Chemical Vapor Deposition of Azidoalkylsilane ... - ACS Publications

Subscriber access provided by the University of Exeter

Article

Chemical Vapor Deposition of Azidoalkylsilane Monolayer Films Rita Vos, Cedric Rolin, Jens Rip, Thierry Conard, Tim Steylaerts, Maria Vidal Cabanilles, Karen Levrie, Karolien Jans, and Tim Stakenborg Langmuir, Just Accepted Manuscript • DOI: 10.1021/acs.langmuir.7b04011 • Publication Date (Web): 30 Dec 2017 Downloaded from http://pubs.acs.org on December 31, 2017

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.

Langmuir 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 43 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

Langmuir

Chemical Vapor Deposition of Azidoalkylsilane Monolayer Films AUTHOR NAMES. Rita Vos,1,* Cedric Rolin, 1 Jens Rip, 1 Thierry Conard, 1 Tim Steylaerts, 1 Maria Vidal Cabanilles, 1 Karen Levrie, 1,2 Karolien Jans1 and Tim Stakenborg1 (*) Author to whom correspondence should be addressed.

AUTHOR ADDRESS. 1

imec, Kapeldreef 75, B-3001 Leuven, Belgium.

2

Katholieke Universiteit Leuven, Kasteelpark Arenberg 10, B-3001 Leuven, Belgium.

KEYWORDS. Self-assembled monolayer, vapor-phase, silanization reaction, click chemistry, biointerface.

ACS Paragon Plus Environment

1

Langmuir 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 2 of 43

ABSTRACT. N3-functionalized monolayers on silicon wafer substrates are prepared via the controlled vaporphase deposition of 11-azidoundecyltrimethoxysilanes at reduced pressure and elevated temperature. The quality of the layer is assessed using contact angle, ATR-FTIR and ellipsometry measurements. At 60 °C, longer deposition times are needed to achieve monolayers with similar N3-density compared to depositions at 145 °C. The monolayers formed via the vapor phase are denser compared to a solvent-based deposition process. ATR-FTIR measurements confirm the incorporation of azido-alkyl chains in the monolayer and the formation of siloxane bridges with the underlying oxide at both deposition temperatures. XPS shows that the N3-group is oriented upwards in the grafted layer. Finally, the density was determined using TXRF after click reaction with chloro-hexyne and amounts to 2.5E14 N3-groups/cm2. In summary, our results demonstrate the formation of a uniform and reproducible N3-containing monolayer on silicon wafers hereby providing a functional coating that enables click reactions at the substrate.


2

Page 3 of 43 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

Langmuir

TEXT. Introduction Over the last decades, the research and development of biosensor and lab-on-chip devices has been expanding rapidly due to their great potential and benefits for point-of-care diagnostics and various other bio-analytical applications.1-4 For the fabrication of such devices, a reliable immobilization of bioreceptors on the sensor surface is essential and a variety of coupling strategies exists to immobilize biomolecules on a surface.5 In general, a more robust, reproducible, and potentially even oriented coupling is obtained using a covalent immobilization strategy.6,7 To achieve such covalent coupling, functional groups that can react with the bioreceptor molecule directly or via a cross-linker need to be introduced. The formation of self-assembled monolayers (SAMs) using organosilanes is widely used to tailor the surface properties of OH-bearing substrates such as silica.8,9 Solvent-phase depositions are widely established and relatively well understood but very sensitive to both surface adsorbed water and traces of moisture in the ambient leading to uncontrolled polymerization and multilayer formation.10-14 The deposition via the vapor phase can be better controlled as moisture variations can be reduced or even eliminated with dedicated vacuum equipment resulting in a better reproducibility and quality of the coating. Moreover, the total chemical consumption is lower and no environment unfriendly or harmful solvents are required thereby significantly reducing chemical cost and waste production. Because vapor-phase silanizations produce highquality and reproducible monolayers in a cost-effective manner, they are easy to integrate in a high-throughput production sequence. For vapor-phase depositions, relatively low system background pressures often in combination with high temperatures are required to reach sufficient partial pressures of the source material in


