Refractive Index Sensing Using Visible Electromagnetic Resonances

Jan 30, 2017 - Plasmonic metal nanostructures, in colloidal or surface-supported forms, have been extensively studied in the context of metamaterials ...
0 downloads 11 Views 1MB Size
Subscriber access provided by Van Pelt and Opie Library

Article

Refractive index sensing using visible electromagnetic resonances of supported CuO particles 2

Mariano Daniel Susman, Alexander Vaskevich, and Israel Rubinstein ACS Appl. Mater. Interfaces, Just Accepted Manuscript • DOI: 10.1021/acsami.6b15726 • Publication Date (Web): 30 Jan 2017 Downloaded from http://pubs.acs.org on January 30, 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.

ACS Applied Materials & Interfaces 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 32

ACS Applied Materials & Interfaces

1 2 3 4 5 6

Refractive index sensing using visible 7 8 9 10 1 12

electromagnetic resonances of supported Cu2O

13 14 15

particles 16 17 18 19 20 21 22 23 24 25 26 27 28 29

Mariano D. Susman,* Alexander Vaskevich, Israel Rubinstein Department of Materials and Interfaces, Weizmann Institute of Science, Rehovot 7610001, Israel KEYWORDS. Cuprous oxide, high-n dielectric, Mie scattering, magnetic dipole resonance, sensing.

30 31 32 3

ABSTRACT. Plasmonic metal nanostructures, in colloidal or surface-supported forms, have been 34 35 36

extensively studied in the context of metamaterials design and applications, in particular as 37 38

refractometric sensing platforms. Recently, high refractive index (high-𝑛) dielectric sub39 40 41 42 43

wavelength structures have been experimentally shown to support strong Mie scattering resonances, predicted to exhibit analogous refractive index sensing capabilities. Here we present

4 45

the first experimental demonstration of the use of supported high-𝑛 dielectric nano/microparticle 46 47 48

ensembles as refractive index sensing platforms, using cuprous oxide as a model high-𝑛 material. 49 50

Single-crystalline Cu2O particles were deposited on transparent substrates using a chemical 51 52

deposition scheme, showing well-defined electric and magnetic dipolar resonances (EDR and 5

54

53

MDR, respectively) in the visible range, which change in intensity and wavelength upon changing 56 57 58

1 59 60 ACS Paragon Plus Environment

ACS Applied Materials & Interfaces

Page 2 of 32

1 2 3

the medium refractive index (𝑛𝑚 ). The significant modulation of the MDR intensity when 𝑛𝑚 is 4 5

modified appears to be the most valuable empirical sensing parameter. The Mie scattering 6 7

properties of Cu2O particles, particularly the spectral dependence of the MDR on 𝑛𝑚 , are 8 9 10

theoretically modelled to support the experimental observations. MDR extinction changes (i.e., 1 12 13

refractive index sensitivity) per particle are >100 times higher compared to localized surface

14 15

plasmon resonance (LSPR) changes in supported Au nanoislands, encouraging the evaluation of 16 17

Cu2O and other high-𝑛 dielectric particles and sensing modes in order to improve the sensitivity 18 19 20 21 22 23 24 25 26 27 28 29

in optical (bio)sensing applications.

30 31 32 3 34 35 36 37 38 39 40 41 42 43 4 45 46 47 48 49 50 51 52 53 54 5 56 57 58

2 59 60 ACS Paragon Plus Environment

Page 3 of 32

ACS Applied Materials & Interfaces

1 2 3

INTRODUCTION 4 5

High refractive index (high-𝑛) dielectric sub-wavelength particles have recently gained 6 7

increasing scientific interest as novel optical metamaterials and in nanophotonic devices, such as 8 9 10

all-dielectric nanoantennas active in the optical frequency domain,1, 1 12 13

2

nanolasers3,

4

and

waveguides.5 The directional light scattering by high-𝑛 dielectric nanoantennas in the visible range

14 15

has also been the subject of research,6-8 demonstrating that high-𝑛 particles can fulfill the first 16 17

Kerker condition of zero backscattering formulated for magnetic particles.6, 9 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32

While metal nanostructures have been extensively studied for controlling and manipulating light on the nanoscale, dielectric materials show lower dissipative losses and fluorescence quenching properties than their plasmonic counterparts, potentially allowing to overcome certain limitations of plasmonic structures in photonic applications. Mie scattering solutions to Maxwell’s equations for the interaction of electromagnetic radiation with sufficiently large spherical high-𝑛 dielectric particles are known to theoretically predict the

