Significant Enhancement of Negative Secondary Ion Yields by Cluster

Subscriber access provided by CMU Libraries - http://library.cmich.edu

Article

Significant Enhancement of Negative Secondary Ion Yields by Cluster Ion Bombardment Combined With Cesium Flooding Patrick Philipp, Tina Bernadette Angerer, Sanna Sämfors, Paul Blenkinsopp, John Stephen Fletcher, and Tom Wirtz Anal. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.analchem.5b02635 • Publication Date (Web): 17 Sep 2015 Downloaded from http://pubs.acs.org on September 24, 2015

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Analytical Chemistry is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 19

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

Significant Enhancement of Negative Secondary Ion Yields by Cluster Ion Bombardment Combined With Cesium Flooding Patrick Philipp1*, Tina B. Angerer2, Sanna Sämfors,3 Paul Blenkinsopp4, John S. Fletcher2,3, Tom Wirtz1 (1) Advanced Instrumentation for Ion Nano-Analytics (AINA), MRT Department, Luxembourg Institute of Science and Technology (LIST), L-4422 Belvaux, Luxembourg (2) Department of Chemistry and Molecular Biology, University of Gothenburg, SE-412 96, Gothenburg, Sweden (3) Department of Chemistry and Chemical Engineering, Chalmers University of Technology, SE-412 90, Gothenburg, Sweden (4) Ionoptika Ltd, Chandlers Ford, Southampton, UK ABSTRACT: In secondary ion mass spectrometry (SIMS), the beneficial effect of cesium implantation or flooding on the enhancement of negative secondary ion yields has been investigated in detail for various semiconductor and metal samples. All results have been obtained for monatomic ion bombardment. Recent progress in SIMS is based to a large extent on the development and use of cluster primary ions. In this work we show that the enhancement of negative secondary ions induced by the combination of ion bombardment with simultaneous cesium flooding is not only valid for monatomic ion bombardment, but also for cluster prima ry ions. Experiments carried out using and bombardment on silicon show that yields of negative secondary silicon ions can be optimised in the same way as by and bombardment. Both for monatomic and cluster ion bombardment, the optimisation does not depend on the primary ion species. Hence, it can be assumed that the silicon results are also valid for other cluster primary ions and that results obtained for monatomic ion bombardment on other semiconductor and metal samples are also valid for cluster ion bombardment. In SIMS, cluster primary ions are also largely used for the analysis of organic matter. For polycarbonate, our results show that bombardment combined with cesium flooding enhances secondary ion signals by a factor 6. This can be attributed to the removal of charging effects and / or reduced fragmentation, but no major influence on ionisation processes can be observed. The use of cesium flooding for the imaging of cells was also investigated and a significant enhancement of secondary ion yields was observed. Hence, cesium flooding has also a vast potential for SIMS analyses with cluster ion bombardment.

Introduction Secondary ion mass spectrometry (SIMS) is known for its excellent detection limits combined with high lateral and high mass resolution. At the same time, sensitivity and detection limits are largely dependent on the ionisation processes of secondary ions. The latter depend on the local sample composition and hence complicate quantification. The ionisation mechanisms can also be influenced by the choice of the primary ion species and their implantation into the sample. Implanted primary ions change the near-surface composition of the sample and modify the ionisation probabilities. For instance, the emission of negative secondary ions can be largely enhanced by the use of electropositive primary ion species. For metallic and semiconductor samples, the use of cesium primary ion beams is most common since, in this configuration, a same species is used for the sputtering of the sample and the enhancement of the negative secondary signal by cesium implantation.1,2 However, this method has the disadvantage that the cesium surface concentration depends on impact energy and incidence angle of the primary ions and on most instruments these parameters cannot be chosen freely. This drawback has been circumvented by the development of a cesium flooding methodology which was first developed by Bernheim et al.3,4 and later optimised by Wirtz et al.5,6 Pri-

mary ion species can be chosen according to the application and the cesium surface concentration is controlled by the simultaneous flooding of the sample surface with cesium vapour. With this setup, the cesium concentration can be changed over a large range7 and negative secondary ion yields can be increased by several orders of magnitude. For elements with an electron affinity higher than carbon, total ionisation becomes even possible.8,9 Initial studies used gallium and cesium primary ion bombardment. More recently, similar results have been obtained with helium and neon primary ions,10 showing that the previously obtained results are valid for any kind of monatomic primary ion species. The ionisation processes for the aforementioned results on metallic and semiconductor samples were always described by the electron tunnelling model where the ionisation process is described by a tunneling process of electrons between the sample surface and the sputtered atom.11 For this model, the ionisation probability of negative secondary ions depends on the work function of the sample surface and the electron affinity of the sputtered atom. This was verified both for Cs+ primary ion bombardment and primary ion bombardment combined with simultaneous cesium flooding. Studies on work function changes under bombardment include the work of Gnaser on , , and 12, energy distributions and fragmentation of graphite,13 and during the depth profiling of