3

Langmuir 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 4 of 43

an enclosed chamber. Heating has the additional benefit to initiate the surface attachment reactions.15 From the different organosilanes that are suitable for vapor-phase deposition, amino (-NH2), thiol (-SH) and epoxy-containing molecules are mostly selected, as they provide a functional group that can be used for bioreceptor molecule coupling. 16-18 However, these functional groups have a limited stability because they are prone to oxidation or hydrolysis. In addition, they are mainly used for random couplings of the biomolecules on the biosensor surface. So, there is still a clear need for a robust and stable coupling layer that allows also a directed or site-selective immobilization. Recently, there is an increased interest for bioorthogonal ‘click’ reactions that meet strict requirements for rate, selectivity and biocompatibility.19,20 Most commonly used for these coupling reactions are the Staudinger ligation between azides (N3) and tertiary phosphines or the azide-alkyne Huisgen cycloaddition. In both cases, N3 react very selectively and with high yield under relatively mild conditions. Moreover, the N3- functional group is sufficiently stable in aqueous biocompatible environments. All of this makes N3-based click reactions a powerful tool for bioconjugation.21 The vapor-phase deposition of N3-containing organosilanes has been presented previously.22 Because of the broad application range of N3-based conjugations and the resulting necessity for a reproducible deposition process that can be implemented in a manufacturing environment, a more thorough investigation of the vapor-phase reaction of N3-silanes is justified. Therefore, we use a unique combination of different techniques to analyze in detail the uniformity, thickness, roughness, composition, density and reactivity of the N3-SAM and we describe the effect of process parameters on the layer quality and density. To our knowledge, our work results in the first peer-reviewed paper of such layers deposited via the vapor phase. To illustrate the use of a


4

Page 5 of 43 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

Langmuir

vapor-phase deposited N3-SAM for the covalent immobilization of bioreceptor molecules, we present a prostate specific antigen (PSA) recognition bio-assay using Surface Plasmon Resonance (SPR) measurements.

Materials and methods Chemicals 11-Azidoundecyltrimethoxysilane (N3-silane) was purchased from ABCR GmbH. 6-chloro-1hexyne, tris[(1-benzyl-1H-1,2,3-triazol-4-yl)methyl]amine (TBTA), tetrakis(acetonitrile)copper(I) hexafluorophosphate (Cu-catalyst), (+)-sodium L-ascorbate (SA), dimethyl sulfoxide (DMSO), toluene and N,N-diisopropylethylamine (DIPEA) were purchased from Sigma-Aldrich. VLSI-grade acetone and isopropyl alcohol (IPA) were obtained from Honeywell. Dibenzylcyclooctyne polyethylene glycol N-hydroxysuccimide ester (DBCO-PEG4NHS) was purchased from Jena Bioscience. The prostate specific antigen (PSA) protein was obtained from Scipac and the monoclonal mouse antibody against PSA (Anti-PSA) from Fujirebio Diagnostics. Substrates P-type silicon wafers with resistivity >1 Ohm.cm and crystal orientation (purchased from MEMC Electronic Materials) were used as substrates for vapor-phase deposition. The wafers were pre-cleaned using a two-step cleaning sequence with an O3-based step for oxidation and diluted HF for subsequent oxide removal.23 A clean chemical oxide was regrown using ozonated deionized water (O3-DIW) in the final rinsing step. Experiments were done either on full wafer


5

Langmuir 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 6 of 43

or on Si-wafer pieces of 3 x 3 cm prepared using a diamond-tip scriber pen for dicing. Particles generated during the dicing are removed using sonication in IPA followed by drying using a stream of N2-gas. Immediately before the coating, the substrates received an additional 15 min. UV/O3-clean (UVO-cleaner, Jelight Company Inc.) to remove any air-borne organic contamination and to render a perfectly clean Si/SiO2-substrate with an oxide thickness of ~2.2 nm and covered with Si-OH groups (Si/SiO2). SAM deposition Vapor-phase silanisation was performed using a home-built stainless-steel vacuum chamber large enough to hold a 200 mm wafer. The entire vacuum chamber with a total volume of ~1.2 dm3 was heated in an oven (WU 6100, Thermo Scientific) with a temperature controller to ensure a uniform and stable temperature. Two different temperatures were evaluated, namely 60 and 145 °C. After loading the samples, the required amount of the N3-silane (i.e. 100 µl for 1 hour and shorter deposition times and 200 µl for longer depositions unless indicated otherwise) was dispensed in a small aluminum cup placed in the vacuum chamber and the pressure was lowered to 25 mbar using a vacuum pump (Buchi V-700). The samples were exposed to the generated N3-silane vapors during a certain time after which the chamber was purged 3x with N2 prior to unloading the samples. To determine the effect of subsequent depositions on the same sample, this procedure was repeated as such and with an intermediate immersion in DIW during 5 min. followed by N2 blow-dry. Coating was carried out on 200mm wafers or on Si-wafer pieces that were placed on a 200 mm carrier wafer that was also cleaned by UV/O3 before the deposition. To assess the uniformity of the coating, always three Si-wafer coupons of 3 x 3 cm were positioned at different locations (centre – top – right) on the carrier wafer. As a reference,