3 34

existence of magnetic and electric dipole resonances (MDR and EDR, respectively), as the first 35 36

and second lowest energy resonances.10-12 The excitation of displacement currents inside the 37 38

particles by the incident radiation leads to a transversal magnetic field, acting as an oscillating 39 40 41 42 43

dipole and giving rise to MDRs at appropriate excitation frequencies.13 In spherical particles, the MDR is expected to occur when the wavelength of the light inside the particle (𝜆𝑝 ) is equal to the

4 45 46

particle diameter (𝑑), i.e., 𝜆𝑝 = 𝜆0 ⁄𝑛 ≈ 𝑑,12, 14 where 𝜆0 is the wavelength in free-space. The 47 48

resulting artificial magnetism in the optical range can be used to design materials with useful 49 50 51 53

52

properties, such as negative refractive index metamaterials, with applications in cloaking and superlensing.5

54 5 56 57 58

3 59 60 ACS Paragon Plus Environment

ACS Applied Materials & Interfaces

Page 4 of 32

1 2 3

Prominent MDRs and EDRs in high-𝑛 dielectric particles have been recently demonstrated for 4 5

SiC,15 barium strontium titanate,16 Te,17 and Si,14, 18 among others, in a wide electromagnetic 6 7

spectral range, from the microwave to the visible. Due to its very high refractive index (𝑛 ~ 3.8 at 8 9 10

633 nm),1 Si shows very strong resonances in the Vis-NIR, measured on the single particle level 1 12 13

using dark-field microscopy8 and in highly monodisperse colloidal Si dispersions by transmission

14

measurements.19 15 16 17

The energy band-gap and the 𝑛 value of semiconducting/insulating materials are inversely 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32

correlated,20 whilst many semiconducting materials commonly show a relatively high 𝑛.20, 21 Copper(I) chalcogenides, which are typically p-type semiconductors, exhibit moderately high refractive indices, ~2.5 at 500 nm.22 These materials are expected to present Mie scattering resonances. However, the optical properties of Cu2-xS, Cu2-xSe, and Cu2-xTe nanocrystals have been shown to be strongly dependent on the degree of self-doping (x).23 The higher the Cu(I) vacancy concentration, the higher the free carrier density, resulting in the apparition of localized surface

3 34

plasmon resonance (LSPR) bands in the NIR. Metal oxides have also been proposed to display 35 36

similar plasmonic properties.24, 25 For such materials, particularly with large particle sizes, it may 37 38

be challenging to distinguish between the sources of resonance, possibly resulting from a 39 40 41 42 43

combination of plasmonic features and optical Mie scattering resonances. Cuprous oxide (Cu2O) is a direct band-gap p-type semiconductor26, 27 (Eg ~ 2.1 eV) with a

4 45

refractive index similar to those of the chalcogenides. However, its typical free carrier 46 47 48

concentration is relatively low26 and it is harder to dope; therefore, the frequently observed 49 50

resonant bands in the Vis-NIR spectrum of Cu2O particles have been seldom attributed to LSPRs.28 51 53

52

Such resonances have been better explained by Mie scattering processes from sufficiently large 5

54

particles,29-33 for which the experimental allocation of EDR and MDR bands in this oxide has been 56 57 58

4 59 60 ACS Paragon Plus Environment

Page 5 of 32

ACS Applied Materials & Interfaces

1 2 3

presented only recently, using single-particle spectroscopy.34 Nanoparticulate Cu2O has also been 4 5

shown to be SERS active,35 as well as Si particles.36, 37 Notably, determining whether resonance 6 7

bands can be assigned to plasmonic features or to dielectric dipolar resonances may be useful in 8 9 10

improving the properties of related optical devices. 1 12 13

García-Cámara et al. have modeled the optical spectra of Si nanospheres in different dielectric

14 15

media, predicting that the spectral properties of high-𝑛 dielectric particles may be suitable for 16 17

refractometric sensing.21 Other sensing modes, involving polarized light, were also theoretically 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32

explored.38 Nonetheless, we are aware of only one demonstration of the sensing potential of high𝑛 all-dielectric platforms, carried out at the single particle level,39, 40 with no such reports on ensemble measurements. Plasmon-based refractometric sensing technologies,41 including surface plasmon resonance (SPR) and LSPR, the latter based primarily on Au and Ag nanostructures (and potentially on Cu, Al and Pd),42-45 have been studied extensively as label-free transducers for sensing of chemical

