Role of different receptor-surface binding modes on the morphological

Feb 8, 2019 - Johannes Daniel Bartl , Paolo Scarbolo , Denis Brandalise , Martin Stutzmann , Marc Tornow , Luca Selmi , and Anna Cattani-Scholz. Langm...
0 downloads 0 Views 2MB Size
Subscriber access provided by TULANE UNIVERSITY

Interfaces: Adsorption, Reactions, Films, Forces, Measurement Techniques, Charge Transfer, Electrochemistry, Electrocatalysis, Energy Production and Storage

Role of different receptor-surface binding modes on the morphological and electrochemical properties of peptide nucleic acid-based sensing platforms Johannes Daniel Bartl, Paolo Scarbolo, Denis Brandalise, Martin Stutzmann, Marc Tornow, Luca Selmi, and Anna Cattani-Scholz Langmuir, Just Accepted Manuscript • DOI: 10.1021/acs.langmuir.8b03968 • Publication Date (Web): 08 Feb 2019 Downloaded from http://pubs.acs.org on February 8, 2019

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 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 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.

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 67 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

Role of different receptor-surface binding modes on the morphological and electrochemical properties of peptide nucleic acid-based sensing platforms Johannes D. Bartl †, Paolo Scarbolo ‡, Denis Brandalise ‡, Martin Stutzmann †, Marc Tornow §,#,¶, Luca Selmi ∥ and Anna Cattani-Scholz †,¶,*



Walter Schottky Institute (WSI) and Physics Department, Technische Universität München, Am Coulombwall 4, 85748 Garching bei München, Germany



Dipartimento Politecnico di Ingegneria e Architettura (DPIA), Università degli Studi di Udine, Via delle Scienze 206, 33100 Udine, (Italy)

§

Molecular Electronics, Department of Electrical and Computer Engineering,

Technische Universität München, Theresienstr. 90, 80333 München, Germany

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 67

Fraunhofer Research Institution for Microsystems and Solid State Technologies (EMFT), Hansastr. 27d, 80686 München, Germany



Center for Nanoscience (CeNS), Ludwig-Maximilians-Universität München, Geschwister-Scholl-Platz 1, 80539 München, Germany



Dipartimento di Ingegneria “Enzo Ferrari”, Università degli Studi di Modena e Reggio Emilia, Via Vivarelli 10, 41125, Modena, Italy

ACS Paragon Plus Environment

2

Page 3 of 67 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

ABSTRACT

Label-free detection of charged biomolecules, such as DNA, has experienced an increase in research activity in recent years, mainly to obviate the need of elaborate and expensive pre-treatments for labeling target biomolecules. A promising label-free approach is based on the detection of changes in the electrical surface potential on biofunctionalized silicon

ACS Paragon Plus Environment

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 67

field effect devices. These devices require a reliable and selective immobilization of charged biomolecules on the device surface.

In this work, self-assembled monolayers of phosphonic acids (SAMPs) are used to prepare organic interfaces with a high density of peptide nucleic acid (PNA) bioreceptors, which are a synthetic analog to DNA, covalently bound either in a multidentate (║PNA) or monodentate (┴PNA) fashion to the underlying silicon native oxide surface. The impact of the PNA bioreceptor orientation on the sensing platform’s surface properties is characterized in detail by water contact angle measurements, atomic force microscopy (AFM), X-ray photoelectron spectroscopy (XPS), cyclic voltammetry (CV) and electrochemical impedance spectroscopy (EIS). Our results suggest that multidentate binding of the bioreceptor via attachment groups at the γ-points along the PNA backbone leads to the formation of an extended, protruding and netlike 3D-metastructure. Typical “mesh” sizes are on the order of 8 ± 2.5 nm in diameter, with no preferential spatial orientation relative to the underlying surface. Contrary, the monodentate binding provides a spatially more oriented metastructure comprising cylindrical features, of a typical size

ACS Paragon Plus Environment

4

Page 5 of 67 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

of 62 ± 23 nm × 12 ± 2 nm. Additional cyclic voltammetry measurements in a redox buffer solution containing a small and highly mobile Ru-based complex, reveal strikingly different insulating properties (ion diffusion kinetics) of these two PNA systems. Investigation by electrochemical impedance spectroscopy confirms that the binding mode has a significant impact on the electrochemical properties of the functional PNA layers represented by detectable changes of the conductance and capacitance of the underlying silicon substrate in the range of 30-50 % depending on the surface organization of the bioreceptors in different bias potential regimes.

KEYWORDS Surface functionalization, organophosphonate interfaces, PNA bioreceptors, cyclic voltammetry, impedance spectroscopy

ACS Paragon Plus Environment

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 67

INTRODUCTION Molecular assays based on the hybridization of target DNA to complementary DNA oligomer probes, which are attached to a solid surface, were first reported nearly four decades ago (1) and have since been applied extensively in the field of molecular diagnostics using biosensors (2). In modern DNA biosensing detection schemes, DNA

ACS Paragon Plus Environment

6

Page 7 of 67 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

oligomer probes are often substituted by peptide nucleic acid (PNA), a promising synthetic analogue of DNA, in which the negatively charged sugar-phosphate backbone is replaced by neutral, achiral, and repeated N-(2-aminoethyl) glycine (AEG) units linked by peptide bonds (3). PNA single strands hybridize to complementary single-stranded oligonucleotide sequences in agreement with Watson-Crick base-pairing rules by establishing hydrogen bonds between complementary nucleobases (4,5). Moreover, the modification of the PNA backbone by various functional groups can be achieved relatively simply, enabling further tuning of the PNA properties (3,6,7). Consequently, a wide variety of novel detection protocols using PNA were developed (8). For example, a promising amplification strategy for the detection of single-stranded DNA was recently shown using PNA probes immobilized on Au-surfaces. The study was based on an in-situ growth mechanism of electroactive polymers through surface-initiated, electrochemically mediated atom transfer radical polymerization (SI-eATRP), illustrating the important role of PNA probes as central building blocks in novel device fabrication (9,10). The introduction of side chains at the γ-position of the PNA backbone does not impair the formation of DNA and RNA duplexes, but rather allows for an improvement of the binding

ACS Paragon Plus Environment

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 67

affinity (6,11). In fact, γ-modifications introduce an inherent helicity to the PNA resulting in a higher rigidity of the PNA molecule and, therefore, enhance DNA duplex formation (6). It is likely that these advantageous properties can be translated also to surfaceimmobilized PNA probes, which has, to our knowledge, not yet been reported in literature. In addition, surface-immobilized PNA probes are relevant for biosensing applications due to the stability of the PNA-DNA duplex, and their suitability for the fabrication into different configurations (12-14). Of high interest is the ability to design interfaces, in which the bioreceptor’s surface orientation can be controlled specifically. In field-effect transistor (FET) devices, for example, DNA or RNA recognition is based on the (indirect) detection of changes in the electrical surface potential and, thus, requires reliable and selective adsorption of charged biomolecules close to the sensing surface (12-14). In conventional FET sensor devices, DNA and PNA probes have been attached by various chemical cross-linking strategies oriented vertically away from the gate-oxide surface. For these devices, the ionic strength of the hybridization buffer is an important factor (15,16). To minimize the electrolyte screening effects, to increase the sensitivity, and to minimize the variability, a PNA probe orientation parallel to the sensor surface was recently suggested

ACS Paragon Plus Environment

8

Page 9 of 67 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

(17). The corresponding hybridization studies performed with radiolabeled DNA ((γ-32P)ATP at the 5′ end) already showed that the estimated equilibrium dissociation constants of the horizontally and vertically tethered PNA-DNA duplex are of the same order on magnitude (KD ≈ 5 nM), indicating a sufficient hybridization affinity for many electronic biosensing applications. Recently, we have demonstrated that PNA oligomers can be grafted on silicon/silicon oxide surfaces in a similar fashion also through organophosphonate chemistry with satisfactory yields (18). Self-assembled monolayer of phosphonic acids (SAMPs) are stable under acidic, neutral, and physiological conditions and exhibit an increased longterm stability in aqueous environments compared to silane-based interfaces (19). Short (15 bases), single-stranded PNA molecules,