ACS Paragon Plus Environment


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Si and SiO2 by van der Heide and Azzarello.14 Similar results were obtained for and bombardment combined with simultaneous cesium flooding on silicon, aluminium, indium phosphide, gallium arsenide and nickel by Philipp et al.15 Recently, it was discussed that the experimentally obtained energy spectra of the sputtered secondary ions do not depend on the velocity of the sputtered matter, but according to the electron tunnelling they should.11 Hence, ionisation processes for metal and semiconductor samples are now believed to depend on the local electronic structure of the sample and the electronic properties of the sputtered atom or cluster.16 At this point it is not clear how the ionisation mechanisms get influenced by the perturbation caused by the impact of the primary ion. Up to now all results have been obtained for monatomic ion bombardment, but the combination of cluster ion bombardment with simultaneous cesium flooding or cesium cobombardment has never been investigated. For cluster ions, secondary ion formation can be expected to change from monatomic ion bombardment since the sample surface gets heavily disturbed by the cluster impact and the surface is no longer well defined in the area atoms and clusters are sputtered from.17 Yet, a significant enhancement of the negative secondary ion yields would also be highly beneficial for SIMS using cluster ion bombardment. In this work, we are going to investigate how and ion bombardment combined with simultaneous cesium flooding affects ionisation processes of secondary ions sputtered from silicon and polycarbonate and by how much negative secondary ion yields can be enhanced. Some preliminary tests are also carried out on cells. In this way, conclusions on how cluster ion bombardment compares to monatomic bombardment in terms of ionisation processes can be drawn and new applications, for inorganic devices but also for organic and biological samples, might emerge. Up to now cluster ion bombardment was successful in organic depth profiling18 and the characterisation of biological systems,19 but suffers from low secondary ion yields. Low energy primary ions used for the sputtering of the sample and combined with 15 keV ions for the analysis were already successful in depth profiling polycarbonate.20 Later on the same technique was extended to the depth profiling of PMMA and PS,21 amino-acid and sugar films,22 as well as biological films.23 Hence, under given experimental conditions, cesium has an influence on the fragmentation of molecules and the emission of negative secondary ions from organic samples, and the same can be expected for and cluster ions combined with simultaneous cesium flooding.

Page 2 of 19

Experimental setup Ionoptika J105 instrument An Ionoptika J105 instrument was used for the experi ments.24 The instrument is equipped with a 40 keV ion gun and a 40 keV gas cluster ion beam (GCIB), the latter providing a nominal cluster size of 4000 Ar atoms (in fact argon gas with 8% CO2 is used as this has been shown to produce larger gas clusters for similar pressure)25, written as in this paper. Both ion guns provide continuous primary ion beams while secondary ions enter a buncher and get compressed for a time focus at the entrance of a reflectron-ToF analyser. For bombardment, the primary ion current was set to 41 pA or 100 pA. The current density was changed by modifying the raster size from 75 × 75 µm2 to 1200 × 1200 µm2 for the 100 pA current and from 50 x 50 µm2 to 600 x 600 µm2 for the 41 pA current. For irradiation of silicon, the current was set to 65 pA and the raster size was change from 75 × 75 µm2 to 800 × 800 µm2. For polycarbonate, the intensity was set to 43 pA and the depth profile was acquired over a 300 µm × 300 µm field of view using 32 × 32 pixels. The fluence to acquire the frames was 2.98 × 1011 ions/cm-2 per Z-layer and equal to 1.49 × 1014 ions/cm-2 for the whole depth profile (500 Z-layers). For the cell analysis, the current was set to 3 pA and images were acquired over a 256 µm × 256 µm field of view using 128 × 128 pixels. Cesium flooding system The cesium flooding system, which was initially developed at the Luxembourg Institute of Science and Technology (LIST)5,26, was modified to fit on the Ionoptika J105 instrument. In short, the cesium flooding system is made of (i) a reservoir, (ii) an isolation valve with flange for rough pumping and, (iii) a tube with another isolation valve and a feedthrough to guide the cesium vapour to the centre of the analysis chamber (Figure 1). The reservoir contains an ampoule with metallic cesium. The latter can be opened under vacuum via a screw mechanism (Figure 1a). The reservoir and the tube guiding the vapour to the analysis chamber are heated in order to control the partial pressure of the cesium vapour. Their temperature can be adjusted separately. The cesium is guided through the tube as close as possible to the sample (Figure 1b). In this work, the temperature of the inner tube was fixed to 120°C and that of the outer part with reservoir was set to 50°C for the experiments of silicon and to 115°C for the depth profiles on polycarbonate. Measuring the Cs deposition rate directly in the J105 instrument with a Quartz microbalance was not possible. Based on previoulsy measured deposition rates, it is estimated to be about 0.6 Å/s at 50°C and about 2 Å/s at 115°C.

Figure 1: Pictures of the cesium flooding system installed on the Ionoptika J105 instrument: a) view of the cesium reservoir mounted to the analysis chamber, b) view inside the chamber with the inner tube of the flooding system pointing towards the centre of the chamber.