6

Page 7 of 43 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

Langmuir

the N3-silane was deposited via the solvent phase using toluene. The coating was done in a glovebox under controlled N2-atmosphere. Clean Si samples were immersed during 24 hours in toluene containing 2 v/v-% of the N3-silane and 0.03 M of DIPEA. After the deposition, the samples were removed from the toluene and extensively rinsed with acetone and IPA during 10 min. To further crosslink the attached N3-silanes at the surface, the samples were baked during 30 min at 110 °C under N2-atmosphere. The N3-silane and samples coated with N3-SAM are protected from light during deposition as alkyl azides can undergo photolytic decomposition.24 At room temperature under normal laboratory light, the decomposition is slow (~1 % per day) 25 but as a precaution, the N3-coated samples are stored in darkness as much as possible. Characterization of functionalized substrates Contact angle (CA) measurements were performed as a quick method to assess the overall quality of the SAM using an OCA 20 Contact Angle System (DataPhysics) in static mode with sessile drops of 1 µL. Each reported contact angle is the average of at least 3 measurements taken on different positions. The thickness of the layers was checked using a Sentech SE400adv-PV laser ellipsometer (SENTECH Instruments GmbH) at a wavelength of 632.8 nm. As the refractive indices of SiO2 and organic monolayers are very similar, a one-layer model with a refractive index n = 1.5 and extinction coefficient k = 0 was used to determine the total thickness of the native oxide and the self-assembled N3-silane layers. The actual layer thickness was calculated by subtracting the oxide thickness of a reference sample that received only the cleaning sequence without coating.


7

Langmuir 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 8 of 43

To verify the composition and the presence of the chemical groups on the surfaces, Attenuated Total Reflectance Fourier Transform Infrared Spectroscopy (ATR-FTIR) and X-ray Photoelectron Spectroscopy (XPS) measurements were used. ATR-FTIR measurements were performed using a Thermo Scientific Nicolet 6700 spectrometer (Thermo Fisher Scientific Inc.) with a VariGATR grazing angle accessory (Harrick Scientific Products Inc.). The incident angle was 65 degrees with p-polarized light and a Ge-crystal as internal reflection element. Spectral acquisition was done with a mercury cadmium telluride (MCT/A) detector with liquid nitrogen cooling at a spectral resolution of 4 cm-1 and 100 scans. To enable sample-to-sample comparison, the measured spectra are normalized towards the band at 900cm-1 that is related to the longitudinal optical (LO) asymmetric stretching of germanium oxide at the Ge-crystal and that is caused by the partial reflection and enhancement of the evanescent wave at the Si-substrate. 26,27 XPS measurements were carried out in Angle Resolved mode using a Theta300 system (Thermo Instruments). A total of 16 spectra were recorded at exit angles between 22 and 78 degrees as measured from the normal of the sample. The measurements were performed using a monochromatized Al Kα X-ray source (1486.6 eV) and a spot size of 400 micrometer. To avoid degradation of the layer by exposure to X-rays as has been reported in literature,28 short measurements were executed on 5 positions on the surface (starting with ca. 3-4 min N1s and C1s at each position). Data processing was performed on the sum of the 5 spectra recorded for each element.