ACS Paragon Plus Environment

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 67

Scheme 1. Single-stranded peptide nucleic acid (PNA, red) is coupled to the selfassembled monolayer of phosphonic acids (SAMPs, blue) either in a vertical (┴PNA) (pathway A) or horizontal (║PNA) (pathway B) fashion using standard maleimido-based cross-coupling (green) methods (not drawn to scale).

modified with reactive SH sites at different points of the backbone, were coupled to SAMPs-terminated silicon native oxide either in a vertical (┴PNA) or horizontal (║PNA) surface conformation (Scheme 1). The applied PNA immobilization chemistry depends on standard maleimido cross-coupling (20,21) and exploits both the hydrolytic stability of organophosphonate anchor groups (22) and the protective properties of ethylene glycol units (EG) against non-specific surface binding of charged biomolecules (23-25) making

ACS Paragon Plus Environment

10

Page 11 of 67 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

them attractive bioreceptor platforms for DNA/RNA detection in a physiological environment. The applied PNA surface immobilization scheme provides high control and reproducibility over the organic interface. Since the organic interface defines primarily the overall device sensing characteristics, we investigated the impact of two distinct PNA interfacial surface conformations on the morphological and electrochemical properties of PNA-based devices, in terms of yield, stability, and receptor density. In this work, an in-depth characterization of these PNA systems was performed by using several surface analysis techniques, including water contact angle measurements, atomic force microscopy (AFM), and X-ray photoelectron spectroscopy (XPS). By cyclic voltammetry (CV) and electrochemical impedance spectroscopy (EIS), we evaluated the influence of the different PNA surface orientations with respect to the field-effect sensor response, i.e. differential capacitance and conductance changes. We can show that different PNA binding modes result in different surface orientations of these bioreceptors, which has a significant impact on the respective surface properties, in particular affecting the electrical

ACS Paragon Plus Environment

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 67

impedance, which may open a potential pathway to further tune the sensing capabilities of PNA-based biosensors for future applications.

EXPERIMENTAL SECTION

ACS Paragon Plus Environment

12

Page 13 of 67 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

General. 2-{2-[2-Hydroxy-ethoxy]-ethoxy}-ethyl phosphonic acid (referred to as SAMPs, ≥95 % (GC)) was purchased from SiKÉMIA. 3-(Maleimido) propionic acid Nhydroxysuccinimide acid-NHS ester (referred to as Linker, ≥98.5 % (HPLC)) was ordered from Sigma-Aldrich and tris(hydroxymethyl)aminomethane (TRIS, ≥99.9 %) was bought from Carl Roth GmbH & Co. KG. Single-stranded 15-mer peptide nucleic acid (PNA, ≥99.9 % (HPLC)) molecules with a standard base sequence of 5’-TGT-ACA-TCA-CAACTA-3’ were obtained from Panagene Inc. The backbone of the molecules was slightly modified to generate two different PNA types. The full base sequence of type 1 (║PNA) is 5’-Ac-Lys-TGT*-ACA-TC*A-CAA-C*TA-Lys-3’, where * refers to a Lys-γ modification with a C6-SH group (C6-SH linker) and the full type 2 (┴PNA) sequence is 5’-Cys-TGT-ACATCA-CAA-CTA-Lys-3’. A schematic representation of both PNA structures can be found in the SI (see Scheme S1). Methods describing various strategies to fabricate and modify PNA can be found elsewhere (26). Ultrapure water (DI H2O, 18.2 MΩ·cm at 25 °C, Merck Millipore) was used to prepare standard buffer solutions of 50 mM TRIS (pH = 7.55 at 26.3 °C) with 100 mM NaCl and redox buffer solutions containing 50 mM TRIS (pH = 7.56 at 26.1 °C) with 100 mM NaCl, 1 mM Hexaammineruthenium(II) chloride (Sigma-Aldrich,

ACS Paragon Plus Environment

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 67

99.9 %) and 1 mM Hexaammineruthenium(III) chloride (Sigma-Aldrich, 98 %) for electrochemical measurements. All buffer solutions were deoxygenated by N2-saturation for at least 45 min after transfer to the electrolyte chamber and between individual measurements. Common reagents were purchased from Sigma-Aldrich and high-grade organic solvents were ordered from Merck (VLSI Selectipur, ≥99.9 %) and used without further purification, respectively. Highly p-doped (boron, 0.0175 Ω·cm) silicon wafers (100), with a thin native oxide layer (~1 nm) (Siltronic AG) were cut into 1 × 0.8 cm2 pieces. Chips used as working electrodes during electrochemical measurements were further coated by physical vapor deposition (PVD) with a Cr (15 nm)/Au (150 nm) layer to obtain an ohmic contact on the backside (see SI). OH-Surface Termination. Cr/Au back contacted silicon chips (see SI) were cleaned subsequently with deionized water (DI H2O), 2-propanol, acetone and 2-propanol in an ultrasonication bath (37 kHz) for 10 min, respectively, and dried under a N2 stream. The cleaned chips were treated by an oxygen plasma (TePla GmbH, GIGAEtch 100-E) with 200 W power at a pressure of ≤ 0.1 mbar for 300 s, transferred in a pre-heated (70 °C) Basic Piranha solution (5 : 1 : 1 ratio of DI H2O : H2O2 (BASF, VLSIn Selectipur, 31 %) :

ACS Paragon Plus Environment

14

Page 15 of 67 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

NH4OH (Sigma-Aldrich, 28-30 % NH3)) for 10 min,

cleaned with DI H2O in an

ultrasonication bath (37 kHz) for 10 min, dried in an N2-atmosphere at 120 °C for 5 min as reported previously (18) and immediately used for surface functionalization. Self-Assembled Monolayers of Phosphonates (SAMPs). SAMPs of 2-{2-[2-Hydroxyethoxy]-ethoxy}-ethyl phosphonic acid molecules (SAMPs for further reference) were prepared using the T-BAG method (tethering by aggregation and growth) with two full TBAG cycles as reported previously (18,21,27). The chips were immersed in an 80 ml tetrahydrofuran (THF) solution (Sigma-Aldrich, anhydrous, inhibitor-free, ≥99.9 %) containing 25 μmol·L-1 of the phosphonic acid molecules. After each T-BAG cycle, the samples were stored in an N2-atmosphere at 120 °C for at least 18 h. After the first TBAG cycle, the chips were cleaned twice in ethanol (EtOH; Sigma-Aldrich, absolute, reag. ISO, reag. Ph. Eur., ≥99.8 %), then in a mixture of DI H2O and THF (3 : 1), followed by cleaning in DI H2O in an ultrasonic bath for 10 min, respectively, and dried under a N2 stream. After the second T-BAG cycle, the chips were cleaned twice in EtOH, then in a mixture of DI H2O, THF and Triethylamine (TEA; Sigma-Aldrich, ≥99.5 %) (10 : 3 : 1), followed by cleaning in DI H2O in an ultrasonic bath for 10 min, respectively, and dried

ACS Paragon Plus Environment

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 67

under a N2 stream to remove potentially formed phosphonic acid multilayers from the surface. PNA-Immobilization. SAMPs-functionalized samples were subsequently cleaned with DI H2O and EtOH in an ultrasonication bath (37 kHz,) for 10 min, respectively, and dried with N2. Two of the cleaned chips were transferred back to back in a gas-tight reaction tube containing a 19 mM solution of 3-maleimidopropionic acid-NHS (Linker for further reference) in dry acetonitrile (20.094 mmol dissolved in 5 ml dry acetonitrile) and stored vibration-free at room temperature under an N2-atmosphere for at least 20 h. After functionalization, the chips were cleaned two times with dry acetonitrile in an ultrasonication bath (37 kHz) for 5 min, respectively, and dried under a N2 stream to remove excess Linker molecules. Coupling of the PNA molecules was achieved by standard Michael addition under mild reaction conditions (23): a 25.1 μmol·L−1 PNA stock solution was prepared by dissolving 25.1 nmol of the respective PNA in 1000 μL of DI H2O at 50 °C in an ultrasonication bath (37 kHz) for at least 10 min. Two of the treated chips were transferred in a 5 ml gas-tight reaction tube filled with a mixture of 200 μL PNA stock solution and 800 μL of DI H2O (1 : 4; 6.3 μmol·L−1), sealed gas-tight under an N2-