In the current setup it is essential that the cesium vapour covers the whole analysis area. To check this point, an area of the copper sample holder was exposed to cesium vapour overnight and a stage scan of 8 mm × 8 mm of the sample holder was carried out afterwards. The Cs+ image shows clearly that the cesium covers an area of about 2.5 mm × 3.8 mm, which is significantly larger than the typical raster size on the Ionoptika J105 instrument (Figure 2). The image has been recorded us ing the primary ion beam.

Figure 2: secondary ion image of the 8 mm x 8 mm stage scan of the sample holder after exposing the sample holder to the cesium vapour overnight. The image was recorded using the ion beam.

Results and discussion bombardment combined with simultaneous cesium flooding. Experiments for bombardment combined with simultaneous cesium flooding have been carried out for two different primary ion currents (Figure 3). Secondary ion intensities have been normalised by dividing the recorded data (counts/frame) by the product of frame duration, sputter density and analysed area to get an intensity in counts/Coulomb. For both series of measurements, the global behaviour is the same. Secondary ion intensities are lowest at high current densities and increase with decreasing density. This is in agreement with ionisation models since for a constant cesium deposition rate the cesium surface concentration decreases with increasing current density. Important information is that the negative secondary ion yields of silicon can be enhanced by the cesium flooding. Apparently, the intensities are lower than those of the different cluster ions ( to ). This is due to the transmission of the J105 mass spectrometer which is optimised for high masses and which is significantly lower at the low mass range. For equal transmission, the intensity would be highest.27 Without cesium flooding, the secondary ion intensities would be increasing linearly with the current density.

Normalised intensity (counts/C)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60



Page 3 of 19

105

a) Increasing Cs concentration

104 103 102 101 Ip = 41 pA 0

10 b) 105

SiSi2-

104

Si3Si4-

103

Si5Si6-

102

Si7Si8-

1

Si9-

10

Si10-

Ip = 100 pA

100 10-4

10-3

10-2

10-1

Current density (pA/µm2) Figure 3: Secondary ion intensities of different silicon clusters normalised to primary ion current as a function of current density for bombardment combined with simultaneous Cs flooding: a) for a primary current of 41 pA, b) for a primary current of 100 pA. During the experiments, the temperature of the cesium reservoir was kept constant at 50°C.

The secondary ion intensities of the different clusters reach a maximum at one point, though not necessarily for the same current density. Systematically, the optimal combination of current density and cesium deposition rate for silicon cluster ions is obtained at a lower higher current density that for . This is related to the electron affinities of the different secondary ions. For silicon clusters, the exact values change slightly depending on the source,28,29 but the general trend remains the same. The electron affinity of is lowest (1.4 eV), followed by (2.1 eV), and (2.2 -2.3 eV), (2.3 eV) and (2.5 eV).28,29 According to ionisation models, higher electron affinities should produce higher negative secondary ionisation probabilities. In addition, it has been shown that full ionisation can be obtained for once a high enough cesium surface concentration is obtained.8 Since the different clusters have higher electron affinities than monatomic silicon, their ionisation should be optimised for lower cesium concentrations, i.e. higher current densities. The relative changes in secondary cluster intensities largely agree also with the vary-

3



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ing electron affinities. , which has one of the lowest electron affinities, reaches its relative highest intensity at a smaller current density than the other clusters. Our results show, that the cesium flooding does not only allow for the optimisation of ions, but also of the different cluster ions as long as their electron affinity is equal to or higher than for monatomic silicon. A similar behaviour of the intensities of secondary ion clusters was already observed for bombardment. Gnaser has already shown that under bombardment the ionisation probabilities of secondary clusters correlates well with the work function change of the surface, hence the electron tunnelling model describes well the ionisation process.30 For the emission of secondary clusters, but the exponential dependence of the ionisation probability on the work function change could not be observed. A direct comparison of our data with literature results is not possible since a measurement of work function changes is not possible on the J105 instrument. Dissociation energies, or dissociation probabilities, do not change significantly for the different clusters,28,29 so that they should have no major influence on the secondary ion intensities. Table 1: Sputter efficiency as a function of current density for 41 pA bombardment of silicon combined with simultaneous Cs flooding. current density (pA/µm2)

Sputter yield

1.64 x 10-2

11.0

1.82 x 10-3

10.4

6.56 x 10

-4

16.9

3.35 x 10

-4

16.5

2.02 x 10-4

15.6

1.36 x 10-4

-35.0

Table 2: Sputter efficiency as a function of current density for 100 pA bombardment of silicon combined with simultaneous Cs flooding. current density (pA/µm2)

Sputter yield

1.78 x 10-2

45.4

-3

16.0

1.60 x 10-3

13.2

4.44 x 10

4.00 x 10-4

0.8

1.56 x 10-4

-6.3

6.94 x 10-5

-22.8

For several reasons, the highest secondary ion intensities are not obtained at the lowest current densities: the sputter rate decreases with lowering current densities and full ionisation has already been obtained at higher current densities. In addition, for the lowest values there is no net sputtering but matter is deposited (Table 1 and Table 2). This is not specific to bombardment, but has also been observed for monatomic ion bombardment.31 The competition between cesium deposition and sputtering induced by the beam starts shifting towards deposition. Normally cesium does not stick well on itself, but