8

Page 9 of 43 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

Langmuir

Quantification of the N3-group density at the surface was done on full 200mm sized wafers after reaction with chloro-hexyne using click chemistry (see Scheme I). To that end, the samples were incubated for 1 h with a solution of 0.44 mM Cu-catalyst, 0.44 mM TBTA, 0.6 mM SA and 150 µM chloro-hexyne in a mixture of DMSO/H2O (4:5 v/v-ratio). After incubation, the substrates were thoroughly rinsed with DMSO, IPA and DIW. As the Cu-catalyst used for the click reaction might leave residual CuClx precipitates, an additional diluted HCl rinse (pH2; 10 min) was applied to remove any remaining Cu species followed by another short DIW rinse before N2 blow dry. The background Cl-contamination levels on our samples was therefore also determined by including several reference samples such as N3-coated samples without any treatment or with diluted HCl rinse, and on a N3-coated sample after click reaction without the catalyst. The resulting chloride-surface concentration was measured with total reflection X-ray fluorescence (TXRF) using a FEI-Atomika 8300W system equipped with a Mo X-ray tube. Measurements were performed in the Mo Kα excitation mode with the X-ray incident angle set at 2.3 mrad and an acquisition time of 1000 sec. After click reaction, measurements were done over

5 different positions on the wafer and average values for Cl are reported.


9

Langmuir 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 43

Scheme I. Cu-catalyzed click reaction between chloro-hexyne and N3-groups. Bio-assay The immobilization of the anti-PSA capture antibodies and the detection of PSA were carried out on a Biacore3000 system (GE Healthcare, United Kingdom) using SiO2-covered Au substrates as described previously.29 The substrates were cleaned with 15 min. UV/O3-clean prior to N3-SAM vapor deposition at 145 °C and 25mbar as described above. The capture antibodies were immobilized using DBCO-PEG4-NHS as a cross-linker (see Scheme II) and the binding of PSA was monitored.

Scheme II. Covalent conjugation of proteins using DBCO-PEG4-NHS as cross-linker after strain-promoted click reaction between the DBCO and the N3-group.


10

Page 11 of 43 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

Langmuir

Results and discussion (a) Optimization of vapor-phase deposition parameters. Time dependence of monolayer formation at different temperatures. To reach sufficiently high concentrations in the vapor phase, heating under reduced pressure is necessary as the N3-silane used in this investigation has a high boiling point at atmospheric pressure outside of the thermal stability range of azido-alkyl compounds. For the N3-silane, a boiling point around 100-105 °C has been reported at a pressure of 0.053 mbar.30 This can be extrapolated in good approximation to ~350 °C at 1 bar.31 This temperature is clearly too high as N3-alkyl compounds will decompose at temperatures above 180 °C with the release of N2 and the formation of a nitrene intermediate.32 To avoid decomposition, the temperature of the vaporphase reactions in our work was limited to 145°C. At a pressure of 25 mbar, the temperatures are still below the boiling point of ~212 °C allowing a relatively slow and therefore controlled evaporation of the N3-silane molecules. The time dependence of monolayer formation at the surface was monitored at two different temperatures by contact angle, ellipsometry and ATR-FTIR measurements and is shown in Figure 1. The contact angles obtained at 145 °C saturate after 1 hour deposition and are in good agreement with layers deposited via the solvent phase in-house (see supplementary info, Figure S.1) as well as those published previously in literature.33 The 1 hour deposition yielded very uniform results across a 200 mm wafer substrate as indicated by the small standard deviation


11

Langmuir 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 12 of 43

obtained by averaging values from samples placed at the top/centre/right of the carrier wafer as shown in Figure 1.

Figure 1. Time dependence of the N3-SAM formation after vapor-phase deposition at 60 °C () and 145 °C () as monitored using contact angle (a), ellipsometric thickness (b) and area


12

Page 13 of 43 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

Langmuir

integrated intensity of the N3-stretching at 2100 cm-1 (c). The dashed line in each graph represents the respective value for N3-SAM prepared via solvent phase.