ACS Paragon Plus Environment

16

Page 17 of 67 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

atmosphere and stored vibration-free for at least 24 h. Low PNA concentration and an elongated reaction time ensure the formation of uniform PNA-terminated surfaces with low densities of physisorbed PNA aggregates on top of the chemisorbed PNA. After the treatment, the samples were cleaned with DI H2O and EtOH in an ultrasonication bath (37 kHz,) for 10 min, respectively, and dried with N2 to remove non-covalently bound PNA molecules from the surface. A reaction mechanism of the Linker- and PNA-attachment can be found in the SI (Scheme S2). Contact angle measurements. Water contact angle measurements were performed on an OCA 15Pro contact angle system (DataPhysics Instruments) under ambient conditions. Data acquisition and evaluation was realized with SCA 20 - contact angle (DataPhysics Instruments, ver. 2.0). To estimate an average Young’s contact angle (𝜃𝑦), 3 μL of deionized H2O was dispensed with a rate of 1 μL·s-1 from a 500 μL Hamilton syringe on the sample surface and after ~3 s (reaching the equilibrium shape) an image was taken for further processing. The procedure was repeated at least five times on different representative spots on the surface and the standard deviation (error) was calculated.

ACS Paragon Plus Environment

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 67

Atomic force microscopy. Surface topography and roughness were investigated by atomic force microscopy (AFM, MultiMode 8, Bruker Corp.) in tapping mode and in contact mode under ambient conditions using NSG30 (TipsNano) for standard characterization and AC 240TS-R3 (Oxford Instruments) silicon cantilever tips for high-resolution scans. Data acquisition and processing was achieved with NanoScope Analysis (Bruker Corp., ver. 9.0). Tapping mode micrographs with a size of 5 × 5 μm2 (overview) were taken to probe for holes in the monolayer and potential multilayer formation, whereas 1 × 1 μm2 tapping mode micrographs were used to obtain a detailed surface morphology. To estimate the organic layer thickness, scratching in contact mode was performed over areas of 1 × 1 μm2 with a deflection setpoint of 0.18 V. At least three different regions adequately distributed over the sample surface were mapped to provide a representative description of the surface morphology. The surface roughness was evaluated via the root mean square (RMS) average of height deviations taken from the mean image data plane of 1 × 1 μm2 tapping mode micrographs. To receive a more representative description of the surface morphology, the measured RMS-values were averaged considering 1 × 1 μm2 tapping mode images of at least three different samples of each functionalization step. In

ACS Paragon Plus Environment

18

Page 19 of 67 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

addition, step profiles were obtained from a 980 x 250 nm2 area cutouts of 1 × 1 μm2 tapping mode micrographs. The 1 × 1 μm2 tapping mode images were further processed using ImageJ 1.52a (NIH USA, ver. 1.8.0) to determine typical “mesh” sizes and structures of the different PNA overlayers. For the ║PNA type at least 175 and for the ┴PNA variant at least 100 individual overlayer units per 1 × 1 μm2 tapping mode image were considered. X-ray

photoelectron

spectroscopy.

X-ray

photoelectron

spectroscopy

(XPS)

measurements were performed with an in-house built UHV setup equipped with a hemispherical energy analyzer PHOIBOS 225 (Specs GmbH; Berlin, Germany). The dual (Mg and Al) X-ray tube anode was operated at an emission current of 20 mA with a voltage of 12.5 kV at a base pressure of ≤8×10-9 mbar with non-monochromatic Al 𝐾𝛼 radiation (1486.6 eV). The data were recorded under normal emission and in medium area mode, corresponding to a circular acceptance area of 2 mm in diameter. Spectra were recorded with SpecsLab (SPECS Surface Nano Analysis GmbH, ver. 2.85), processed with CasaXPS (Casa Software Ltd, ver. 2.3.17) and further charge corrected by shifting the maximum of the Si 2p peak to the literature value of 99.4 eV (28). The

ACS Paragon Plus Environment

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 67

basic method used to calculate the surface composition from XPS high resolution spectra is presented in the SI. Electrochemical Measurements. For electrochemical characterization, a standard three electrode setup was used: surface-treated silicon chip (working electrode), an encapsulated Ag/AgCl reference electrode (BASi, filled with 3.0 M NaCl) and a spiral platinum wire (counter electrode) (Scheme 2). Each part of the electrolyte chamber was made from chemically inert Teflon. Potentials were applied and monitored by a PGSTAT 204 potentiostat electrochemical system (Metrohm Autolab). Data acquisition and evaluation was achieved by Metrohm Autolab Nova (Metrohm Autolab, ver 1.11). During measurements, the chamber was stored in a Faraday cage in complete darkness. The active electrochemical interface (contact area between the silicon chips and electrolyte) was estimated to be 0.189 cm2 ± 0.05 cm2 and measured currents were normalized accordingly. I-V curves of cyclic voltammetry measurements were recorded at different scan rates (10 to 500 mVs−1) in a potential window of −1.5 to 0.5 V (the limits of the potential window were

ACS Paragon Plus Environment

20

Page 21 of 67 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

Scheme 2. Schematic representation of the measurement cell used for electrochemical characterization of differently functionalized silicon chips (shown for a ║PNA system). The measurement cell (Teflon) is comprised of a standard three electrode setup: ohmic contacted silicon (working electrode), an encapsulated Ag/AgCl reference electrode and a spiral platinum wire (counter electrode). P-type silicon chips were mounted between the socket and the cylindrical measurement chamber.

chosen to guarantee the repeatability of the measured cyclic voltammograms of the SAMPs layer). The complex impedance (absolute value |Z| and phase) was measured as a function of frequency between 0.1 Hz and 100 kHz with an AC signal amplitude set to

ACS Paragon Plus Environment

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 67

10 mV (single sine), for a set of different bias potential values (-0.7 to 0.5 V in steps of 0.1 V). Prior to individual frequency sweeps, the system was equilibrated by applying the corresponding potential for 2 min. To evaluate the impedance data, a modular lumped elements equivalent circuit was developed, in which each layer of the sensor is represented by separate parallel RC-elements connected in series. Comparison of different sets of measurements corresponding to the pristine and differently functionalized surfaces enables the extraction of layer specific electrical impedance parameters. Functionalization-dependent conductances and capacitances of the interface region are calculated before/after SAMP-and PNA-deposition.

ACS Paragon Plus Environment

22

Page 23 of 67 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 Contact angle measurements. Wetting properties of silicon native oxide, SAMPs and PNA coated samples were investigated by water contact angle measurements (Table S1). After SAMPs surface functionalization, the water contact angle increased from θ  26 ±

ACS Paragon Plus Environment

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 67

1 to 45 ± 3°, indicating a decrease of the surface hydrophilicity, as previously reported (18,21). For the ║PNA type a further increase of the water contact angle to θ  52 ± 2° is observed, as expected. Contrary, the ┴PNA type interface shows a more hydrophilic behavior with contact angles of θ  35 ± 3°. We assign our finding to different local regions of surface-exposed hydrophobic and hydrophilic functional groups for the two PNA interface systems. PNA oligomers possess regions with a high number of heteroatoms in the bases, prone to increase the wettability of the interface if sufficiently exposed on the surface (3). The observed disparate wetting behavior for the two PNA interface systems is an initial indication that the PNA binding mode can influence the overall wetting properties of the fabricated interface systems. Atomic force microscopy. Atomic force microscopy (AFM) was used to characterize the surface morphology (Figure 1). A comparable surface roughness (RMS) is observed before (0.17 ± 0.04 nm) and after monolayer deposition (0.15 ± 0.02 nm), with no evidence of pinholes or multilayer formation. These observations are confirmed over different fabrication batches, indicating the high reliability of the SAMPs deposition method applied in this work. AFM scratching experiments indicated the thickness of the

ACS Paragon Plus Environment

24

Page 25 of 67 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

SAMPs layers (𝑑𝑆𝐴𝑀𝑃𝑠) to be on the order of 0.3-0.5 ± 0.1 nm (Figure 1b), which is in reasonable agreement with previous studies (21). However, the experimentally determined SAMPs thickness is slightly lower than expected for a well-packed monolayer with a backbone chain length of ca. 0.9 nm. We assign our findings to minor distortions of the hydrogen bonding network at the top of the hydroxylated film and to the presence of oxygen

Figure 1. AFM tapping mode images of a silicon native oxide sample after Piranha treatment (a) and of SAMPs of 2-{2-[2-Hydroxy-ethoxy]-ethoxy}-ethyl phosphonic acid after scratching in AFM contact mode (b).