Page 4 of 19

in the current situation the low sputter rates facilitate the oxidation in the irradiated area with the build-up of a cesium oxide. For a larger Cs flux, the deposition of matter is expected to occur at higher primary current densities since the balance between net sputtering and deposition depends on the ratio between current density and Cs flux. It is also known, that 32-34 beams start depositing carbon for too low impact energies. In particular, on silicon highly stable silicon carbide can form.35 However, the 40 keV impact energy used in this work should be high enough to avoid this issue. The enhancement of secondary ion yields observed under cluster ion bombardment shows that a significant amount of cesium must be present in the area from which the secondary ions are sputtered. Hence, the cesium layer is able to restore in between two overlapping impacts. This is not obvious because the cluster ion bombardment creates a much larger perturbation of the surface than monatomic ion bombardment. The large atomic radius of cesium which allows only for a low solubility in silicon and leads to its diffusion from the bulk to the sample surface on time scales smaller than the time difference between two consecutive and overlapping impacts may help at this point.36,37 The diffusion of cesium to the sample surface is also supported by the formation of cesium oxide dots once the sample is taken out of vacuum, although this happens on much longer time scales over several days.38 In addition, the adhesion of cesium on cesium is low, facilitating its surface diffusion to areas with low cesium surface concentrations. For some reason, the primary current influences also the whole processes leading to higher secondary ion intensities for a given current density when using a higher primary current (Figure 3). At about 2 × 10-2 pA/µm2, the intensity is about 5 times higher when using a current of 100 pA compared to 41 pA and the difference in intensities reduces only slightly to about a factor 4 at current densities around 1.5 × 10-4 pA/µm2. As the data is plotted as a function of current density, the primary ion current is not expected to have any influence. For both primary currents, the curves are also parallel one to the other while the slope of the Si cluster ions is smaller for the 41 pA than for the 100 pA primary current. A possible explanation for this could not be found, but the residual oxygen in the instrument may alter the ionisation mechanisms. For cesium ion bombardment or monatomic ion bombardment combined with cesium flooding, studies focused on the enhancement of the intensity. Some small cluster ions are sputtered, but their intensities are slightly lower than the intensity. Our results show that for the ion bombard ment, secondary ion clusters from to contribute significantly to the total secondary ion current. Since the direct sputtering of small silicon clusters is a common mechanism for oxygen bombardment,39 the same can be expected for the bombardment. Considering the local perturbation induced by the cluster impact, ionisation models like the electron tunnelling model which are based on a flat surface and a welldefined electronic band structures, seem not adapted. Instead, a model based on the local chemical environment, as suggested by K. Wittmaack, 16 should be able to describe the processes. The variation of the useful yield of the different secondary silicon clusters supports this even more (Figure 4). Here the useful yield is defined as: = !"!#"! /

4


%"& '(""!)! , where !"!#"! is the number of Si ions or Si cluster ions which gets detected during the SIMS measurement and %"& '(""!)! is the number of Si atoms which gets sputtered from the analysed area. The latter is calculated after measuring the crater volume with a profilometer. It should also be noted that only the data from Figure 3 where net sputtering is observed has been used to calculate the useful yields. For the conditions where deposition of matter has been observed, the number of sputtered silicon atoms cannot be calculated. Hence, the data points with the lowest current densities cannot be found in Figure 4.

10-2 10

Ip: 41 pA

-3

SiSi2-

Si3-

Si3-

4 5 6

Si4-

Si Si

10

100 pA

SiSi2Si

10-4

Useful yield

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Si5Si6-

-5

10-6 10-7 10-8 10-4

Increasing Cs concentration

10-3

10-2

10-1

Current density (pA/µm2) Figure 4: Useful yield variations of different secondary silicon clusters as a function of current density, both for a primary current of 41 pA and 100 pA. The cesium deposition rate was the same for all data points. For the conditions where deposition of matter has been observed, the number of sputtered silicon atoms cannot be calculated. Hence, the data points with the lowest current densities cannot be found in this picture.

Over the whole range of primary current densities of Figure 4, all intensities show the same monotonous increase in useful yield of three to four orders of magnitude. The same enhancement has already been obtained for gallium bombardment,8 as well as helium and neon bombardment10 combined with simultaneous cesium flooding. Hence, similar ionisation mechanisms can be assumed for the different species. As expected, the difference in secondary ion intensities for the two primary ion currents is also visible for the useful yields.