13

Langmuir 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 14 of 43

At 60 °C it took much longer before saturation was achieved. The measured contact angles also showed more variation depending on the position of the sample in the reaction chamber. Only at very long deposition times, such as 24 hours, the contact angles were more uniform across the wafer with similar values compared to the vapor-phase process at 145 °C. This time-dependent non-uniformity confirms that the evaporation and transport of silane molecules in the vapor phase plays an important role to get uniform surface coatings. Besides higher concentrations of the N3-silane in the vapor phase, the faster grafting at higher temperatures might also be related to a higher reactivity of the silane molecules.15,34 The thickness of the layers as determined by ellipsometry confirms these trends and similar thicknesses are measured for saturated monolayers both at 60 and 145 °C. More information on the formation of a monolayer was obtained by ATR-FTIR measurements. The ATR-FTIR spectrum after silane deposition shows a clear peak at ~2100 cm-1 that is related to the asymmetric stretch vibration of the N3-group. The integrated area can be used to monitor the relative surface concentration. The results depicted in Figure 1 demonstrate that the surface concentration of N3-groups saturated after 1 hour at 145 °C while much longer deposition times were necessary to achieve a comparable surface concentration at a low temperature of 60 °C. For the monolayers formed in the vapor phase, higher N3-concentrations are measured compared to N3-SAM’s deposited via the solvent phase although they display similar contact angles and thicknesses. Similar observations have been done in the past for instance for the grafting of alkylsilane layers where a denser and more ordered SAM is formed via the vapor phase then when using toluene as a solvent.35,36


14

Page 15 of 43 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

Langmuir

Figure 2. Contact angle (a), ellipsometric thickness (b) and area integrated intensity of the N3stretching at 2100 cm-1 (c) for N3-SAMs formed via the vapor phase at 145 °C after 1 hour deposition with 100 µl silane (x), after 1 hour deposition with 200 µl silane (x’), after 2 hours deposition with 200 µl silane (2x’), after 2 depositions of 1 hour using each with fresh 100 µl silane (x + x) and after the latter with an intermediate DIW rinse in between (x+H2O+x).


15

Langmuir 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 16 of 43

Dosing of N3-silane and moisture. To further improve the density of the N3-SAM deposited via the vapor phase, increasing amounts of N3-silane and subsequent depositions have been tried at a deposition temperature of 145 °C. The measured contact angles, N3-intensity and film thickness are reported in Figure 2. When the volume of N3-silane is doubled from 100 to 200 µl or for multiple cycles no change in contact angle, film thickness and N3-surface concentration is observed. Like for solvent deposition, water has an important effect on the formation of SAMs and sequential introductions of controlled amounts of silane precursor and water vapor have been proposed in literature to increase the film density.37-39 Therefore, the effect of adding moisture was evaluated using a short DIW rinse in between two subsequent depositions. As shown in Figure 2, only the measured thickness increased, while the contact angle decreased. The concentration of N3-groups at the surface, however, remained constant indicating that the observed changes were not due to incorporation of more N3-silane molecules in the layers. The observed thickness change must be related to the incorporation of moisture inside the film. Lowe et al.22 added MgSO4·7H2O for the controlled release of water during the deposition of N3silanes at 110 °C and thereby obtained an increased monolayer coverage. However, the addition of water during the vapor-phase deposition implies an increased risk for polymerization of the silane molecules in the vapor and must be meticulously controlled. In our work, the exposure of the deposited monolayers to water did not influence the concentration of N3-groups in the layer. A possible explanation for this discrepancy could be that the substrates already contain enough surface adsorbed water to fully complete the surface reaction during the deposition.


16

Page 17 of 43 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

Langmuir

(b) Characterization of the N3-SAM. N3-SAM growth via siloxane formation at different temperatures. The new chemical bonds introduced by grafting of the N3-silane at a SiO2 surface were further investigated with ATR-FTIR. Figure 3,a shows the region above 1300cm-1 for N3-SAMs prepared via vapor-phase coating at two different temperatures. The deposition time was selected such that a saturated monolayer was formed (i.e. 1hour deposition at 145 °C and 24 hours at 60 °C).


