylecgonine, codeine, etc.).2 – 4

The formation of charged molecular species is alsoobserved from nonpolar molecules, even if in low yield,

Copyright 2005 John Wiley & Sons, Ltd. J. Mass Spectrom. 2005; 40: 1550–1557

1552 S. Cristoni et al.

MS capillaryentrance hole

Liquid

Liquid

Liquid

High-densityvapor

Vapor expansionand diffusion

Surface

(a)

(b)

Sample injection

Sample injection

Figure 2. Scheme of an orthogonal APCI source operating in (a) no-discharge conditions and (b) surface-activated chemicalionization (SACI) conditions.

without the metallic surface. This result indicates that, atleast partially, the ionization takes place in the regionsshown in Fig. 2(a). It could be a result of thermal effectsbut, considering the thermal lability of some of the analytesinvestigated, the survival of intact molecules seems toexclude this possibility.

A reasonable ionizing mechanism might involve colli-sional phenomena occurring either in the high-density vaporor in the dilute gas-phase region. The kinetic energy acquiredby the molecules of the expanding gas could lead to effectivecollisional phenomena with the neutral species present insidethe SACI source. The high density (atmospheric pressure)of the gas could lead to effective internal energy depositionthrough multicollisional phenomena, with the formation ofionic species, which in turn activate possible ion–moleculereactions. Alternatively, it could be hypothesized that pro-tonated molecules are generated by collisionally induceddecomposition of solvated analyte molecules through chargepermutation and/or proton exchange processes.

In order to evaluate this aspect, two simple calculationswere performed. They are approximate, but can give an ideaof the order of magnitude of some properties of the speciesemerging from the heated nebulizer of the API source.

Consider a tube of cross section s (m2) in which a liquidis injected with a speed vl (m/s), a volume flow V

žl (m3/s)

and mass flow mž

l (kg/s). At the exit of the tube, the vapor isejected with a speed vv, a volume flow V

žv, and a mass flow

mž

v. Of course, the mass flow of the injected liquid (mž

l) mustbe equal to the mass flow of the vapor emerging from thetube (m

žv):

mž

l D mž

v �1�

and, considering that

�i� Vž

l D mž

l

�l��l D density of liquid sample�

�ii� Vž

v D mž

v

�v��v D density of the vapor�

�iii� Vž

v D svv

one can write

Vž

l�l D Vž

v�v ) Vž

v D Vž

l�l

�v�2�

Further, since

Vž

l D svl

and

Vž

v D svv ) vv D Vž

v

s�3�

from Eqns (2) and (3) it is possible to calculate the vaporspeed as

vv D Vž

l�l

s�v�4�

In Eqn (4) the only term that is unknown is the densityof vapor emerging from the vaporizer, which is stronglydependent on its temperature. But, once the temperature isknown, the �v values can be found from the literature.7



The temperature of the vapor flowing out from thecapillary (mostly nitrogen as the nebulizer gas) was deter-mined for conditions in which the nebulizer temperature wasfixed (400 °C). Different solvents, i.e. H2O, CH3OH, CH3CN,CHCl3, CH3CN/H2O (1 : 1) and CH3OH/H2O (1 : 1), wereinjected into the capillary at different volume flow rates (10,15 and 30 µl/min). A microthermocouple was mounted justat the exit of the nebulizer device and the vapor temperatureswere measured.

In absence of any sample, i.e. by flowing N2 only, atemperature of 135 °C was measured. By injecting differentsolvents at different flow rates, only slight changes wereobserved, with temperatures in the range 135–145 °C. Forthese reasons, further calculations were performed for anexit vapor temperature of 140 °C.

The �v values, together with the calculated vapor speedsand the molar kinetic energies, are reported in Table 1.The highest value is obtained for water, correspondingto a kinetic energy of 10.56 meV. The lowest value isobserved for chloroform (1.60 meV), while intermediatevalues are observed for methanol and acetonitrile (5.86and 4.59 meV respectively). It is to be emphasized thatthese values represent the lower limits, calculated byconsidering homogenous vaporization of the liquid sample.Considering the laminar flow regime present inside acapillary line, higher speeds and hence higher kineticenergies are to be expected, even if of comparable orderof magnitude.