*+ , bombardment combined with simultaneous cesium flooding Previous experiments have been repeated to study the influence of cluster size and composition on the enhancement of secondary ion yields. For primary ion beam combined with cesium flooding, the general trend stays the same than for (Figure 5). However, the sputter rate under bombardment is extremely low, making it difficult to sputter through the native oxide layer and leading to a partial oxidation of the irradiated area. This explains also the presence of

many oxygen-containing secondary cluster ions, e.g. -. , - , - . , - . and - . (data not shown). It should be noted that the secondary ion intensities shown in Figure 5 have been taken from the equilibrium regime, i.e. after obtaining constant secondary ion intensities. Depending on the secondary ion cluster, secondary ion intensities normalised to the primary current increase by about one order of magnitude. produces a behaviour similar than the beam, except that the optimum secondary ion intensities do not seem to be reached. Probably a higher Cs deposition rate would have been required. For , the highest secondary ion intensity of 5 × 102 counts/C is obtained for a current density of about 10-4 pA/µm2. For bombardment, much higher current densities are required to reach the opti mum enhancement. Although sputter rates are much lower than those obtained with , a higher Cs deposition rate seems to be required for an optimum enhancement. Probably clusters are more effective in removing the cesium at the sample surface than the beam. The larger size of the cluster seems to favour also the emission of larger sec ondary cluster ions. For , the highest secondary ion intensi it is obtained for ty was obtained for clusters, for the larger / clusters. For , using a high primary ion current is even more . The highest normalised intensity does important than for not go beyond 5400 counts/C, which is about an order of mag nitude lower than for . At the same time, the intensity stays all the time at the background level, which is around 10 counts/C. The same is true for other secondary ion intensities lying below this threshold. This observation is a related to the much lower sputtering yield of on silicon than for bombardment.

104


Page 5 of 19

Increasing Cs concentration

Si Si2 Si3

103

Si4 Si5 Si6

102

Si7 Si8

101 65 pA 10

0

10-4

10-3

10-2

10-1

Current density (pA/µm2) Figure 5: Secondary ion intensities of different silicon clusters normalised to primary current as a function of current density for bombardment combined with simultaneous Cs flooding. The primary current was set to 65 pA. During the experiments, the temperature of the cesium reservoir was kept constant at 50°C.

5


Analytical Chemistry Useful yields could not be calculated for , because the craters were too shallow to get any reliable data on crater dimensions. Estimating any values is difficult because sputter rates for and bombardment differ significantly. Nevertheless, for and primary ion clusters the secondary ion signals increase by about 1 order of magnitude, so that the enhancement of the useful yields might be similar to those obtained with the beam. For monatomic ion bombardment combined with simultaneous cesium flooding, it was already shown that the enhancement can be obtained on metallic or semiconductor samples.8 As the general trend is the same for cluster primary ions, it can be expected that the results obtained in this study on silicon can be generalised to other semiconductors or metals.

*+ , bombardment with cesium flooding applied to the depth profiling of PC

Intensity (counts/frame)

106

a)

105

104

103

102

Intensity (counts/frame)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

b)

105

C6H5O

104

C8H5O C9H9O 167 195 C15H15O

103

10

359 405

2

0

2000

4000

6000

Time (s) Figure 6: Depth profiles in PC using a) bombardment only, b) bombardment combined with simultaneous cesium flooding. For the depth profile with cesium flooding b), the temperature of the cesium reservoir was set to 115°C.

In addition to the experiments carried out on silicon, the combination of cluster bombardment and simultaneous cesium flooding was also tested on PC. Depth profiling was carried out only with the because it is the state-of-the-art beam for this kind of applications. A first advantage of the cesium flooding is the removal of the charging effect and the opera-

Page 6 of 19

tion of the flood gun which is no longer needed. In addition, overall secondary ion intensities are higher with flooding than without. When using no cesium flooding, secondary ion intensities increase up to a maximum in the pre-equilibrium regime (Figure 6a). Afterwards, some fragmentation leads to lower intensities in steady state conditions. In general, intensities in steady state are about six times lower than at the beginning of the depth profile. When switching on cesium evaporation, the increase of secondary ion intensities in the pre-equilibrium regime stays the same. The maximum intensity reached by the different secondary ion clusters stays also the same. However the decrease towards steady state conditions is no longer observed but all signals stay at the maximum level (Figure 6b). Hence, the yields of all secondary ion clusters have been enhanced by about a factor six. Maximum secondary ion intensities which stay at the same maximum level than without flooding suggests also that the cesium flooding is not enhancing the ionisation probability significantly, but removes the charging effect efficiently and / or reduces the fragmentation caused by the bombardment. Hence, the effect could be similar to low-energy depth profiling of polymers observed by Houssiau et al.20,21 where the cesium prevents radicals formed under ion bombardment to react with each other and to recombine. In general, the enhancement caused by the cesium flooding is also far less on the polymer sample than on inorganic samples. The main advantages are the removal of the charging effect and the disappearance of the secondary ion decay towards steady-state conditions.