17

Langmuir 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 18 of 43

Figure 3. ATR-FTIR spectra in the region 1300 – 3700 cm-1 (a) and 750 – 1300 cm-1 (b) of N3SAMs formed on Si/SiO2 via the vapor phase at 60 °C (24 hours deposition) and at 145 °C (1


18

Page 19 of 43 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

Langmuir

hour deposition). The spectrum of a blank Si/SiO2 sample after UV/O3 clean is included as a reference. The spectrum of a native Si/SiO2-substrate after UV/O3-clean is included as a reference. As already mentioned above, the introduction of the N3 functional group is clearly visible by its asymmetric stretching band at 2100 cm-1, while the symmetric stretching vibration of the N3 at 1348 cm-1 has lower intensity and may be overlapping with CH2 bending deformations. Clear bands of CH2 and CH3 stretching vibrations in the region 3000-2800 cm-1 can be observed on all samples including the ‘clean’ reference sample after UV/O3. Although measurements were done within 1 hour after preparation of the samples, the presence of these bands on the Si/SiO2 reference sample is related to inevitable organic contamination from air exposure and demonstrates the surface sensitivity of these ATR-FTIR measurements. Similarly, all spectra show a spurious band at 1740 cm-1 which we also attribute to the addition of organic contamination during the coating and/or transport of the samples to the ATR-FTIR tool. For the N3-SAMs, there is a significant increase in intensity for the doublet related to the CH2 symmetric and asymmetric stretching confirming the incorporation of the undecyl-chain in the monolayer. Especially the position of the asymmetric stretching band is very sensitive to the arrangement of the alkyl chains in the SAM. For crystalline n-alkane chains, the asymmetric stretching modes appear around 2916-2918 cm-1 while it shifts to higher wavenumbers for disordered chains.40-42 For the N3-SAM deposited via the vapor phase, the peak position for the asymmetric methylene stretching is ~2928 cm-1 indicating a non-crystalline layer. This finding agrees with the disordered structure identified by Steinruck et al.43 for SAMs of organosilanes with relatively short alkyl chains while closely packed crystalline domains were only reported for longer chain lengths, namely 14 or more CH2-units. Moreover, the presence of the terminal azido-group can


19

Langmuir 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 20 of 43

further contribute to the formation of a non-closely packed layer due to its larger size compared to CH2-units. At 1470 cm-1 a peak due to the CH2 scissor vibration is observed for all samples. The CH3 asymmetric stretching from the trimethoxy-groups should appear as a band at 2975 – 2950 cm-1 while the symmetric stretching has lower intensity and overlaps with the CH2 symmetric stretching. The intensity of the CH3 stretching bands after vapor-phase deposition is low and comparable to the bands on the Si/SiO2 reference sample. This indicates that they are related to residual organic contamination and suggests the complete hydrolysis and/or reaction of the methoxy-groups both for the deposition at 60 and 145 °C. An additional confirmation for the condensation reaction of the trimethoxy-groups at the surface is the absence of two distinct Si-OCH3 bands at 1190 and 1100cm-1 due to CH3 rocking and asymmetric Si-O-C stretching respectively which should occur with strong intensity.44 Lowe et al.22 have measured unhydrolyzed methoxy groups remaining after vapor deposition of silanes at 110°C and therefore added a source of water to form a dense, completely hydrolyzed monolayer. The full hydrolysis of methoxy-groups observed in our work is further supported by the lack of impact of water exposure on the deposited layers as described above. The spectral area below 1300 cm-1 (see Figure 3, b) is dominated by the characteristic peaks related to Si-O vibrations. The longitudinal optical (LO) phonon mode from the Si-O-Si asymmetric stretching is visible with ATR-FTIR and appears as a band located at ~1230 cm-1 with an unsymmetrical shape. 45 From the spectra, a shift in the peak position and an increase in the intensity of the LO-band is observed after N3-SAM formation. The peak positions and relative intensities as given by the peak height are summarized in Table 1.


20

Page 21 of 43 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

Langmuir


21

Langmuir 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 22 of 43

Table 1. Peak position and peak height of the LO mode of the Si-O stretching after N3-SAM formation via the vapor phase at 60 °C (24 hours deposition) and 145 °C (1 hour deposition). Si/SiO2 samples before coating, after removal of the N3-SAM and after heating are included as references. Calculated# thickness increase (Å)

Peak position Si-O (cm-1)

Peak height Si-O (a.u.)