Another calculation was performed by considering thatordinary average thermal energies are of the order of kT(k D Boltzmann constant, T D absolute temperature). By thisapproach, energies for water were calculated to be 35 and57 meV/mol at 413 and 673 K respectively.

Both calculations indicate that the kinetic energies of thesolvent molecules are at least three orders of magnitude

Table 1. �v values (calculated at 140 °C), vapor speed(m/s) and molar kinetic energies (meV) calculated for H2O,CH3OH, CH3CN and CHCl3. Other details are given in thetext

Solvent �v (140 °C) vv (m/s) Ekmol (meV)

H2O 0.536 238 10.56CH3OH 0.955 133 5.86CH3CN 1.23 104 4.59CHCl3 3.58 36 1.60

less than those necessary to promote effective gas-phasecollision-induced ionization and decomposition processesand, consequently, collisional phenomena cannot be retainedresponsible for the ion generation.

Recently, a new ionization mechanism, called the clus-terelectric effect, has been proposed by Gebhardt et al.8 It hasbeen observed that neutral clusters of water (as well asSO2) molecules lead, by impact with a solid surface, to theproduction of positive and negative fragment ions, even ifthe kinetic energy of the neutral species is lower than itsionization energy. This phenomenon could be invoked torationalize the results obtained by no-discharge APCI, con-sidering the impact of water (or other polar solvent) neutralclusters on the vaporizer line surface. The ions so formedmight act as protonating agents, leading to the formationof [M C H]C species of the analyte. However, this model isdifficult to be applied in the present case: in fact previousresearch8 showed that all positively charged fragments carryan alkali ion acting as catalyzer for the ion pair formation.However, we cannot exclude the presence, at trace level,of alkali atoms on the vaporizer line, but their low levelcannot justify the high yield of ion production observed inno-discharge APCI experiments.

300 305 310 315 320 325 330 335 340 345 350 355

m/z

300 305 310 315 320 325 330 335 340 345 350 355

m/z

0

20

40

60

80

100

0

20

40

60

80

100

345.2

347.3

348.3

(a)

(b)

Rel

ativ

e ab

unda

nce

341.5

342.5327.4

328.5

309.4 340.4312.5343.6 351.3313.5 329.6

319,4307.3 335.4314.5 348.8325.7 354.4304.3324.2

Figure 3. No-discharge API spectra of 21-DF, 10 ng/ml solutions in (a) H2O and (b) CH3OH.



Another aspect was considered, moving from physical toa chemical phenomenon. At 20 °C, the dissociation constantfor water (pKw) is 14.1669, while it is 13.0171 at 60 °C and12.4318 at 90 °C.9 Considering that

Kw D [HC][OH�]

it follows that the [HC] concentration changes from 8.3 ð 10�8

to 6.1 ð 10�7 M passing from 20 to 90 °C. In other words, the[HC] concentration shows a clear increase with respect totemperature. These values are just related to pure water:the possible (and expected!) presence of electrolytes even atlow concentration would lead to a further increase of [HC]concentration. Consequently, it is more than reasonable toassume that during heating, but before solution vaporization,a decrease of pH of the solution takes place with theformation of protonated molecules of the analyte.

To investigate this point, a series of experiments wasperformed by analysis of 21-deoxycortisol (21-DF) underdifferent conditions. Ten nanograms per milliliter solutionsof 21-DF in CHCl3, H2O and CH3OH were injected directlyinto the source at a flow rate of30 µl/min. In the case of thesolution in CHCl3 no [MH]C ions of 21-DF were observed;this negative result can be well justified by the low dis-sociation constant of chloroform10 (pK�CHCl3�

a D 25), i.e. bythe ineffectiveness of chloroform as a protonating agent.The results obtained for the water and methanol solutionsare reported in Figs 3(a) and 3(b), respectively. In the for-mer case, an abundant (2.39 ð 106 counts/s) [MH]C ion isdetectable, with practically no chemical background. In thecase of the methanol solution, a lower signal intensity wasobtained (3.78 ð 104 counts/s), and the [MH]C peak wasobserved together with a high level of chemical background,making its identification practically impossible.