Perspective for cell imaging Imaging of cells is one important application for ToF-SIMS analysis with cluster beams. Some preliminary tests combing bombardment with simultaneous cesium flooding on freeze dried fibroblast cells, cultured on indium tin oxide coated glass microscope slides (Bruker GmbH), show some encouraging results (Figure 7). Without flooding, the different primary features in the cell can be seen. Using the same current and dwell time, but putting the cesium reservoir at a temperature of 115°C, the secondary ion intensities could be improved significantly while preserving the spatial information. The migration of molecules seems also to be avoided. In general, the enhancement is largely mass dependent and changes from a factor 2 to a factor 20 (Figure 8). The enhancement is less for lower temperatures of the cesium reservoir. A higher enhancement at more elevated temperatures of the reservoir cannot be excluded, i.e. at higher cesium deposition rates, but the current setup did not allow for higher temperatures. At this point it is already clear that much higher cesium desposition rates are required than on inorganic samples. On silicon, optimum conditions were obtained with the cesium reservoir at 50°C and the current ranging between 41 and 100 pA. This is related to the much higher sputter yields on organic matter. Figure 8 shows the enhancement for some selected frames. The evolution of secondary ion intensities with fluence is shown in Figure 9. Without cesium flooding, most signals show an increase in the pre-equilibrium regime and stay mostly constant afterwards. Signals of masses 61, 77 and 89 show no major variation right from the beginning. With cesium flooding, the increase of secondary ion intensities in the pre-

6


equilibrium regime is more pronounced for some masses and less for others. The intensity of mass 72 is significantly enhanced and the signal variation in the pre-equilibrium regime reduced. For the intensity at mass 166, the signal variation in the pre-equilibrium regime is more pronounced which leads to a significant enhancement of the intensity in steady-state conditions. For the intensity at mass 185, the shape of the curve remains largely the same and all data points get similarly enhanced. The signals of the three masses, which were mostly constant without cesium flooding, are not much affected by the flooding. Thus, the effect of the cesium flooding is largely mass-dependent.

hanced still needs to be investigated. The same is true for the influence of cesium on the mobility of salts or molecules in cells. Repeating the same experiments with is also of interest because larger molecular fragments could be obtained than with the clusters. layer 1 layer 4 layer 7 layer 11

20

15

10

5

48 49 50 60 61 63 72 73 74 77 79 84 89 97 131 147 166 167 185 201 262

0

Mass (m/z) Figure 8: Enhancement induced by the cesium flooding for different masses as a function of depth (layer) for the cells shown in Figure 7. Layers 1, 4, 7 and 11 correspond to 1.17 x 1013, 4.68 x 1013, 8.19 x 1013 and 1.29 x 1014 ions/cm2. The enhancement is calculated relative to the data without flooding. 108

Intensity (counts)

a) No Cs

Figure 7: Cell images of three different masses with and without cesium flooding, with the temperature of the cesium reservoir set 0 0 to 115°C: a) = 23 without Cs flooding, b) = 23 with flooding, c) 0

0 1

1

= 42 without flooding, d) 0

0 1

1

= 42 with flooding, e)

= 456 without flooding, and f) = 456 with flooding. Image 1 1 size is 256 µm × 256 µm, 128 × 128 pixels. The primary fluence to acquire the images of a given layer was 1.17 × 1013 ions/cm-2. The images correspond to layer 11 after sputtering the cells to an accumulated fluence of 1.17 × 1014 ions/cm-2. Maximum counts per pixel are shown for each image using a linear colour scale.

Hence, cluster ion bombardment combined with cesium flooding enhances secondary ion yields for cells, but to determine the optimum analysis conditions and to understand which secondary ions gets mostly enhanced, a far more detailed study is required. It can be expected that cesium interacts and reacts differently with the various molecules in cells, depending on their chemical composition. How and by how much the ion yields of the different molecules can be en-

107 106 105 104 b) With Cs

Intensity (counts)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Enhancement

Page 7 of 19

107 106 61 72 77 89 131

5

10

104 0.0

4.0x1013

8.0x1013

166 167 185 262

1.2x1014

Fluence (ions/cm2)

Figure 9: Intensity evolution of some clusters of Figure 8 as a function of depth, a) without cesium flooding, b) with cesium flooding.

7



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

(1)

Conclusions For the first time cluster ion bombardment has been combined with simultaneous cesium flooding. As cluster ion bombardment perturbs the sample surface much more than monatomic ion bombardment, it could be expected that cesium flooding has no or a reduced effect on secondary ion mecha nisms. Our results show the opposite: and bombardment combined with simultaneous cesium flooding produce exactly the same enhancement on metallic and semiconductor samples than , , . and irradiation. For monatomic ion bombardment, the enhancement does not de pend on the primary ion species. For and bombardment both cluster ions produce the same trend. However, the larger clusters favour the emission of larger secondary ions. Hence, it can be assumed that the ionisation mechanisms must be similar for monatomic and cluster primary ions on semiconductor and metal samples and results obtained for and bombardment combined with cesium flooding are also valid for cluster ion bombardment. The primary ion cur and bomrent is also an important parameter. For bardment, reliable data could only be obtained with a current of 100 pA. For lower currents but identical current densities, the useful yields are lower by about an order of magnitude. For polycarbonate, cesium flooding removes the charging effect leading to an easier tuning of the instrument and making the use of the flood gun unnecessary. In addition it increases the secondary ion intensities of all cluster ions by about a factor 6 and the decrease of secondary ion intensities after the transient regime is removed. The enhancement corresponds rather to the removal of the charging effects and / or reduced fragmentation than to an increase of the ionisation probability of the different secondary ions. Preliminary results on biological cells show that secondary ion yields can be enhanced by a factor 2 – 20 when combining bombardment with simultaneous cesium flooding. Most likely the yields are not fully optimised because the results were obtained with the highest possible cesium deposition rate the experimental setup allowed for. Higher temperatures were not possible with the current setup, but will be possible for future experiments. Other parameters, like the influence of molecule composition on the enhancement of secondary ion yields need to be investigated too in order to fully control the flooding process and apply it successfully to the imaging of cells.

