Organic Ice Resists - Nano Letters (ACS Publications)

Nov 20, 2017 - We discovered that frozen layers of simple organic molecules can be patterned by an electron beam for lithography applications. Most ex...
0 downloads 30 Views 2MB Size
Subscriber access provided by Queen Mary, University of London

Communication

Organic ice resists William Tiddi, Anna Elsukova, Hoa Thanh Le, Pei Liu, Marco Beleggia, and Anpan Han Nano Lett., Just Accepted Manuscript • DOI: 10.1021/acs.nanolett.7b04190 • Publication Date (Web): 20 Nov 2017 Downloaded from http://pubs.acs.org on November 28, 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.

Nano Letters 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 13

Nano Letters

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

1

Organic ice resists

2 3

William Tiddi, Anna Elsukova, Hoa Thanh Le, Pei Liu, Marco Beleggia, Anpan Han*

4 5

DTU Danchip/CEN, Technical University of Denmark, 2800 Kgs. Lyngby, Denmark

6

Abstract

7

Electron-beam lithography (EBL) is the backbone technology for patterning nanostructures and

8

manufacturing nanodevices. It involves processing and handling synthetic resins in several steps, each

9

requiring optimization and dedicated instrumentation in cleanroom environments. Here, we show that

10

simple organic molecules, e.g. alcohols, condensed to form thin-films at low temperature demonstrate

11

resist-like capabilities for EBL applications and beyond. The entire lithographic process takes place in a

12

single instrument, and avoids exposing chemicals to the user and the need of cleanrooms. Unlike EBL

13

that requires large samples with optically flat surfaces, we patterned on fragile membranes only 5-nm-

14

thin, and 2 x 2 mm2 diamond samples. We created patterns on the nm to sub-mm scale, as well as three-

15

dimensional structures by stacking layers of frozen organic molecules. Finally, using plasma etching, the

16

organic ice resist (OIR) patterns are used to structure the underlying material, and thus enable

17

nanodevice fabrication.

18

Keywords

19

Electron-beam lithography, condensed organic molecules, ice lithography, focused electron-beam induced

20

deposition, 3D lithography, nanostructured diamond

21

Text

22

Electron-beam patterning at the nanometer scale enables research areas such as plasmonics and

23

metamaterials1, optomechanical2 and nanofluidics3 systems, quantum computing4, advanced electronics

24

and 2D material devices5. The electron beam lithography (EBL) process (Figure 1a) includes four steps: spin-

25

coating resin-based resists on optically flat surfaces, resist baking, electron beam (e-beam) exposure, and

1 ACS Paragon Plus Environment

Nano Letters

Page 2 of 13

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

1

development in liquid followed by sample drying. Crucial to the success of the lithography process, these

2

steps require cleanroom environment for reduced particle contamination, and dedicated equipment and

3

chemicals. Resist processing also restricts the choice of sample materials and geometries. Another e-beam

4

based nanopatterning method is focused electron-beam induced deposition6 (FEBID, Figure 1b), in which a

5

precursor gas is introduced into the EBL instrument through a nozzle in close proximity to the sample

6

surface. The energetic e-beam dissociates the gas molecules adsorbed on the surface and deposits a solid

7

at the exposure site. Most of the gas is unreacted and removed by pumps. FEBID’s ultimate resolution is

8

better than EBL7-8, and the lack of spin-coating and development enables patterning on complex sample

9

topographies and creating 3D nanostructures9. However, the simultaneous interaction between gas

10

precursor, sample surface and the e-beam electrons to form the solid deposit results in processing times

11

up to 5 orders of magnitude longer than EBL. In ice lithography10 (IL, Figure 1c), water vapour condenses

12

into a solid layer of ice that coats the sample held at cryogenic temperatures. The e-beam removes the ice

13

and uncovers the underlying substrate, and the resulting patterns are equivalent to those from a positive-

14

tone EBL resist. To realize metal nanostructures, maintaining cryogenic conditions and vacuum, the sample

15

is then subjected to a metallization and lift-off process where the patterned ice acts as mask. Since no resist

16

application is required, IL allows nanopatterning on non-planar or fragile samples such as freely suspended

17

carbon nanotubes11. Unfortunately, since most downstream nanofabrication instruments work at ambient

18

conditions, the need for cryogenic in-situ processing limits IL applications.