3

and biological interactions.46 High-𝑛 dielectric particles are potential candidates for competing 34 35 36

with the refractometric plasmonic-based technologies, possibly under different measurement 37 38

conditions. 39 40 41 42 43

In this work, we experimentally demonstrate the presence of intense optical scattering peaks in supported Cu2O crystal ensembles prepared by chemical depositions (CD) on transparent

4 45

substrates by performing combined collimated and integrating-sphere measurements. The 46 47 48

observed scattering bands are attributed to electric and magnetic Mie scattering dipole resonances 49 50

using qualitative comparison of the experimental spectra with full Mie scattering solutions for 51 52

spherical Cu2O particles. We evaluate experimentally the refractive index sensing properties of 5

54

53

high-𝑛 dielectric particle ensembles by using Cu2O as a model material. The optimal refractive 56 57 58

5 59 60 ACS Paragon Plus Environment

ACS Applied Materials & Interfaces

Page 6 of 32

1 2 3

index sensing mode is found to be the change in the MDR extinction intensity, rather than the 4 5

commonly used (in LSPR sensing) frequency shift. The sensitivity per particle is shown to be more 6 7

than 100 times higher for Cu2O particles compared to common supported Au LSPR transducers.47 8 9 10 1 12 13

RESULTS AND DISCUSSION

14

Supported single-crystalline Cu2O particles were produced according to a previously published 15 16 17

chemical deposition (CD) procedure.43, 48 In order to obtain particles of different sizes, Au-seeded 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32

substrates were immersed for increasing deposition times in a CD solution designed for depositing truncated octahedral particles. Of the various attainable Cu2O morphologies exposing {100} and {111} facets, the truncated octahedral one was chosen based on to its resemblance to spherical morphology. HRSEM images of Cu2O particles supported on quartz (Figure 1) show particles of increasing equivalent diameter (the equivalent diameter is the diameter of a circle of equivalent top-view

3

area), from 253±23 nm to 449±34 nm, corresponding to samples K1 to K4. Respective histograms 34 35 36

are presented in Figure 1E. The particles exhibit a narrow size distribution, with a polydispersity 37 38

of 350 particles). 19 20 21 22 23 24 25 26 27 28 29

Table 1. Quartz- and glass-supported Cu2O particle sizes and spatial distributions, obtained from HRSEM image analyses.

30 31 32 3 34 35

Equivalent Sample Substrate diameter (nm)

Equivalent geometrical cross-section (nm2)

Height (nm)

1.06 × 104 1.67 × 104 2.01 × 104 2.57 × 104 5.03 × 104 8.87 × 104 13.20 × 104 15.83 × 104

52 66 72 81 114 151 185 202

*

Particle surface density (m-2)

Surface coverage (%)

26.0 22.1 21.4 14.9 1.84 2.63 2.30 2.34

22.5 27.2 38.2 30.2 9.1 18.5 21.9 32.3

36 37 38 39 40 41 42 43 4 45 46

G1 G2 G3 G4 K1 K2 K3 K4

Glass Glass Glass Glass Quartz Quartz Quartz Quartz

116 ± 11 146 ± 13 160 ± 14 181 ± 14 253 ± 23 336 ± 25 410 ± 34 449 ± 34

47 48

*The particle height was estimated as 45% of the equivalent diameter. 49 50 51 52 53 54 5 56 57 58

7 59 60 ACS Paragon Plus Environment

ACS Applied Materials & Interfaces

Page 8 of 32

1 2 3

A

4 5 6 7 8 9 10

300 nm

1 12

B

13 14 15

300 nm

16 17

Figure 2. (A) Tilted angle (75º) and (B) cross-sectional HRSEM images of Cu2O particles of 18 19 20 21 22 23 24 25 26 27 28 29

449±34 nm equivalent diameter deposited on quartz. Particles are truncated by the substrate, with

30

(𝑄𝑒𝑥𝑡 , 𝑄𝑠𝑐𝑎 , 𝑄𝑎𝑏𝑠 ) were experimentally determined by combining regular collimated transmission 31 32 3

a height of ca. 45% of the equivalent diameter. As particles become larger, the extinction properties are expected to be increasingly dominated by Mie scattering rather than by absorption. Extinction, scattering and absorption efficiencies

experiments (shown schematically in Figure 3A) with diffuse transmittance and reflectance

34

measurements performed using an integrating sphere detector (Figure 3B-D). The collimated 35 36 37

transmission (𝜏𝑐 ) is related to the sample extinction (E) by E = −𝑙𝑜𝑔 𝜏𝑐 . 38 39