ACS Paragon Plus Environment

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 67

High resolution micrographs of functional layers obtained after immobilization of the PNA oligomers in both binding modes: ║PNA (c) and ┴PNA (d). In the center of Figure 1, a sketch of the overall layer arrangement (presented for the ║PNA case) and estimated thickness of the sensing platform is shown (not drawn to scale). Both PNA overlayers possess roughly the same overall thickness (║PNA ≈ ┴PNA).

(ethylene glycol functionalities) in the backbone chain, which may easily contribute to an increase of disorder in the film and to a reduced packing homogeneity. After immobilization of the backbone-modified PNA oligonucleotides on the maleimidoactivated surface the morphology of the organic interface changed significantly (Figure 1c and d). Both PNA overlayers possess roughly the same layer thickness on the order of 3 nm (see Figure S2). In Figure 2a step profiles

ACS Paragon Plus Environment

26

Page 27 of 67 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. Average profile heights (a) (considered image area: 980 x 250 nm2) taken from the same region on a representative silicon native oxide (black), a SAMPs (blue), a ║PNA (green) and a ┴PNA sample (red) together with the corresponding average surface roughness (RMS). To provide better visibility, the different profiles are arbitrarily shifted along the y-axis. High resolution micrographs (300 x 300 nm2 cutout) of a representative ║PNA (b) and a ┴PNA sample (c) together with a schematic sketch of the corresponding PNA metastructures. On both images, the building blocks comprising the two PNA metastructures can be observed (highlighted by red lines).

ACS Paragon Plus Environment

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 67

of a representative silicon native oxide (black), a SAMPs (blue), a ║PNA (green) and a ┴PNA sample (red) are shown together with the corresponding average surface roughness (RMS). For the ║PNA and ┴PNA interfaces, the average surface roughness is increased similarly to 0.35 ± 0.03 nm and 0.35 ± 0.01 nm, compared to 0.15 ± 0.02 nm for the SAMPs and 0.17 ± 0.04 nm for the Si/SiO2-system, respectively. The relatively large physisorbed agglomerates (less than 6 % surface coverage) were not considered for the calculation of the average surface roughness. PNA immobilization induces a significant change in the average profile heights with respect to the bare substrate (black) and the SAMPs system (blue). In both cases, separated and plateau-like segments are formed, in which PNA molecules are grouped in larger structures of different shape (overlayer building blocks). As expected, the two PNA variants do not form a closed overlayer, due to only moderate grafting-yields on the maleimido-activated surface, as previously shown for similar interfaces (18,21). However, both PNA profiles exhibit intriguing differences. The individual overlayer building blocks of the ┴PNA type (red) appear to be more separated and defined, whereas ║PNA overlayer building blocks (green) are merging into each other and form a more coherent overlayer. We attribute

ACS Paragon Plus Environment

28

Page 29 of 67 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

this difference in surface morphology to the specific binding modes of the individual PNA systems. ║PNA layers form an extended, protruding and netlike 3D-metastructure with no preferential spatial direction on the underlying surface. Typical “mesh” sizes are on the order of 8 ± 2.5 nm in diameter. Contrary, the ┴PNA type forms a more localized, yet extended metastructure comprised from individual cylindrical features, which appears to have an average preferential orientation (parallel orientation to each other) on the surface. These cylindrical features possess a typical size of 62 ± 23 nm × 12 ± 2 nm. These differences in surface morphology for the two PNA systems were consistently found on at least three different samples originating from three different batches, respectively. Moreover, prolonged sonication in standard solvents did not induce any significant change in the morphology observed by AFM. X-ray

photoelectron

spectroscopy.

X-ray

photoelectron

spectroscopy

(XPS)

characterization was carried out to monitor and quantify changes in the surface chemical composition before and after individual functionalization steps. The obtained results are summarized in Figure 3 and

ACS Paragon Plus Environment

29

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 30 of 67

Figure 3. High-resolution scans of the Si 2p core level region of a representative silicon native oxide sample after Piranha treatment (black), a SAMPs (blue) of 2-{2-[2-Hydroxy-ethoxy]-ethoxy}-ethyl phosphonic acid, a ║PNA (red) and a ┴PNA sample (red, dashed lines) (a) and corresponding high-resolution scans of the C 1s region (b) and N 1s core level region (c). Both the bare substrate and SAMPs layer do not contain nitrogen in relevant quantities. A quantitative comparison between the average surface composition for the differently functionalized systems is given in Table 1.

Table 1. The decrease of the Si 2p signal after SAMPs formation (Figure 3a, blue) with respect to the bare substrate (black) is in good agreement with the increasing C 1s signal related to the deposition of phosphonic acid molecules (Figure 3b, blue). Further attenuation of the Si 2p signal can be observed after PNA immobilization (Figure 3a, red). The overall damping of the Si 2p signal for both PNA systems is nearly equal, which indicates a comparable average thickness of the two PNA metastructures. After monolayer formation a shoulder of the C 1s peak positioned at higher binding energies

ACS Paragon Plus Environment

30

Page 31 of 67 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

(~287.3 eV) is observed, which can be assigned to the C-C-OH and C-C-O structural components (29) of the ethylenglycol backbone of the monolayer (see Scheme 1). Adventitious carbon and additional carbon contamination entrapped in the native oxide layer are detected at ~285.4 eV. The amount of surface carbon is further increased after PNA immobilization as expected. (Figure 3b, red). The relative broad and asymmetric C 1s signal is typical for PNA interfaces and related to the presence of carbon atoms in a wide range of chemical states. The

ACS Paragon Plus Environment

31

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 32 of 67

Figure 4. High-resolution scans of the C 1s core region of a representative ║PNA (a) and a ┴PNA sample (b) sample and the corresponding high-resolution scans of the N 1s core region (║PNA (c) and ┴PNA (d)) together with the components providing the best fit of the measurement data. A quantitative comparison between the structural features of both PNA types and the respective fitting components for the N 1s and the C 1s region are given in Table S2 and S3, respectively. For all considered fitting solution the residual standard deviation was close to unity (>0.97).

extended shoulder at higher binding energies (~289.2 eV) is attributed to carbon in a highly oxidized environment, whereas the less pronounced shoulder at lower binding energies (~285.2 eV) is mostly linked to aromatic carbon. The slightly different C 1s peak shape of both PNA samples (in particular visible in the lower binding energy regime) arises from the different backbone modification of the two PNA oligomers (see Scheme S1). The corresponding C 1s deconvolution for the two PNA types is presented in Figure 4a and b. A detailed assignment of the different carbon components is provided in the SI (see Table S2). Table 1. Average surface composition [at. %] of the differently functionalized systems obtained by XPS high-resolution scans (see Figure 3). Reliable determination of the phosphorus amount is not possible due

ACS Paragon Plus Environment

32

Page 33 of 67 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

to superposition with a silicon plasmon loss feature (30). Nitrogen traces on Si/SiO2 surfaces are related to the Basic Piranha surface activation (see Exp. Sec.).