(2) (3) (4)

(5) (6) (7) (8) (9) (10) (11) (12) (13) (14) (15) (16) (17) (18) (19) (20) (21) (22) (23)

(24) (25) (26) (27)

AUTHOR INFORMATION

(28)

Corresponding Author

(29)

* E-mail : [email protected] The authors declare no competing funding interest.

(30) (31) (32)

ACKNOWLEDGMENT

(33)

Notes

The MS analysis was performed at the National Center for Imaging Mass Spectrometry, Sweden, part of the GU/Chalmers Bioanalytical Center.

(34)

REFERENCES

(36)

(35)

Page 8 of 19

Storms, H. A.; Brown, K. F.; Stein, J. D. Anal. Chem. 1977, 49(13), 2023. Stingeder, G. Anal. Chem. 1994, 297, 231. Bernheim, M.; Slodzian, G. Journal de Physique Lettres 1977, 38 (15), L325-L328. Bernheim, M.; Rebière, J.; Slodzian, G. Negative Ion Emission from Surfaces Covered with Cesium and Bombarded by Noble Gas Ion. Benninghoven, A., Evans, C. A., Powel, R. A., Shimizu, R., Storms, H. A., Eds.; Springer: London, 1979; p ;40. Wirtz, T.; Migeon, H. N. Appl. Surf. Sci. 2004, 231-232, 940-944. Wirtz, T.; Migeon, H. N. Appl. Surf. Sci. 2004, 222, 186197. Philipp, P.; Wirtz, T.; Migeon, H. N.; Scherrer, H. Int. J. Mass Spectrom. 2007, 261 (2-3), 91-99. Philipp, P.; Wirtz, T.; Migeon, H. N.; Scherrer, H. Int. J. Mass Spectrom. 2006, 253 (1-2), 71-78. Philipp, P.; Wirtz, T.; Migeon, H. N.; Scherrer, H. Appl. Surf. Sci. 2006, 252 (19), 7205-7207. Pillatsch, L.; Vanhove, N.; Dowsett, D.; Sijbrandij, S.; Notte, J.; Wirtz, T. Appl. Surf. Sci. 2013, 282 (0), 908-913. Wittmaack, K. Surf. Sci. Rep. 2013, 68 (1), 108-230. Gnaser, H. Phys. Rev. B 1996, 54(23), 16456. Gnaser, H. Nucl. Instrum. Methods Phys. Res. B 1999, 149 (1-2), 38-52. van der Heide, P. A. W.; Azzarello, F. V. Surf. Sci. 2003, 531, L369-L377. Philipp, P.; Wirtz, T.; Migeon, H. N.; Scherrer, H. Int. J. Mass Spectrom. 2007, 264 (1), 70-83. Wittmaack, K. Anal. Chem. 2014, 86 (12), 5962-5968. Garrison, B. J.; Postawa, Z. Mass Spectrom. Rev. 2008, 27 (4), 289-315. Mahoney, C. M. Mass Spectrom. Rev. 2010, 29 (2), 247293. Fletcher, J. S.; Lockyer, N. P.; Vickerman, J. C. Mass Spectrom. Rev. 2011, 30 (1), 142-174. Mine, N.; Douhard, B.; Brison, J.; Houssiau, L. Rapid Commun. Mass Spectrom. 2007, 21 (16), 2680-2684. Houssiau, L.; Douhard, B.; Mine, N. Appl. Surf. Sci. 2008, 255, 970-972. Wehbe, N.; Houssiau, L. Anal. Chem. 2010, 82 (24), 10052-10059. Brison, J.; Mine, N.; Wehbe, N.; Gillon, X.; Tabarrant, T.; Sporken, R.; Houssiau, L. Int. J. Mass Spectrom. 2012, 321−322 (0), 1-7. Fletcher, J. S.; Rabbani, S.; Henderson, A.; Blenkinsopp, P.; Thompson, S. P.; Lockyer, N. P.; Vickerman, J. C. Anal. Chem. 2008, 80 (23), 9058-9064. Angerer, T. B.; Blenkinsopp, P.; Fletcher, J. S. Int. J. Mass Spectrom. 2015, 377 (0), 591-598. Philipp, P.; Wirtz, T.; Migeon, H. N.; Scherrer, H. Appl. Surf. Sci. 2004, 231-232, 754-757. Lyon, I.; Henkel, T.; Rost, D. Appl. Surf. Sci. 2010, 256 (21), 6480-6487. Yang, J.; Xu, W.; Xiao, W. Journal of Molecular Structure: THEOCHEM 2005, 719 (1−3), 89-102. Tam, N. M.; Nguyen, M. T. Chem. Phys. Lett. 2013, 584 (0), 147-154. Gnaser, H. Appl. Surf. Sci. 2003, 203-204, 78-81. Philipp, P; Wirtz, T.; Migeon H. N. unpublished results Gillen, G.; Batteas, J.; Michaels, C. A.; Chi, P.; Small, J.; Windsor, E.; Fahey, A.; Verkouteren, J.; Kim, K. J. Appl. Surf. Sci. 2006, 252 (19), 6521-6525. Lee, J. L. S.; Seah, M. P.; Gilmore, I. S. Appl. Surf. Sci. 2008, 255 (4), 934-937. Kozole, J.; Winograd, N. Appl. Surf. Sci. 2008, 255 (4), 886-889. Krantzman, K. D.; Kingsbury, D. B.; Garrison, B. J. Appl. Surf. Sci. 2006, 252 (19), 6463-6465. Barry, P. R.; Philipp, P.; Wirtz, T. J. Phys. Chem. C 2014, 118 (7), 3443-3450.