Arginine

1.40E+08

1.20E+08

1.00E+08

8.00E+07

6.00E+07

4.00E+07

2.00E+07

0.00E+00

1.60E+08

coun

ts/s

coun

ts/s

coun

ts/s

−2.00E+06

−5.00E+07

5.00E+07

1.50E+08

1.00E+08

2.50E+08

2.00E+08

3.00E+08

3.50E+08

4.50E+08

4.00E+08

0.00E+00

1.60E+07

1.40E+07

1.20E+07

1.00E+07

8.00E+06

6.00E+06

4.00E+06

2.00E+06

0.00E+000 500 1000 1500 2000

0 500 1000 1500 2000

0 500 1000 1500 2000

0 500 1000 1500 20000 500 1000 1500 2000

0 500 1000 1500 2000

0 500 1000 1500 20000 500 1000 1500 2000

(c)

(b)

(a)

0 500 1000 1500 2000 2500

0 500 1000 1500 2000 2500

V

0 500 1000 1500 2000

V

V

0 500 1000 1500 2000 2500

Figure 4. Trends of [MH]C signal intensity vs voltage applied to the metal surface obtained for (a) arginine; (b) 21-DF and(c) PHGGGWGQPHGGGWGQ peptide.



These results are consistent with the above collisionalhypothesis (in fact pK�CH3OH�

a D 15.5, higher than that ofwater)11 which, in our opinion, could be considered agood rationale of the mechanism of the no-discharge APCIphenomenon.12 – 14

However, these considerations cannot explain the largeincrease in ion production (typically from 104 to 106 counts/s)observed when a metallic surface is mounted at 45° withrespect to the direction of vapor emission (Fig. 2(b)). In theprevious investigations of the SACI mechanism,1 it wasshown that the best results are obtained not by holding thesurface at ground potential, but by leaving it floating, i.e.insulated from ground.1 Interestingly, it was observed thatin the latter condition the ionization efficiency increased withincreasing nebulizing time, i.e. the total elapsed time afternebulization was started. This result might be explained byconsidering that a number of ions, generated by the abovedescribed mechanisms, are deposited on the surface; byincreasing the nebulizing time, this number increases to thepoint where a suitable potential is induced on the surface. Ifthis hypothesis is correct, a positive voltage applied to thesurface should lead to a significant increase in the yield ofpositive ions.

To investigate this proposal, an external power supplywas used to apply negative or positive potentials in therange �400 to C400 V to the surface. In all the cases, negativevoltages led to negative results: no signal could be detected.On the contrary, as shown by the results reported in Fig. 4,the ion intensity (counts/s) is strongly dependent on thesurface voltage. Thus, in the case of SACI analysis of arginine

(Fig. 4(a)) a very weak signal was observed for zero surfacepotential. Increasing signal intensities are observed as thesurface potential is raised, until a maximum is observed atC150 V, followed by a very rapid intensity drop for higherpotentials. Similar trends were obtained also in the cases of21-DF and the PHGGGWGQPHGGGWGQ peptide (Fig. 4(b)and (c), respectively). It is interesting to note that the potentialat the intensity maximum is analyte dependent; this effectmust reflect in some way the different ionization and/orfocusing conditions created by the presence of the surface.In general, for a range of analytes, surface potential valuesin the range 150–300 V were found to result in sensitivitymaxima. In this regard, it is important to note that the ionsleave the region of the vaporizer tube at or near groundpotential, but experience a negative voltage gradient fromthe surface (¾C100 V) to the entrance of the heated capillary(C25 V).