Si/SiO2

SAMPs

Si [at. %]

60.6 ± 1.9

50.9 ± 2.7

32.6 ± 3.0

32.6 ± 1.6

O [at. %]

33.3 ± 1.2

37.2 ± 1.2

28.7 ± 2.1

28.0 ± 1.9

C [at. %]

5.0 ± 0.9

10.8 ± 1.9

27.9 ± 3.3

27.7 ± 1.7

N [at. %]

1.1 ± 0.3

1.1 ± 0.3

10.8 ± 1.8

11.7 ± 1.8

║PNA

┴PNA

Interestingly, the experimentally obtained carbon ratios agree well with the theoretical carbon structure of both PNA types, except for slight deviations related to contributions from the underlying organic layers. Moreover, we observe that some of these discrepancies can only arise from different PNA surface orientation related to the individual binding modes of PNA functional layers which might result in two distinct attenuation patterns. In Figure 3c, high-resolution scans of the N 1s core region for the differently functionalized samples are presented. Nitrogen cannot be detected in relevant quantities on both the silicon substrate (black) and the organophosphonate monolayer

ACS Paragon Plus Environment

33

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 34 of 67

(blue). After PNA immobilization a significant increase of the surface nitrogen is detected. Due to the PNA’s molecular structure the amount of surface nitrogen allows for an acceptable correlation with the grafting yield of the PNA probes (Scheme S1). The corresponding N 1s deconvolution analysis for the two PNA types is depicted in Figure 4c and d. A detailed assignment of the three nitrogen components is provided in the SI (see Table S3). For both PNA systems, the experimentally derived values are in good agreement with the theoretical composition. Deviation from the expected values might be related to contributions from the Linker, containing one nitrogen atom in a ― N < environment per Linker molecule (21). In Table 1 the average surface composition given in atomic percentage [at. %] of the differently functionalized systems is presented. The statistical errors are obtained by comparing the respective surface ratios of at least five silicon native oxide samples after Basic Piranha treatment, five SAMPs-terminated, and three different PNA samples of each type (each sample from a different fabrication batch). Even though, all interfaces were produced ex-situ under ambient conditions, the amount of surface nitrogen and,

ACS Paragon Plus Environment

34

Page 35 of 67 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

therefore, the PNA-grafting yields are rather comparable effectively illustrating the robustness and reproducibly of the proposed PNA surface functionalization scheme. Cyclic voltammetry. Cyclic voltammetry (CV) measurements were performed on representative silicon native oxide, SAMPs and PNA samples to investigate the insulating properties (ion diffusion kinetics) of the different organic layers. To this end, a small and highly mobile Ru-based redox couple was added to the standard buffer solution (50 mM TRIS containing 100 mM NaCl and 1 mM Hexaammineruthenium(II/III); pH = 7.56 at 26.1 °C). The redox complex can perform a simple one electron transfer: [Ru(NH3)6]3 + + e ― ⇌ [Ru(NH3)6]2 + . This complex can penetrate the different organic films and participate in redox reactions on the electrode surface. However, the penetration rate depends on the screening properties of the respective organic structure on top of the silicon electrode. For example, a densely packed monolayer introduces a noticeable penetration barrier for the redox couple and, consequently, the intensity of oxidation and reduction peak current densities are decreased, and the associated peak current densities are shifted to higher or lower potentials, respectively. In Figure 5a representative

ACS Paragon Plus Environment

35

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 36 of 67

Figure 5. (a) Representative cyclic voltammograms of a silicon native oxide sample after Piranha treatment (black), a SAMPs (blue), a ║PNA (red, solid lines) and a ┴PNA sample (red, dashed lines) in a redox buffer solution presented for a scan rate of 400 mVs-1. (b) Individual reduction peak current densities are plotted against the square root of the corresponding scan rates v for each system (same color labels). The respective linear fits allow for a more detailed understanding of the change of the insulating properties (diffusion kinetics) of differently functionalized systems.

cyclic voltammograms of a silicon native oxide sample after Piranha treatment (black), a SAMPs (blue), a ║PNA (red, solid lines) and a ┴PNA sample (red, dashed lines) in a redox buffer solution (see Exp. Sec.) are presented for a scan rate of 400 mVs-1. The reduction peak current density of the SAMPs system (blue) is smaller compared to the bare silicon substrate (black) and, in addition, the corresponding peak potential is shifted slightly to more negative bias potentials (higher overpotential). Since the penetration rate of the Ru-

ACS Paragon Plus Environment

36

Page 37 of 67 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

complex depends on the insulating properties of the respective organic structure on top of the silicon electrode, the SAMPs layer presents a noticeable barrier even for the highly mobile redox complex in an aqueous environment. In the same time frame, only a fraction of Ru-complexes (smaller peak current density) can reach the SAMPs-coated surface and simultaneously require a higher driving force (more negative peak potential) compared to the bare silicon reference. The observed decrease in the reduction peak current density and a shift to more negative potential is preserved for the ║PNA system (red, solid lines), but the overall impact is small compared to the effects of the SAMPs layer. Interestingly, the insulating properties of the ┴PNA system (red, dashed lines) is noticeable different from the ║PNA analog (red, solid lines). The peak potential is shifted to more positive potentials (similar peak potential as the silicon reference), indicating that the immobilization of ┴PNA molecules results in an increased permeability for the Rucomplex with respect to the SAMPs layer (blue). Since the different organic layers exhibit adequate stability under our electrochemical measurement conditions (see Figure S3a), the observed disparity should be related to inherently different screening properties of the two PNA layers. In Figure 5b, the average reduction peak current densities (averaged

ACS Paragon Plus Environment

37

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 38 of 67

over at least three different measurements; see Figure S3b) of the differently functionalized systems are plotted against the square root of the respective scan rates (10-500 mVs-1), allowing for a more detailed description of the insulating properties. For all differently functionalized surfaces, a linear dependence can be observed, indicating that the electrons are transferred from the semiconductor electrode to the electrolyte and reduce the redox species in a diffusion-limited process (31). Compared to the transfer rates of the bare silicon substrate (black), the decrease for the ║PNA system (red, solid lines) is unexpectedly small relative to the observed decrease after SAMPs-grafting (blue). Consequently, the addition of the ║PNA overlayer onto the SAMPs system raises the penetration barrier for the Ru-redox couple only slightly, which is surprising, since AFM measurements indicate that the PNA layer is roughly 4-5 times thicker than the respective SAMPs coating (see Figure 1). These findings further support the idea of a spacious, netlike

║PNA

metastructure spanning over the linker-activated SAMPs-

terminated surface, which can be easily traversed by the Ru-complex compared to the densely-packed SAMPs layer. Interestingly, the change in transfer rates for the ┴PNA system is significantly smaller than the one of the ║PNA layer. Since XPS measurements

ACS Paragon Plus Environment

38

Page 39 of 67 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

suggested a comparable amount of ║PNA and ┴PNA molecules on the silicon substrates, we attribute this difference in surface permeability to the different PNA surface morphologies, which are related to the specific binding modes (at least in an aqueous environment). We propose that the rather poor insulting properties of the ┴PNA type are related to its rotational degrees of freedom in solution (see Scheme 1). This rotational mobility compared to the rigidity of the ║PNA network appears to influence the underlying organic layer so that the electrode surface can be more easily accessed by the redox complex. We assume that the free rotation of individual ┴PNA units introduces a stress to the underlying layer(s), resulting in a reduction of the density of the SAMPs layer and, thus, a lower penetration barrier. Electrochemical impedance spectroscopy. Electrochemical impedance spectroscopy (EIS) measurements were performed on representative silicon native oxide, SAMPs and PNA samples using the Teflon cell presented in Scheme 2. EIS raw data of the differently functionalized systems and corresponding fits illustrating the quality of the fitting algorithm can be found in the SI (Figure S4). The electrolyte was a 50 mM TRIS buffer solution containing 100 mM NaCl salt concentration (pH = 7.55, at 26.3 °C). The potential range