8


Page 9 of 19 (37)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

(38) (39)

Analytical Chemistry Philipp, P.; Barry, P.; Wirtz, T. Surf. Interface Anal. 2014, 46 (S1), 7-10. Ngo, K. Q.; Philipp, P.; Kieffer, J.; Wirtz, T. Surf. Sci. 2012, 606 (15−16), 1244-1251. Barry, P. R.; Philipp, P.; Wirtz, T.; Kieffer, J. J. Mass Spectrom. 2014, 49 (3), 185-194.

9



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 10 of 19

Insert Table of Contents artwork here

10


Page 11 of 19

1 2 3 4 5 6 7 8 9 10



Analytical Chemistry Page 12 of 19

1 ±3. 8mm 2 3 4 5 6 7 ±2. 5mm ACS Paragon Plus Environment 8 9 10

Page 13 of 19

N o r m a lis e d in te n s ity ( c o u n ts /C )


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


a )

1 0

5

In c r e a s in g C s c o n c e n tr a tio n

1 0

4

1 0

3

1 0

2

1 0

1

Ip = 4 1 p A

1 0

0

1 0

5

b )

S iS i2

4

1 0 3

1 0 2

1 0 1

1 0

Ip = 1 0 0 p A

1 0

-4

-

S i4

-

S i5

-

S i6

-

S i7

-

S i8

-

S i9

-

S i1

0

1 0

S i3

1 0

-3

1 0

-2


C u r r e n t d e n s ity ( p A /µ m

2

1 0

)

-

-1

0

-


1 0 1 0 1 0

U s e fu l y ie ld

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45

Page 14 of 19

-2

Ip: 4 1 p A S iS i2

-3

-4

1 0

-5

1 0

-6

1 0

-7

1 0

-8

1 0 0 p A S iS i2 -

S i3

-

S i3

-

S i4

-

S i4

-

S i5

-

S i5

-

S i6

-

S i6

-


1 0

-4

1 0

-3

1 0

-

-2



2

1 0

)

-1

Page 15 of 19

4

1 0


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45



S i S i2 S i3

1 0

3

S i4 S i5 S i6

1 0

2

S i7 S i8 1

1 0

6 5 p A 1 0

0

1 0

-4

1 0

-3

1 0


-2


2

1 0

)

-1

In te n s ity ( c o u n ts /fr a m e ) In te n s ity ( c o u n ts /fr a m e )

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

1 0

6

1 0

5

1 0

4

1 0

3

1 0

2

1 0

5

1 0

4

a )


Page 16 of 19

b )

C 6

H 5

O

C 8

H 5

O

C 9

H 9

O

1 6 7 1 9 5

1 0

3

C

1 0

1 5

H

1 5

3 5 9 4 0 5 2

0

2 0 0 0

4 0 0 0


T im e ( s )

6 0 0 0

O

Page 17 of 19

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34




la y la y la y la y

2 0

e r 1 e r 4 e r 7 e r 1 1

1 5

1 0

5

M a s s (m /z )


1 3 1 1 4 7 1 6 6 1 6 7 1 8 5 2 0 1 2 6 2

9 7

8 9

8 4

7 9

7 7

7 4

7 3

7 2

6 3

6 1

6 0

5 0

4 9

0 4 8

E n h a n c e m e n t

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45

Page 18 of 19

Page 19 of 19

8

1 0

In te n s ity ( c o u n ts )

a ) N o C s 1 0

7

1 0

6

1 0

5

1 0

4

b ) W ith C s

In te n s ity ( c o u n ts )

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51


1 0

7

1 0

6

6 1

1 6 6 1 6 7 1 8 5 2 6 2

7 2 5

1 0

7 7 8 9

1 0

1 3 1 4

0 .0

4 .0 x 1 0

1 3

8 .0 x 1 0

1 3


F lu e n c e ( io n s /c m

2

1 .2 x 1 0

)

1 4

Significant Enhancement of Negative Secondary Ion Yields by Cluster

Recommend Documents