At this stage, the most obvious explanation of the effectof the surface on ion-signal intensity is based on its ability tofocus the ion beam onto the mass spectrometer (MS) entranceorifice, as shown schematically in Fig. 2(b). This view isconsistent with the behavior illustrated in Fig. 4; ions ofhigher masses will require higher fields to be guided towardthe entrance capillary orifice. On this basis, the insertion ofthe metallic surface would only result in driving ions tothe MS entrance, which would otherwise be lost inside thesource. If this is true, different geometries of the surfaceshould greatly affect the signal intensity. For this reason,different shapes of the surface were investigated. Insteadof using a planar geometry, concave and convex surfaces,

Rel

ativ

e ab

unda

nce

Rel

ativ

e ab

unda

nce

1615.9

1593.9

1631.81571.8

1632.8

1637.71430.81647.8

1615.8

786.8

1593.8808.9

819.8

1571.8

1631.81632.8

1359.7827.7

777.51337.6861.6 1637.8

[M + H]+

[M + H]+

[M + 2H]2+

[M + Na]+

[M + Na]+

[M + K]+

[M + K]+

200 400 600 800 1000 1200 1400 1600 1800 2000

m/z

200 400 600 800 1000 1200 1400 1600 1800 2000

m/z

0

20

40

60

80

100

0

20

40

60

80

100

(a)

(b)

Figure 5. SACI spectra of the PHGGGWGQPHGGGWGQ peptide obtained by injection of the same solution, (a) with a cylindricalsurface arranged in a concave orientation with respect to the vaporizer and MS entrance capillary and (b) with the cylindrical surfacearranged in a convex orientation.



with their symmetry axis parallel or perpendicular to theplane defined by the nebulizer-MS entrance region, wereinvestigated. The surface area was always maintained at400 mm2.

Some differences were observed on varying the surfacegeometries, as shown by the spectra reported in Fig. 5. Ascan be observed in Fig. 5(a), the molecular cluster is due to[M C H]C, [M C Na]C and [M C K]C all them reflecting thephenomena taking place in solution and/or on the surfacereasonably well. In the case of a surface that was concavetoward both the nebulizer and mass spectrometer entrance,a signal of 1 ð 107 counts/s (Fig. 5(a)) was obtained forthe peptide PHGGGWGQPHGGGWGQ; with the concavityin the opposite direction, i.e. the surface convex towardboth nebulizer and spectrometer entrance orifice, a signal of5.3 ð 107 counts/s (Fig. 5(b)) was obtained. It is remarkablethat, in the latter case, signals due to both singly and doublyprotonated peptides are present, whereas in the concavesurface case only the singly protonated species are observed(Fig. 5(a)). This observation suggested a possible active roleof the surface in the ion production. It is reasonable to assumethat in the latter case the formation of doubly charged speciesoriginate by the interaction of the singly charged ion withthe surface, leading to a further protonation reaction. It isto be emphasized that this phenomenon is observed only inthe case of peptides (in the case of steroids, alkaloids andamphetamines, the formation of doubly charged species hasnot been observed) and can be justified by the presence of alarge number of basic sites in these molecules.

The geometry of the SACI source with the surfacemounted at 45° (Fig. 2(b)) is reminiscent of the experimentalsetup for surface-induced dissociation (SID) experiments,suggesting that SACI might somehow be related to thephenomena extensively studied by Cooks and coworkers.15

SID is effectively employed for decomposition of mass-selected precursor ions; however, as noted by Cooks in a1994 review, a rich chemistry accompanies the collision ofions of translational energy of some tens of electron volts withorganic surfaces.16 This new chemistry was investigated formany different reactions between ions of different nature(e.g. organic,17,18 organometallic,19 silicon clusters20) with

surfaces treated in different ways. What is of interest for thepresent study are the results obtained for organic radicalcations impinging on hydrocarbon surfaces, which showthat hydrogen abstraction takes place.18 Of course, in thepresent case the situation is significantly different from SIDexperiments for the following reasons:

1. SID operates in vacuum conditions (10�7 torr), while SACItakes place at atmospheric pressure;

2. SID is based on the selection of ions of interest exhibitinga closely specified low kinetic energy, impinging on andreacting with a chemically well-defined surface. In SACI,either neutrals or collisionally generated ions can impingeon the surface, whose chemical nature will depend mainlyon the nature of the solvent. In fact, it is reasonable toassume that in SACI, because of the high pressure presentinside the source, an adsorption–desorption equilibriumof solvent molecules on the metallic surface occurs.