Si/SiO2

1230.4 ± 0.3

1.37 ± 0.01

N3-SAM 60 °C

1230.6 ± 0.7

1.45 ± 0.05

0.5 ± 0.3

N3-SAM 145 °C

1232.5 ± 0.3

1.48 ± 0.03

0.6 ± 0.2

N3-SAM 145 °C + UV/O3

1235.7 ± 0.1

1.59 ± 0.06

1.3 ± 0.3

Si/SiO2 + UV/O3

1230.6 ± 0.7

1.41 ± 0.02

0.2 ± 0.1

Si/SiO2 + 1 hour at 145 °C

1231.0 ± 0.1

1.39 ± 0.01

0.2 ± 0.06

#

: Calculated from the increase in Si-O peak height using a slope of 1.74 a.u./nm obtained from the LO-peak height as function of SiO2 thickness (see supplementary info, Figure S.5).

Tian and co-workers demonstrated that the spectral position and the intensity of the LO-band can be used to characterize the newly formed SAM/substrate interface.46 The observed blue shift and increase of the LO phonon mode upon SAM formation was related to the creation of interfacial Si-O bonds between the SAM and the substrate. In our work, a similar increase and a shift of the LO-band is observed via ATR-FTIR after N3-SAM formation. As the observed intensity increase is proportional to the oxide thickness with a slope of 1.74 ± a.u./nm (see supplementary info, Figure S.5), an apparent added thickness of 0.6 ± 0. 2 Å to the existing SiO2 layer can be estimated due to N3-SAM formation via the vapor phase at 145 °C. For N3-SAM deposited at 60 °C, a similar increase in added thickness is found. The estimated thicknesses are comparable to


22

Page 23 of 43 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

Langmuir

the previous work of Tian et al.46 who showed an increase in thickness of 0.8 Å after solventbased deposition of a undecanaltriethoxy silane. To exclude that the observed effects in our work are related to changes in reflection due to differences in refractive index or thickness of the layer, the organic chains of the formed N3SAM were removed with a 5 min. UV/O3 clean. The complete removal of the N3-group and CH2-units is confirmed by the disappearance of the N3 band at 2100 cm-1 and the decrease in CH2 bands (see supplementary info, Figure S.3). For the N3-SAM after this treatment, an additional intensity increase of the LO-band is observed in contrary to a Si/SiO2 blanket. This confirms that only the N3-SAM is oxidized with formation of extra Si-O bonds contributing to the existing SiO2 network and limited oxidation of the substrate occurs. Additionally, Si/SiO2 substrates that were put in the vacuum chamber at 145 °C without the silane chemistry did not show an increased LO-band. The additional increase in LO intensity after oxidation of the alkyl sidechains by the UV/O3 corresponds to the formation of 1.3 Å thick layer. Only the silane molecules that create an interfacial bond contribute to the added thickness after N3-SAM deposition while the thickness after oxidation of the N3-SAM contains additional contributions of all N3-silanes that have reacted at the surface (see Scheme III). As there is a two-fold increase in added thickness after oxidation of the N3-SAM, it can be concluded that the total number of bonds between the added N3-silanes and the substrate equals the number of silane molecules. Besides that, the silanes form a laterally polymerized network47-49 but these lateral bonds are not visible with p-polarized ATRFTIR spectroscopy.


23

Langmuir 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 24 of 43

Scheme III. Schematic representation of the 11-azidoundecyltrimethoxysilane grafting at the Si/SiO2 substrate and after subsequent removal using a short UV/O3-treatment.

Additional prove of the condensation reaction during vapor-deposition of the N3-SAMs can be found in the disappearance of the shoulder around 970cm-1 that can be attributed to the Si-OH stretching42 and that is only clearly visible on all the Si/SiO2 samples without N3-SAM. N3-SAM orientation. The presence and position of the N3-functional groups in the layer was confirmed by XPS measurements. The atomic surface concentrations after reaction with the N3-silane are


24

Page 25 of 43 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

Langmuir

summarized in Table 2 while Figure 4 shows the N1s spectrum.

Table 2. Angle integrated atomic surface concentrations (%) extracted by XPS measurements on N3-SAMs formed via the solvent phase or via the vapor phase during 1 hour at 145 °C.