ACS Paragon Plus Environment

39

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 40 of 67

(-0.7 V to 0.5 V) was chosen so that the corresponding electrical regime at the Si/SiO2 interface spanned from weak inversion to accumulation. The total impedance Z between the electrolyte and the silicon substrate was recorded and converted to the admittance Y, which was then separated in conductance and capacitance terms Y = 1/Z = G + jωC. The modular, lumped elements equivalent circuit model presented in Figure 6a is further used to interpret the datasets. The model comprises four elements describing the bare Si/SiO2 reference system (prior to functionalization), whereas an additional 5th element is introduced to describe the organic parts of the system after functionalization (SAMPs or SAMPs/Linker/PNA); referred to as functional layer, FL). The individual elements represent specific parts of the samples (from left to right): the semiconductor (p-type silicon; purple), the dielectric (native oxide; blue), the functional layer(s) (SAMPs or SAMPs/Linker/PNA; green), the interface electrical double layer (EDL; red) and the bulk electrolyte (light blue). The region of the electrolyte, which is in direct contact with the dielectric or the SAMPs if present, is modeled by a double layer capacitance, CEDL, and a constant phase element (CPE or Warburg admittance) YW = QW (jω)1/2 =

QW (1+j) (ω/2)1/2. At each bias potential, the values of the circuit elements were extracted ACS Paragon Plus Environment

40

Page 41 of 67 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

to model the measurement data as described in detail in the SI. Figure S4a, b, c, and d demonstrate the adequacy of the model to achieve a high-quality fit of the admittance Y presented for the 4-element case of the reference system (Figure S4a) and the 5-element case for the biosensing systems with SAMP, and SAMP/Linker/PNAs (Figure S4b, c and d). After evaluating different descriptions of the native oxide layer, a constant oxide capacitance, irrespective of frequency and bias voltage proved to be sufficient to reproduce the measurement data. Figures 6b and c show the admittance’s real part (Re{Y(ω)}) and the imaginary part (Im{Y(ω)}) of each element in the model at two selected frequencies for the ║PNA (blue) and ┴PNA system (red). For each curve, the data is presented as a function of the bias potential ranging from negative (-0.7 V) to positive (+0.5 V) potentials, which is applied to the substrate terminal (working electrode, WE) with respect to the Ag/AgCl reference electrode (RE) (Scheme 2). The rightmost curves correspond to the total Re{Y(ω)} and Im{Y(ω)} between the outer terminals of the whole stack (WE and RE). The conductance of each cell, which is a representation of the respective cell admittance’s real part, is taken as independent of the frequency except for the interface electrical double layer cell. In the

ACS Paragon Plus Environment

41

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 42 of 67

interface electrical double layer cell, the contribution of a constant phase element (CPE) (mandatory to achieve a good fit for all spectra over the whole bias and frequency range) introduces power law frequency dependencies with non-integer exponents. Thus, curves

ACS Paragon Plus Environment

42

Page 43 of 67 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

ACS Paragon Plus Environment

43

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 44 of 67

Figure 6. (a) Schematic representation of the modular equivalent circuit used as a model to interpret the electrochemical impedance data. Each RC element represents a section of the system. From left to right: the semiconductor (p-type silicon; purple), the dielectric (native oxide; blue), the functional layer(s) (SAMPs or SAMPs/Linker/PNA; green), the interface with its electrical double layer (EDL) capacitance and CPE (red) and the bulk electrolyte (light blue). (b) The admittance’s real part of each equivalent circuit lumped element (conductance) extracted from model fits to the EIS data, at bias potentials ranging from deep depletion to accumulation (-0.7 to +0.5 V) for a ║PNA (blue) and a ┴PNA system (red) presented at two fixed AC signal frequencies of 100 mHz (squares) and 10 kHz (triangles). In the rightmost column the real part of the whole system’s total admittance is reported (labeled as “total” in the modular equivalent circuit). (c) The corresponding admittance imaginary part of each equivalent circuit lumped element (respective capacitance multiplied with ω) extracted for the two PNA systems (same bias range). The rightmost column reports the imaginary part of the whole system’s total admittance.

with squares and triangles overlap for all cells except the interface one. As mentioned above, the capacitance of the native oxide is essentially independent of bias potential and type of functionalization (i.e., Im{Y(ω)} scales proportional to the frequency, Figure 6c). The numeric value of COX = εox/tox is consistent with a thickness tox of about 1 nm; a small but non-negligible conductance GOX is thus expected and in fact observed (Figure 6b),

ACS Paragon Plus Environment

44

Page 45 of 67 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

likely due to direct carrier tunneling through the ultrathin oxide layer. Upon SAMPsgrafting, GOX increases by up to one order of magnitude, whereas further surface modification by PNA (Figure 6b, red and blue) does not change the native oxide conductance significantly. The reason for this behavior is not fully understood yet, but could be related to either the creation or the shift in energy of allowed electronic states in the SiO2 layer, which could assist in the electron tunneling through the thin oxide. Another explanation might be the introduction of a small strain to the native oxide layer (in solution), which results in a slightly increased transparency towards charge transport processes. Note that despite the attachment of the functionalization layer(s), in almost all cases GOX is providing the smallest conductance of all cells and, thus, sets an upper limit to the overall DC conductance of the whole stack of cells in series. In contrast, the silicon substrate conductance GP is relatively large (high doping concentration) and decreases at negative voltages due to the decreasing surface hole concentration (incipient surface depletion). As expected, a similar trend is observed for Im{Y(ω)} (substrate capacitance CP, Figure 6c), due to the formation of a depletion layer of increasing thickness for decreasing bias potentials. The different surface

ACS Paragon Plus Environment

45

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 46 of 67

functionalization variants introduce a well-detectable change of the GP and CP in the order to 30-50 % in the -0.4 V to 0 V bias potential regime. Probably, these changes reflect charge variations in the semiconductor induced by charge rearrangement at the dielectric/electrolyte interface, which is a promising result with respect to the utilization of this type of surface functionalization to produce receptor layers for sensing applications. In addition, the bulk electrolyte conductance (GE) is also relatively large due to the high salt concentration. In contrast, the electrolyte capacitance (CE), is rather small, due to the large distance between the sample and the reference electrode. Since CE cannot be reliably extracted from data only ranging up to 100 kHz, its numerical value was set to zero during the parameter extraction procedure and is, thus, not shown in Figure 6c. The CPE has a clear effect on the electrical double layer conductance GEDL, which exhibits a square root dependence on the angular frequency (approximately an increase of 2.5 orders of magnitude corresponding to the five orders of magnitude frequency increase from 100 mHz to 10 kHz). Interestingly, this trend cannot be observed for the imaginary part of the admittance (Figure 6c), suggesting that the contribution of CEDL,

ACS Paragon Plus Environment

46

Page 47 of 67 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

which is in parallel to the CPE, is in fact dominating Im{YEDL(ω)}. Thus, at a fixed frequency, the bias potential dependence of Im{YEDL} reflects the

Figure 7. Conductance GFL (a,b) and capacitance CFL (c,d) of the functionalized layer (SAMPs/Linker/PNA) extracted from model fits to the lumped element equivalent circuit presented in Figure 6a, for a forward bias potential sweep from -0.7 V to 0.5 V (a,c) and a reverse bias potential sweep from 0.5 V back to -0.7 V (b,d). Stable and repeatable results can be obtained for the ║PNA type, whereas an unstable behavior is observed for the ┴PNA type for the forward sweep.