On the basis of the above considerations, we hypothesizefor SACI the following mechanisms:

1. The metallic surface employed in the experiment isconsidered to be completely covered by the solventmolecules employed in the experiments, thus becoming aproton-rich environment.

2. Neutral analyte molecules, that survived the ionizationprocesses occuring in solution and described above,interact with the surface, and [MH]C species are producedby proton abstraction.

3. The negative potential gradient between the surface(¾100 V) and entrance capillary (25 V) leads to efficienttransfer of the ions so formed into the MS entrancecapillary.

To investigate these ideas, we performed a simpleexperiment based on the deposition, on the metallic surface,of a thin layer of deuterated glycerol. Under this condition,when the analysis of the PHGGGWGQPHGGGWGQ peptidewas performed by SACI, the signal of the [M C DC] ion atm/z 1573 was observed to dominate (Fig. 6). This is goodevidence for the participation of the chemicals present on thesurface in the ionization phenomena occurring in SACI.

[M + D]+

[M + H]+

1560 1565 1570 1575 1580 1585

m/z

1572.9

1571.8

1573.9

1574.9

1575.9

Rel

ativ

e ab

unda

nce

0

50

100

Figure 6. SACI spectrum of the peptide PHGGGWGQPHGGGWGQ obtained by the pretreatment of the metallic surface withdeuterated glycerol.



Further experiments were performed by varying the sur-face angle. The best results were obtained with a small angle(5–10°) with respect to the vaporizer axis. This result can berationalized either by the interaction of a larger populationof neutral molecules of the analyte with the surface or by theoccurrence of phenomena analogous to those observed ingrazing-incidence surface-induced dissociation (GI-SID)).21

Both experimental results and quantum-mechanical mod-els show that the GI-SID process is significantly differentfrom normal SID and result in highly effective inter-nal energy deposition into species interacting with thesurface.

CONCLUSION

In conclusion, the present work has suggested that themechanisms operating in SACI can be summarized asfollows:

1. Estimates of the speed of the neutrals emerging from thevaporizer device of an APCI ion source (operating withoutany corona discharge) seem to exclude that solvent ionscan be formed by gas-phase, multicollision phenomenain the atmospheric pressure environment. The ionizationobserved in the absence of corona discharge and sur-face can be reasonably due to the temperature-dependentincrease of HC production in the solution, before solutionvaporization. However, the partial contribution of phe-nomena analogous to the clusterelectric effect8 cannot beexcluded.

2. A metal surface mounted in front of the vapor streamemerging from the vaporizer leads to a large increase ofthe analyte ion abundance, and this can be rationalized byinteraction of neutral analyte molecules (that had survivedthe ionization phenomena occurring in the solution) withsolvent molecules adsorbed on the surface itself. Thisinteraction, reasonably based on a proton abstractionmechanism, analogous to that observed in SID, leadsto the production of [MH]C species.

The SACI approach has led to an increase of sensitivityand reproducibility and to a decrease of chemical back-ground for a wide range of analytical problems, and it iscurrently employed in our laboratories for the developmentof new and effective analytical procedures.

AcknowledgementsThe authors thank Lorenzo Capucci of Istituto ZooprofilatticoSperimentale della Lombardia e dell’Emilia-Romagna for providingthe LCQDecaXP instrument used in these experiments.

REFERENCES1. Cristoni S, Rossi Bernardi L, Biunno I, Tubaro M, Guidugli F.

Surface-activated no-discharge atmospheric pressure chemicalionization. Rapid Commun. Mass Spectrom. 2003; 17:1973.

2. Cristoni S, Rossi Bernardi L, Gerthoux P, Gonella E, Mocarelli P.Surface-activated chemical ionization ion trap massspectrometry in the analysis of amphetamines in diluted urinesamples. Rapid Commun. Mass Spectrom. 2004; 18: 1847.

3. Cristoni S, Rossi Bernardi L, Gerthoux P, Guidugli F, Gonella E.Surface activated chemical ionization ion trap mass spectrometryin the analysis of drugs in dilute urine samples. Part II: analysisof morphine and other street drugs. J. Mass Spectrom. Submitted.