19 20

We discovered that frozen layers of simple organic molecules can be patterned by an electron beam for

21

lithography applications. Most experiments were performed in a custom instrument consisting of a

22

scanning electron microscope (SEM) modified to include a liquid nitrogen cooled cryostage, a gas injection

23

system (GIS), an e-beam scan control system, and an airlock for sample exchange. Design, implementation

24

and operation of the experimental setup are described in detail elsewhere12. In brief, to form the OIR film,

25

we cool down the sample in the SEM vacuum, and we direct the organic gas through the GIS nozzle onto

2 ACS Paragon Plus Environment

Page 3 of 13

Nano Letters

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

1

the cold sample surface, where it condenses to a uniform layer of ice. We irradiate the OIR thin-film with

2

the e-beam, and then we transfer the sample to the airlock to heat it to room temperature while

3

maintaining vacuum. In the e-beam exposed areas a solid product is formed, whereas the unexposed OIR

4

sublimates within minutes, and it is removed by the vacuum system of the airlock (Figure 1d). The resulting

5

patterns are stable at ambient conditions. As in IL and FEBID, this sublimation step completely replaces the

6

solvent development needed in EBL. In strong contrast to IL, our e-beam patterned OIR can be used ex-situ

7

for further characterization and processing, as the resulting patterns are equivalent to those made by a

8

negative-tone EBL resist. Because the entire lithography happens in one single instrument, cleanroom

9

environment and additional resist processing instrumentation are not required, thereby streamlining

10

nanofabrication.

11

12 13

Figure 1. OIR patterning compared to other e-beam based techniques. (a) EBL – The sample is coated with a resin layer that

14

changes its solubility when exposed to the e-beam. The sample is then removed from the vacuum instrument and the pattern is

15

developed with dedicated solvents. (b) FEBID – A precursor is introduced in the vacuum chamber through a nozzle, it is adsorbed

16

onto the surface and dissociates due to the electron bombardment during exposure, depositing a non-volatile product. (c) IL –

17

Water vapour is condensed onto a cryogenically cooled sample to form an ice thin-film, which is removed by the e-beam. The

18

patterned ice is used as mask for in-situ processes before bringing the sample back to room temperature. (d) OIR – A vapour of a

3 ACS Paragon Plus Environment

Nano Letters

Page 4 of 13

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

1

simple organic compound is first condensed onto the cooled sample to form a uniform layer. Its interaction with the e-beam locally

2

modifies its chemical composition resulting in a non-volatile product. When the sample is heated to room temperature the

3

unexposed OIR sublimates while the exposed patterns are stable, enabling ambient downstream processing.

4

We carefully designed and tested OIRs condensation conditions and e-beam scan parameters. The OIR

5

thickness is controlled by adjusting the GIS leak valve and the total deposition time. We selected and

6

patterned OIRs using a representative subset of important organic chemistry building blocks: increasing

7

hydrocarbon chain length, presence of aromatic rings, and of hydroxyl groups (Figure 2a; see also the

8

comparative Table S1). 1-pentanol (C5H12O) had the lowest source vapour pressure in the GIS, and

9

therefore it provides the lowest deposition rates enabling uniform ultrathin layers (7 nm). This is

10

challenging with resin-based EBL resists. Anisole (methoxybenzene, C7H8O) and nonane (C9H20) ice resists

11

readily formed 20 to 200-nm-thick layers, which made them ideal for lithography applications. The layers