The integrating sphere setup enables determination of the total reflectance (𝜌𝑇 ) by collecting the 40 41 42 43 4

hemi-spherically back-scattered light, which includes both the diffuse and specular reflectance (Figure 3B); as well as the total transmittance (𝜏 𝑇 ), which includes the hemi-spherically forward

45 46

scattered and non-interacting transmitted components (Figure 3C). Exclusion of the purely 47 48 49

transmitted fraction allows estimation of the diffuse transmittance (𝜏𝑑 ) (Figure 3D). 50 51 52 53 54 5 56 57 58

8 59 60 ACS Paragon Plus Environment

Page 9 of 32

ACS Applied Materials & Interfaces

1 2 3

B

A

4 5

I0

6

I

7

Reference beam

Specular reflection

D

D

Detector Sample

8 9 10

C

1

D

12 13

D

14

D

15 16

Spectralon standard

17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32

Figure 3. Schematic presentation of the spectrophotometric measurement configurations used for estimating absorption, scattering and extinction efficiencies. (A) Collimated transmission, (B) total reflection (diffuse + specular reflection), (C) total transmission (diffuse transmitted + noninteracting light), and (D) diffuse transmission. I0 is the incident irradiance, I is the detected noninteracting irradiance, and the orange halo represents the light scattered by the sample. Spectralon was used as reflectance standard. The sample is shown in the forward orientation.

3 34

Raw spectra corresponding to the optical measurements described in Figure 3 are presented in 35 36

Figure 4. Slides were measured in the forward and backward positions, to compare the effect of 37 38 39

the orientation of the particle layer with respect to the substrate. 40 41 42

For large Cu2O particles on quartz, the extinction spectra (Figure 4A) are characterized by the

43 4

presence of two prominent peaks in the range 500-900 nm, that red-shift and increase in intensity 45 46

as the particles’ equivalent diameter becomes larger. The peaks tend to overlap and then disappear 47 48

when the particle size is further decreased, as observed for sample K1 and for smaller glass49 50 51 53

52

supported particles (Figure S2A, Supporting Information). The spectra differ insignificantly for the two slide orientations.

54 5 56 57 58

9 59 60 ACS Paragon Plus Environment

ACS Applied Materials & Interfaces

Page 10 of 32

1 2 3

A

4

B

0.8

5

100

30

0.6

7 8

Q3

0.4

Q2

0.2

9

0.0 200

1

400

800

Q3 Q1

10

1000

1200

No sample

400

Wavelength (nm)

C

90

14

80 Q1

T (%)

15 16

70

17 18

50 0

800

50

Q4

40

Q2

30

Q3 Q2

20 10 Blocked beam

400

20 21 22 23 24 25 26 27 28 29

700

No sample (w/ Spectralon) Quartz (w/ Spectralon)

90

Q4

19

D

Quartz

Q3 60

600

Wavelength (nm)

No sample

13

500

100

d (%)

12

100

Q2

Quartz

0 600

Q4

20

Q1

Quartz

10

T (%)

Q4

Extinction

6

Spectralon

500

600

700

Wavelength (nm)

800

0

Q1

Quartz (w/o spectralon)

400

500

600

700

800

Wavelength (nm)

Figure 4. Raw spectra representing (A) extinction, (B) total reflectance, (C) total transmittance, and (D) diffuse transmittance, measured as shown in Figure 3, for quartz-supported Cu2O particles of (K1) 253±23 nm, (K2) 336±25 nm, (K3) 410±34 nm, and (K4) 449±34 nm equivalent diameter, in the back (full lines) and front (dashed lines) orientations.

30 31 32 3

The extinction spectra at 500 nm, in the spectral range of the large extinction bands. This observation provides a preliminary indication that these peaks may be attributed to scattering processes. The diffuse reflectance is higher for back-oriented slides than

3

for front-oriented ones, particularly that of the highest energy peak. Inversely, the total diffuse 34 35 36

transmittance (Figure 4C) is in general more intense for front-oriented slides. 37 38

The diffuse transmittance spectra (Figure 4D) show similar spectral features to those of the 39 40 41 42 43

diffuse reflectance spectra (Figure 4B), exhibiting better resolution of the two peaks. The higher efficiency of diffuse transmission compared to diffuse reflectance for large particles may be related

4 45

to improved forward scattering compared to back scattering, which is characteristic of the Mie 46 47 48

scattering regime.8, 11 As is the case with the total diffuse transmittance (Figure 4C), front-oriented 49 50

slides generally display higher diffuse transmission compared to back-oriented slides (Figure 4D). 51 52