ACS Paragon Plus Environment

47

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 48 of 67

dependence of CEDL and exhibits a weak minimum, as expected from Gouy-ChapmanStern theory of the EDL (31). Regarding the functionalized layer conductance GFL and capacitance CFL, we observe at both low and high frequencies significant larger numerical values compared to GOX and COX. The functionalized layer is, thus, not as effective in blocking the charge transfer as the native oxide. By measuring samples only functionalized with a SAMPs layer, we could verify that the overall conductance of the organic structure does not significantly change upon

║PNA

immobilization. Consequently, the overall conductance appears to be

governed by the SAMPs layer. This further supports the idea of a, generally, spacious and porous PNA metastructure, which does not fundamentally influence the charge transfer between the semiconductor and the electrolyte. For the ┴PNA (Figure 6b and c, red) system, we observed a noisy and unstable response in a bias potential regime ranging from approximately -0.3 V to 0.3 V, in which CFL increases, and GFL decreases appreciably up to the point where a reliable extraction of meaningful values becomes difficult. However, for ║PNA (Figure 6b and c, blue) systems,

ACS Paragon Plus Environment

48

Page 49 of 67 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

significantly more stable and repeatable measurements could be observed in the whole bias potential range along the sweeps from negative to positive voltages. In order to investigate this behavior, small signal EIS measurements were carried out for positive and negative bias potential sweep directions. Figure 7 compares the extracted values of GFL (a) and CFL (b) for a forward (-0.7 V to 0.5 V; left-hand side) and a reverse potential sweep (0.5 V to -0.7 V; right-hand side). We propose that the observed switching behavior of the ┴PNA from a noisy and variable to a stable conductance and capacitance state, which is also comparable to the overall behavior of the ║PNA system, is induced by a considerable surface reorientation of individual ┴PNA molecules resulting in an overall change of the ┴PNA metastructure. As previously discussed, due to the different binding modes, ┴PNA molecules can rotate freely around their binding axis, whereas ║PNA molecules are rigidly bound to the surface (no possible reorientation without decomposition). The non-existing (║PNA) and existing (┴PNA) switching behavior is likely a result of the different binding modes of the two PNA types (see Scheme 1) and of a bias potential-induced rearrangement of the ┴PNA molecules. Disorder in the ┴PNA layer during the reorientation at intermediate bias

ACS Paragon Plus Environment

49

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 50 of 67

potentials might also explain the observed instability of the functional layer capacitance

CFL of the ┴PNA (Figure 7c, red) compared to the monotonic behavior of CFL for the ║PNA (blue) system. However, a simple change of the overall interface thickness cannot fully explain the observed behavior. The ┴PNA layer does not only contain polarizable functional groups (large portion of the hetero-atoms), but also possesses rotational degrees of freedom and, therefore, can support a redirection of charges during the transition from depletion to accumulation to reach a new equilibration state at the interface. However, during the reverse potential sweep, the previously observed switching behavior of the ┴PNA does not occur anymore. These findings may suggest that at large absolute values of the bias potential the ┴PNA reaches a new equilibration state, which is stable on the time scale of the performed bias potential sweeps. Further studies have to be performed to identify the impact of the ion concentration and the type of ions on the observed switching behavior, which might open an additional pathway to further tune the electronic properties of PNA-based sensing devices.

ACS Paragon Plus Environment

50

Page 51 of 67 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

SUMMARY AND CONCLUSION We have investigated self-assembled monolayers of phosphonic acids (SAMPs) containing two adjacent ethylene glycol (EG) functionalities as starting platforms for the immobilization of differently modified PNA molecules (15 bases) on silicon native oxide surfaces as potential candidates for the fabrication of silicon nanowire sensor devices for label-free DNA detection. AFM and cyclic voltammetry measurements indicated the

ACS Paragon Plus Environment

51

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 52 of 67

formation of closed, densely-packed monolayers on the silicon native oxide with a thickness of roughly 0.3-0.5 nm as a reliable starting platform for further surface functionalization. The presence of this SAMPs made from 2-{2-[2-hydroxy-ethoxy]ethoxy}-ethyl

phosphonic

acid

molecules

was

further

substantiated

by

XPS

characterization, revealing characteristic spectroscopic features. Activation of the phosphonate monolayer was performed by a heterobifunctional maleimido compound. Subsequent functionalization with two different PNA oligonucleotides, a ║PNA and a ┴PNA type, was verified by contact angle, AFM, XPS and cyclic voltammetry measurements. Distinct water contact angles for ║PNA (52°) and ┴PNA type (38°) indicated different surface morphologies. High-resolution AFM measurements revealed that ║PNA molecules form extended, protruding and netlike 3D-metastructures with no preferential spatial direction, whereas the more flexible ┴PNA molecules produce spatially more isolated and anisotropic metastructures. Detailed XPS characterization could distinguish both PNA types due to their individual structural features and showed that the amount of surface-immobilized PNA molecules is roughly comparable. Additional cyclic voltammetry measurements in a redox buffer solution containing a small and highly

ACS Paragon Plus Environment

52

Page 53 of 67 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

mobile Ru-based complex revealed noticeably different insulating properties (diffusion kinetics) of both PNA systems, which we attributed to the disparate PNA metastructures resulting from the two specific binding modes. Impedance measurements carried out in positive and negative bias potential sweep directions (between -0.7 V and 0.5 V) on the differently functionalized surfaces could be analyzed by a modular lumped element equivalent circuit model using five individual elements describing the semiconductor, the dielectric (native oxide), the functional layer(s) (SAMPs or SAMPs/Linker/PNA), the interface electrical double layer (EDL), and the bulk electrolyte. For the ┴PNA type a switching behavior from a noisy and variable to a stable conductance and capacitance state was observed and attributed to a surface reorientation of individual ┴PNA molecules resulting in an overall change of the ┴PNA metastructure. In contrast, the overall more stable behavior of the ║PNA system suggested a higher stability of the corresponding surface. The different surface functionalization variants introduce a well-detectable change of the semiconductor conductance Gp and capacitance Cp in the range of 30-50 % in the (intermediate bias potential regime). Thus, the integration of an active sensor device in

ACS Paragon Plus Environment

53

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 54 of 67

the substrate, e.g., by standard CMOS technology, has the potential to detect the binding events by purely electrical label-free techniques. Based on these results, the modification of PNA backbones is a promising approach to control the orientation of bioreceptor binding sites on silicon surfaces, which may well become an important tool in the functionalization and thus fabrication of silicon nanowire sensor devices for label-free DNA detection.

Future studies will address the impact of the PNA binding mode-

dependent surface orientation on the hybridization affinity towards complementary DNA.

ASSOCIATED CONTENT

Supporting Information.

ACS Paragon Plus Environment

54

Page 55 of 67 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 following files are available free of charge.

Scheme S1 (PDF)

Scheme S2 (PDF)

Table S1 (PDF)

Figure S1 (PDF)

Figure S2 (PDF)

Table S2 (PDF)

Table S3 (PDF)

Figure S3 (PDF)

Figure S4 (PDF)

ACS Paragon Plus Environment

55

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 56 of 67

AUTHOR INFORMATION

Corresponding Author *E-Mail: [email protected]

Notes The authors declare no competing financial interest.

ACS Paragon Plus Environment

56

Page 57 of 67 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

ACKNOWLEDGEMENTS

We acknowledge financial support by the DFG (CA1076/5-1; TO266/7-1), support from the FLAG-ERA CONVERGENCE project via MIUR and the IUNET Consortium. The authors are grateful to C. Paulus, R. Csiki, C. Dore, S. Gremmo and G. Rziga for experimental support. D. Brandalise acknowledges financial support from the University of Udine and ARDISS and is grateful to Prof. Pierpaolo Palestri (University of Udine) for supervision.

ACS Paragon Plus Environment

57

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 58 of 67

ACS Paragon Plus Environment

58

Page 59 of 67 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

REFERENCES

(1) Southern, E. M. Detection of Specific Sequences Among DNA Fragments Separated by Gel Electrophoresis. J. of Mol. Biol. 1975, 98, 503-517.

(2) Piscevic, D.; Knoll, W.; Tarlov, M. J. Surface plasmon microscopy of biotinstreptavidin binding reactions on UV-photopatterned alkanethiol self-assembled monolayers. Supramol. Sci. 1995, 2 (2), 99-106.

(3) Nielsen, P. E.; Egholm, M. An introduction to peptide nucleic acid. Mol Biol. 1999,

1 (1-2), 89-104.

(4) Egholm, M.; Buchardt, O.; Christensen, L.; Behrens, C.; Freie, S. M.; Driver, D. A.; Berg, R. H.; Kim, S. K.; Norden B.; Nielsen, P. E. PNA hybridizes to complementary oligonucleotides obeying the Watson-Crick hydrogen-bonding rules. Nature 1993, 365, 566-568.

ACS Paragon Plus Environment

59

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 60 of 67

(5) Nielsen, P. E.; Egholm, M.; Berg R. H.; Buchardt; O. Sequence selective recognition of DNA by strand displacement with a thyminesubstituted polyamide.