12

were condensed at substrate temperatures ranging from 120 K to 150 K without any observable difference

13

in subsequent performances, indicating a robust process. We also patterned isopropanol (C3H8O) ice, but

14

the results were hard to reproduce, possibly due to its lowest freezing point among the four compounds

15

tested. The critical doses (D80, Figure 2b) of nonane ice for 5 and 20 keV e-beam exposures were reached at

16

3 and 12 mC cm-2, respectively. The analysis of exposed thick layers revealed that OIR resist thickness does

17

not affect the critical dose, as even at 5 keV the electrons interact with the ice layer throughout its entire

18

thickness at each scan point. This enables the patterning of OIR layers ranging from a few to hundreds of

19

nanometer thick using identical patterning time, and is in direct contrast to the gas-based FEBID method,

20

which deposits matter in a monolayer-by-monolayer fashion, leading to a process time proportional to the

21

volume of the deposited material. The OIR critical dose is already 100-fold lower than in IL, allowing

22

nanoscale fabrication as well as sub-mm features (Figure 2c–d) which would be prohibitively time

23

consuming using water ice. The OIR dose is significantly higher than PMMA (250 µC/cm2 at 19 keV), and

24

other advanced EBL resists (10 – 100 µC/cm2)13, and it is difficult to compare to FEBID doses, as these

25

depend on precursor gas flow rate, nozzle and sample geometry, and writing direction14 as well as desired

4 ACS Paragon Plus Environment

Page 5 of 13

Nano Letters

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

1

thickness of the deposits. Nevertheless, a benchmark figure for the FEBID dose required to deposit a single

2

atomic monolayer from an organometallic precursor translates to 3 mC cm-2 14, which is equivalent to the

3

dose we used to pattern the 600-nm-thick OIR layer in Figure 2b. Hence, to deposit layers above 100-nm-

4

thick, as those used for etching mask applications, OIR is up to 3 orders of magnitude more efficient than

5

FEBID.

6

To investigate the resolution of OIR, we patterned and compared single line scans on OIR thin-films (Figure

7

2e–g) and on the negative-tone EBL resist AR-N 7520. We obtained down to 60-nm-wide OIR lines, and 80-

8

nm-wide EBL resist lines. The unexpectedly large EBL resist patterns indicate that feature size is currently

9

limited by our custom instrument.

10 11

Figure 2. OIR patterns. (a) Atomic force microscopy (AFM) images of 20 x 12 µm2 rectangles patterned in OIR thin-films and

12

representative scan-lines across the rectangle surface. Average pattern thickness is in parenthesis. (b) Contrast curves for nonane

13

ice at 5 and 20 keV. The critical dose at 5 keV is unchanged even with a 20-fold time increase in thickness. (c, d) Large area map of

14

Denmark made by exposed nonane ice (dark blue areas). The total exposure time was 75 min. The optical image (c) covers an area

15

of 0.4 x 0.5 mm2. The micrometric DTU logo (9 x 6 µm2) patterned in the Copenhagen region is visible in the SEM inset (d). (e) SEM

5 ACS Paragon Plus Environment

Nano Letters

Page 6 of 13

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

1

image of 100-nm-wide lines patterned in nonane ice. (f) 60-nm-wide line patterned in anisole ice. (g) Tilted SEM view of patterned

2

OIR lines reaching the edge of a 1 x 1 cm2 chip. The OIR layer is uniform on the entire sample surface, up to the very edge. Scale bar

3

is 200 nm.