4. Cristoni S, Sciannamblo M, Rossi Bernardi L, Biunno I,Gerthoux P, Russo G, Chiumello G, Mora S. Surface-activatedchemical ionization ion trap mass spectrometry in the analysisof 21-deoxycortisol in blood. Rapid Commun. Mass Spectrom. 2004;18: 1392.

5. Cristoni S, Rossi Bernardi L, Cantu M, Gerthoux P, Gonella E,Brambilla M, Cavalca V, Zingaro L, Guidugli F. Surface-activated chemical ionization in the analysis of arginine inplasma samples. Rapid Commun. Mass Spectrom. 2005; 19: 123.

6. Cristoni S, Ranaldo M, Bombi GG. Surface Activated ChemicalIonization in the analysis of pesticides. results to be published.

7. http://www.questconsult.com/¾jrm/enthpres.html [accessed2001].

8. Gebhardt CG, Schroder H, Kompa KL. Surface impact ionizationof polar-molecule clusters through pickup of alkali atoms. Nature1999; 400: 544.

9. Schmidt E. Properties of Water and Steam in SI-Units. Springer-Verlag: New York, 1969.

10. Streitwieser A, Heathcock CH. Introduction to Organic Chemistry.Macmillan Publishing Company: New York, 1992.

11. Fox MA, Whitesell JK. Organic Chemistry, 2nd ed., Jones andBartlett Publishers International Georg Thieme Verlag: Stuttgart,New York, 1997.

12. Cristoni S, Rossi Bernardi L. Development of newmethodologies for the mass spectrometry study of bioorganicmacromolecules. Mass Spectrom. Rev. 2003; 22: 369.

13. Cristoni S, Rossi Bernardi L, Biunno I, Guidugli F. Analysis ofpeptides using partial (no discharge) atmospheric pressurechemical ionization conditions with ion trap mass spectrometry.Rapid Commun. Mass Spectrom. 2002; 16: 1686.

14. Cristoni S, Rossi Bernardi L, Biunno I, Guidugli F. Analysis ofprotein ions in the range 3000–12000 Th under partial (nodischarge) atmospheric pressure chemical ionization conditionsusing ion trap mass spectrometry. Rapid Commun. Mass Spectrom.2002; 16: 1153.

15. Dongre AR, Somogyi A, Wysocki V. Surface-induced dissocia-tion: an effective tool to probe structure, energetics and frag-mentation mechanisms of protonated peptides. J. Mass Spectrom.1996; 31: 339.

16. Cooks RG, Ast T, Pradeep T, Wysocki V. Reactions of ions withorganic surfaces. Acc. Chem. Res. 1994; 27: 316.

17. Morris MR, Rieder DE, Winger BE, Cooks RG, Ast T,Chidsey CED Jr Ion Surface Collisions at Functionalized Self-Assembled Monolayer Surfaces. Int. J. Mass Spectrom. IonProcesses 1992; 122: 181.

18. Ast T, Mabud MdA, Cooks RG. Collisions Reactive collisionsof polyatomic ions at solid surfaces. Int. J. Mass Spectrom. IonProcesses 1988; 82: 131.

19. Wainhaus SB, Lim H, Schultz DG, Hanley L. Energy transferand surface-induced dissociation for SiMe3

C scattering off cleanand adsorbate covered metals. J. Chem. Phys. 1997; 106: 10 329.

20. Whetten RL, St John PM. Reactions in cluster-surface collisions.Pittsburgh Conference, Chicago, IL, February 27 to March 3, 1994.

21. Wieghaus A, Schmidt L, Popova AM, Komarov VV, Jungclas H.Grazing incidence surface-induced dissociation of proto-nated peptides generated by matrix-assisted laser desorp-tion/ionization. Rapid Commun. Mass Spectrom. 2000; 14: 1654.


The role of different phenomena in surface-activated chemical ionization (SACI) performance

Documents

Transcript of The role of different phenomena in surface-activated chemical ionization (SACI) performance