Science 1991, 254, 1497-1500.

(6) Dragulescu-Andrasi, S.; Rapireddy, A.; Frezza, B. M.; Gayathri, C.; Gil, R. R.; Ly, D. H. A simple gamma-backbone modification preorganizes peptide nucleic acid into a helical structure. J. Am.Chem. Soc. 2006, 128, 10258-10267.

(7) Rapireddy, S.; He, G.; Roy, S.; Armitage, B. A.; Ly, D. H. Strand invasion of mixed-sequence B-DNA by acridine-linked, gamma-peptide nucleic acid (gamma-PNA). J. Am. Chem. Soc. 2007, 129, 15596-15600.

(8) Singh, R.P.; Oh, B.K.; Choi, J.W. Application of peptide nucleic acid towards development of nanobiosensor arrays. Bioelectrochemistry 2010, 79, 153–161.

(9) Hu, Q.; Wang, Q.; Sun, G.; Kong, J.; Zhang, X. Electrochemically Mediated Surface-Initiated de Novo Growth of Polymers for Amplified Electrochemical Detection of DNA. Anal. Chem. 2017, 89, 9253-9259.

ACS Paragon Plus Environment

60

Page 61 of 67 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

(10) Hu, Q.; Han, D.; Gan, S.; Bao, Y.; Niu, L. Surface-Initiated-Reversible-Addition– Fragmentation-Chain-Transfer Polymerization for Electrochemical DNA Biosensing. Anal. Chem. 2018, 90, 12207-12213.

(11) Dragulescu-Andrasi, A.; Zhou, P.; He, G.; Ly, D. H. Cell-permeable GPNA with appropriate backbone stereochemistry and spacing binds sequence-specifically to RNA. Chem. Commun 2005, 2, 244-246.

(12) Hahm, J.-in.; Lieber, C. M. Direct Ultrasensitive Electrical Detection of DNA and DNA Sequence Variations Using Nanowire. Nano Lett. 2004, 4, 51-54.

(13) Zhang, G.-J.; Zhang, G.; Chua, J. H.; Chee, R.-E.; Wong, E. H.; Agarwal, A.; Buddharaju, K. D.; Singh, N.; Gao, Z.; Balasubramanian, N. DNA Sensing by Silicon Nanowire: Charge Layer Distance Dependence. Nano Lett. 2008, 8, 1066-1070.

ACS Paragon Plus Environment

61

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 62 of 67

(14) De, A.; van Nieuwkasteele, J.; Carlen, E. T.; van den Berg, A. Integrated labelfree silicon nanowire sensor arrays for (bio) chemical analysis. Analyst 2013,

138(11), 3221-3229.

(15) Sørensen, M. H.; Mortensen, N. A.; Brandbyge, M. Screening model for nanowire surface-charge sensors in liquid. Appl. Phys. Lett. 2007, 91, 102105.

(16) Stern, E.; Wagner, R.; Sigworth, F. J.; Breaker, R.; Fahmy, T. M; Reed, M. A. Importance of the Debye Screening Length on Nanowire Field Effect Transistor Sensors. Nano Lett 2007, 7, 3405-3409.

(17) De, A.; Souchelnytskyi, S.; van den Berg, A.; Carlen, E. T. Peptide Nucleic Acid (PNA)-DNA Duplexes: Comparison of Hybridization Affinity between Vertically and Horizontally Tethered PNA Probes. ACS Appl. Mater. Interfaces 2013, 5

(15), 7659-7659.

(18) Bartl, J. D.; Stutzmann, M.; Cattani-Scholz, A.; Tornow, M. Synthesis and optimization of organic sensing platforms for label-free DNA detection. IEEE 12th

ACS Paragon Plus Environment

62

Page 63 of 67 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

International Conference on Nano/Micro Engineered and Molecular Systems (NEMS), Los Angeles, USA, 2017, 118-121.

(19) Cattani-Scholz, A. Functional Organophosphonate Interfaces for Nanotechnology: A Review. ACS Appl. Mater. Interfaces 2017, 9(31), 2564325655.

(20) Cattani-Scholz, A.; Pedone, D.; Dubey, M.; Neppl, S.; Nickel, B.; Feulner, P.; Schwartz, J.; Abstreiter, G.; Tornow, M. Organophosphonate-Based PNA Functionalization of Silicon Nanowires for Label-Free DNA Detection. ACS Nano 2008, 2, 1653-1660.

(21) De, A.; Keim, K.; Tornow, M.; Cattani-Scholz, A. Horizontal γ-PNA immobilization through organophosphonate chemistry for biosensing applications. 2015 IEEE 15th International Conference on Nanotechnology

(IEEE-NANO), Rome, Italy, 2015, 1568-1571.

ACS Paragon Plus Environment

63

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 64 of 67

(22) Silverman, B. M.; Wieghaus, K. A.; Schwartz, J. Comparative Properties of Siloxane vs Phosphonate Monolayers on A Key Titanium Alloy. Langmuir 2005,

21, 225-228.

(23) Cattani-Scholz, A.; Pedone, D.; Blobner, F.; Abstreiter, G.; Schwartz, J.; Tornow, M.; Andruzzi, L. PNA-PEG Modified Silicon Platforms as Functional BioInterfaces for Applications in DNA Microarrays and Biosensors.

Biomacromolecules 2009, 10, 489-496.

(24) Ogaki, R.; Zoffmann-Andersen, O.; Jensen, G. V.; Kolind, K.; Kraft, D. C. E.; Pedersen, J. S.; Foss, M. Temperature-Induced Ultradense PEG Polyelectrolyte Surface Grafting Provides Effective Long-Term Bioresistance against Mammalian Cells, Serum, and Whole Blood. Biomacromolecules 2012, 13, 3668-3677.

(25) Choi, N.-S.; Lee, Y. M.; Cho, K. Y.; Ko, D.-H; Park, J.-K. Protective layer with oligo(ethylene glycol) borate anion receptor for lithium metal electrode stabilization. Electrochem. Commun. 2004, 6, 1238-1242.

ACS Paragon Plus Environment

64

Page 65 of 67 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

(26) Sharma, C.; Awasthi, S. K. Versatility of peptide nucleic acids (PNAs): role in chemical biology, drug discovery, and origins of life. Chem. Biol. Drug. Des. 2017, 89, 16-37.

(27) Cattani-Scholz, A.; Liao, K.-C.; Bora, A.; Pathak, A.; Krautloher, M.; Nickel, B.; Schwartz, J.; Tornow; M.; Abstreiter, G. New Molecular Architecture for Molecular Electronics. Angew. Chem. Int. Ed 2011, 50, A11-16.

(28) Wanger, C. D.; Riggs, W. M.; Davis, L. E.; Moulder, J. F.; Muilenberg, G. E.

Handbook of X-ray Photoelectron Spectroscopy; Perkin-Elmer Corp., Physical Electronics Division, Eden Prairie, Minnesota USA, 1979.

(29) Lindberg, B. J.; Hamrin, K.; Johansson, G.; Gelius, U.; Fahlman, A.; Nordling, C.; Siegbahn, K. Molecular Spectroscopy by Means of ESCA II. Sulfur compounds. Correlation of electron binding energy with structure. Physica

Scripta 1970, 1, 286.

ACS Paragon Plus Environment

65

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 66 of 67

(30) Gouzman, I.; Dubey, M.; Carolus, M. D.; Schwartz, J.; Bernasek, S. L. Monolayer vs. multilayer self-assembled alkylphosphonate films: X-ray photoelectron spectroscopy studies Surf. Sci. 2006, 600, 773-781.

(31) Bard A. J.; Faulkner L. R. Electrochemical Methods: Fundamentals and

Applications, Second edition; John Wiley & Sons, Inc., New York USA, 2001.

ACS Paragon Plus Environment

66

Page 67 of 67 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

TOC Graphic. Schematic representation of PNA surface grafting by means of phosphonate chemistry (top). Equivalent circuit model describing the individual parts of the PNA device (bottom).

ACS Paragon Plus Environment

67