The above orientation-dependent results suggest that the presence of the substrate alters the 53 5

54

angular distribution of scattered light, to which the integrating sphere is particularly sensitive. 56 57 58

11 59 60 ACS Paragon Plus Environment

ACS Applied Materials & Interfaces

Page 12 of 32

1 2 3

Smaller sized particles prepared on glass substrates barely show extinction signals at >500 nm 4 5

(Figure S2A, Supporting Information), although these samples display prominent scattering in this 6 7

spectral range by integrating sphere experiments (Figure S2B and D, Supporting Information). 8 9 10

Note that the higher reflectance of the glass substrates compared to quartz ones is attributed to 1 12 13

the smaller glass thickness, leading to lower light absorption by the substrate while promoting

14

multiple internal reflections, hence higher back-reflection. 15 16 17

Although the highest hemi-spherical scattering is observed for front-oriented samples measured 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32

in the forward direction (Figure 4D), a large part of the corresponding near-field is located in a region occupied by the substrate. In the context of sensing, the back-orientation represents the most useful configuration, as the forward-scattered field, predominant in the Mie scattering regime, would be forward-directed, pointing toward the medium. For this reason the back orientation was chosen for continuing the optical analysis. To estimate the absorption and scattering components, for each measurement mode, the

3

corrected sample response (𝑃) was calculated from the raw data by correcting with respect to 0% 34 35 36

and 100% spectral references (Figure S3, Supporting Information), thus considering the substrate 37 38

and setup contributions, using the following expression: 39 40 41 42 43

𝑃 = 𝑃100% 𝐼

𝐼𝑠 −𝐼0% 100% −𝐼0%

(1)

where 𝐼𝑠 , 𝐼0% and 𝐼100% are the raw measured spectra of the sample, the 0% and 100% references, 4 45 46

respectively, and 𝑃100% is the spectral characteristics of the 100% standard used (spectralon). 47 48

Corrected spectra are shown in Figures S4 and S5 (Supporting Information). 49 50 51 53

52

As expected, at low surface coverages 𝜏𝑑 < 𝜏𝑐 < 𝜏 𝑇 , as most of the light is transmitted without major interactions. At larger coverages 𝜏𝑑 > 𝜏𝑐 , as the fraction of directly transmitted light

54 5

decreases. The collimated measurements are largely correlated with the 𝜏𝑑 component, mainly 57

56 58

12 59 60 ACS Paragon Plus Environment

Page 13 of 32

ACS Applied Materials & Interfaces

1 2 3

because 𝜌𝑇 < 𝜏𝑑 and as 𝜌𝑇 is comparatively smooth. As a result, the bands observed in the 4 5

collimated transmission are largely defined by forward scattering processes. 6 7

The law of conservation of energy dictates that 1 = 𝛼 + 𝑠 + 𝜏, where 𝛼 is the absorptance, 𝑠 is 8 9 10

the scattering, and 𝜏 is the transmittance of the sample; hence, the previous measurements can be 1 12 13 14 15

used to estimate the absorption component. Given that 𝜌𝑇 + 𝜏 𝑇 ≅ 𝑠 + 𝜏, 𝛼 can be estimated as 𝛼 = 1 − 𝜌𝑇 − 𝜏 𝑇 . The extinction, absorption and scattering cross-sections, 𝜎 (and therefore the

16 17

corresponding efficiencies) can then be estimated from the previously calculated corrected spectral 18 19 20 21 22 23 24 25 26 27 28 29

responses as: E = −𝑙𝑜𝑔 𝜏𝑐 = 𝜎𝑒𝑥𝑡 𝑁

(2)

−𝑙𝑜𝑔(𝜌𝑇 + 𝜏 𝑇 ) = 𝜎𝑎𝑏𝑠 𝑁

(3)

−𝑙𝑜𝑔(1 − 𝜌𝑇 − 𝜏 𝑇 ) = 𝜎𝑠𝑐𝑎 𝑁

(4)

where 𝑁 is the particle surface density.

30 31 32 3

Contributions from each optical process, i.e., extinction, scattering and absorption efficiencies calculated using equations 2 to 4, are summarized in Figure 5 and Figure S6, Supporting

34 35

Information. The increasing contribution of scattering in the >500 nm region is unequivocally 36 37 38

attributed to Mie scattering processes. 39 40

Considering that the extinction values are