4

To further probe the resolution performances of OIRs, we used a transmission electron microscope (TEM)

5

fitted with a gas cell and a cryostage to condense OIR layers on electron-transparent membranes and

6

perform analogous lithography experiments operating the TEM at 80 keV. First, we used our custom

7

instrument to test whether OIR is compatible with the delicate 3-mm-wide TEM substrates consisting of 5-

8

nm and 50-nm-thick silicon nitride membranes with holes. We deposited and patterned OIR layers on the

9

membranes (Figure 3a; see also Figure S1) without any modification to the setup, while spin-coating EBL

10

resist on such samples would have required custom holders and careful handling or, for more irregular

11

surfaces, alternative strategies such as thermal evaporation of the resist15, which however relies on

12

additional high temperature processes and still requires development in liquid solvents. Liquid

13

development is undesired because the interfacial forces involved in the drying process may damage fragile

14

nanostructures. To recreate the experimental conditions of OIR lithography in the TEM, we introduced

15

nonane vapour through the GIS connected to the gas cell, condensing OIR thin-films on cold 5-nm-thick

16

silicon nitride membranes. We focused the TEM e-beam to scan and expose patterns in the form of dots

17

arrays. After exposure, we heated the cryostage in-situ and immediately observed the resulting structures

18

(Figure 3b). The significantly improved feature size (down to 10 nm, Figure 3c) confirms that our SEM

19

results are far from the ultimate resolution of this technique, and suggests that by improving the optics and

20

stability of our custom instrument we might achieve with OIRs critical dimensions in line with EBL.

6 ACS Paragon Plus Environment

Page 7 of 13

Nano Letters

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

1 2

2

Figure 3. OIR patterning in an environmental TEM. (a) TEM bright-field image of a 20x20 µm square and lines previously patterned

3

in our custom instrument on a 20-nm-thick “holey” Si3N4 membrane. (b, c) TEM dark-field (b) and bright-field (c) images of exposed

4

nonane OIR dots patterned on 5-nm-thick Si3N4 membranes. In this case, OIR condensation, exposure, sublimation and imaging

5

were all performed inside the TEM using its 80 keV beam. Scale bar in (b) is 500 nm, the highlighted dots in (c) are 18 and 10 nm in

6

diameter.

7

Another major advantage of OIR is 3D patterning capabilities, which we demonstrate with multi-layered

8

structures realized in our custom instrument using an iterative condense-exposure process (Figure 4a; see

9

also Supporting Movie 1 for a 3D animation). All three OIR layers are uniform within sub-nanometer surface

10

roughness and unaffected by the stacking, which suggests that many layers could be added for more

11

complicated and higher aspect ratio structures. We then modified this 3D lithography process using

12

condense-expose-sublimate cycles. Using the first 3D process, the top layers present the cumulative

13

thickness as in an additive manufacturing process. In the modified process, any additional top layer follows

14

the profile of the previously exposed features underneath (Figure 4b) and allows multiple patterning, which

15

is an important technique used to increase pattern density16. For these configurations, OIR processing is

16

more straightforward than EBL, because all steps can be performed in-situ while EBL requires multiple

17

resist spin, bake and development steps. Furthermore, EBL resists such as PMMA cannot be stacked in the

18

first place, because the solvents for spin-coating damage the underlying features. Additive 3D routines in

19

low temperature FEBID were introduced by Bresin et al.17, who reported on optimizing growth

20

performances for direct writing of platinum structures up to several micrometers thick. The deposits are

7 ACS Paragon Plus Environment

Nano Letters

Page 8 of 13

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

1

platinum clusters embedded in an amorphous carbon matrix, and the method appears powerful for direct

2

writing of metallic nanostructures. However, platinum deposits are not suitable for nanoscale lithography

3

applications since, unlike organic resists, platinum cannot be removed by chemical or physical processing

4

without relying on aggressive chemistries.

5 6

Figure 4. Three-dimensional multi-layered structures. (a) AFM 3D image and its central scan line for a multi-layered structure

7

obtained by patterning 3 layers of anisole ice. (b) OIR-enabled approaches to multilayer patterning, and AFM images of the results.

8

In the AFM images, L1 originates from exposing the first ice layer, and L2 from the second. Scale bar is 1 µm.

9

To create useful devices, lithographic patterns must be transferred into an underlying functional material.