C1s

N1s

O1s

Si2p (SiO2)

Si2p (Si)

UV/O3

6.7

0.1

38.5

14.4

40.4

N3-SAM solvent

30.5

3.2

26.3

12.4

27.6

10

N3-SAM 145 °C

23.3

3.8

34.1

14.8

23.9

6

11

3

ideal

C/N

3.66


25

Langmuir 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 26 of 43

Figure 4. N 1s XPS spectra acquired on N3-SAM deposited after 1 hour via the vapor phase at 145 °C. The higher binding energy N1s band at 404.1 eV is assigned to the electron-deficient, central nitrogen atom while the band at 401.5 eV is assigned to the two outer nitrogen atoms in the N3group.50 The intensities of both bands reflect the expected 1:2 ratio. Although the measurements were carried out under conditions where the degradation of the N3-group under X-ray exposure is minimized, a third band caused by photo-oxidized species is present with low intensity at lower binding energy. The C/N-ratio calculated from the reported atomic surface concentrations of the N3-SAM after vapor deposition in Table 2 is higher than the ideal value for a perfect monolayer. This might be partially related to the observed degradation. In agreement with the ATR-FTIR results, additional background contamination is also evident from the C-content measured on reference samples that received only a UV/O3 clean immediately before loading in the XPS tool. Table 2 shows also higher C-concentrations for N3-SAM deposited via the liquid phase and a corresponding higher C/N-ratio confirming a less densely formed SAM again in agreement with the relative densities derived from the ATR-FTIR data. Information on the ordering of the different species in the layer can be obtained by comparing measurements done at different angles due to the decreased effective analysis depth at large angles. The atomic surface concentrations as measured at different angles from the normal of the sample for the N3-SAM after vapor-phase deposition is given in Figure 5 and in the supplementary info, Table S.1. As measurements performed at an exit angle of 78 deg are more surface sensitive compared to at 20 deg, these data confirm that the N is mainly occurring near the surface suggesting an upward orientation of the N3-group in the obtained layer.


26

Page 27 of 43

50 40

O1s

Atomic conc. (%)

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

Langmuir

Si2p (Si)

30 20

C1s

Si2p (SiO2)

10 N1s 0 30

40

50

60

70

Angle (deg)

Figure 5. Atomic surface concentrations as measured at different angles from the normal of the sample for the N3-SAM after 1 hour vapor-phase deposition at 145 °C. Determination of N3-group density. The number of N3-groups that is a measure of the surface density of the SAM was further quantified using TXRF measurements. The N3-concentration was determined indirectly after click reaction with chloro-hexyne and measuring the resulting Cl concentration. As the azidealkyne click reaction occurs with high yield, the concentration of Cl at the surface is a measure for the N3-groups present at the outer SAM surface. Indeed, Figure 6 shows the almost complete disappearance (> 90%) of the N3-stretching vibration in the ATR-FTIR spectrum of the vapordeposited N3-SAM after click reaction with chloro-hexyne. The remaining intensity around 2100 cm-1 is at least partially due to a H2O combination band. In absence of the Cu-catalyst, no reaction of the N3-group occurs indicating that the N3-group is stable and does not decompose


27

Langmuir 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 28 of 43

under the conditions used for the click reaction. Moreover, as the chloro-hexyne compound has a purity >98%, it can be concluded that mainly Cl-containing reaction products are formed.

Figure 6. ATR-FTIR spectra in the region 1300 – 3700 cm-1 of N3-SAMs formed on Si/SiO2 via 1 hour vapor-phase deposition at 145 °C before and after Cu-catalyzed click reaction with chlorohexyne. The spectrum of a blank Si/SiO2 sample after UV/O3 clean and of the N3-SAM after reaction without the Cu-catalyst are included as references.


28

Page 29 of 43 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

Langmuir

The reference samples included in the measurements to check the background Cl-concentration, display sufficiently low levels of Cl compared to the signal after click reaction with the chlorohexyne (see Table 3). Moreover, the diluted HCl rinse can remove all Cu-precipitates from the surface. As X-ray irradiation might damage the organic layer, the effect of TXRF measurement time upon the measured concentrations was verified and found to have no impact (see supplementary info, Table S.2). The reported Cl-concentration after surface reaction with the Clhexyne and the small variation (

Chemical Vapor Deposition of Azidoalkylsilane ... - ACS Publications

Recommend Documents