10

The lithography patterns act as masks during pattern transfer, and they must be chemically stable. Indeed,

11

OIR patterns were stable when placed in different organic solvents, and mild acids and bases. We used OIR

12

patterns as protective etch masks in reactive ion etching to create silicon nanostructures (Figure 5a-b, see

8 ACS Paragon Plus Environment

Page 9 of 13

Nano Letters

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

1

Supporting Information for details of the etch process). A good etch mask must not be consumed during

2

etching of the underlying material, with their relative etch rate defined as the selectivity. For nonane ice

3

patterns, selectivity over silicon was 1:6, which is the same value obtained with AR-N 7520. Using oxygen

4

plasma, we can remove the patterned OIR without leaving any residues or harming the underlying silicon.

5

This ensures a pristine surface for further processing or final applications. The good chemical stability, high

6

selectivity, and the fact that they can be removed by oxygen plasma, suggest that these features might be

7

amorphous hydrogenated carbonaceous solids akin to FEBID deposits18-19. EDX and EELS analysis of the

8

exposed OIR on the TEM grids showed carbon and oxygen (Figure S2), but both methods are not directly

9

sensitive to the expected hydrogen content.

10

One main advantage of OIR is that we can handle small samples such as the TEM membrane samples used

11

in our TEM experiments. As a further demonstration, we patterned OIR onto diamond chips as small as 2 x

12

2 mm2 to be used as etch mask (Figure 5c-e). Diamond substrates and nanopillars are e.g. used for

13

nitrogen-vacancy centres for quantum technology applications20. In contrast to EBL resist coating, uniform

14

OIR layers of well-controlled thickness can be deposited on the entire sample surface, and the one-step

15

lithography reduces sample manipulation to a minimum.

9 ACS Paragon Plus Environment

Nano Letters

Page 10 of 13

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

1 2

Figure 5. Fabrication of nanowires and nanopillars by plasma etching. (a) AFM profiles evolution of OIR lines on a silicon substrate

3

at three different steps in the etch process: as patterned, after silicon etch, and after removal of the residual OIR. The etch

4

selectivity between the patterned OIR and silicon is 1:6. (b) SEM view of 400-nm-tall silicon fins made with OIR and reactive ion

5

etching. Scale bar is 500 nm. (c) Tweezers holding a 2 x 2 mm diamond chip used for fabricating diamond quantum devices. (d)

6

SEM view of an array (pitch 10 µm) of diamond nanopillars fabricated using patterned OIR as the etch mask. (e) Close-up of

7

diamond pillars. Scale bar is 1 µm.

8

In conclusion, OIR offers unique advantages to nanofabrication beyond the state-of-the-art. These ice

9

layers of simple organic molecules can be used as negative-tone resists in e-beam based patterning without

10

requiring spin-coating or developing in liquid solutions. Both larger sub-mm- and nanostructures can be

11

created by condensing these molecules on a cooled substrate and scanning the e-beam according to the

12

desired pattern. The unexposed organic ice sublimates away, leaving no residues, while the exposed areas

13

are carbonaceous structures stable at ambient conditions, which, if needed, can be removed in a gentle

2

10 ACS Paragon Plus Environment

Page 11 of 13

Nano Letters

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

1

oxygen plasma step. After the lithography process, the initial substrate remains pristine and undamaged.

2

Similar to additive manufacturing technologies, we also made 3D structures by stacking multiple layers.

3

While we used homogeneous OIR stacks in these experiments, we believe there is unexplored potential in

4

combining chemistries and advanced 3D structures consisting of heterogeneous multilayer stacks. Still to be

5

discovered are e.g.; organic compounds providing desirable electrical properties for the resulting features;

6

additional interlayer dynamics (e.g. chemical reactions, sacrificial layers) which could further increase the

7

complexity of 3D structures and the range of potential applications.

8

We have only started to explore the potential of OIR for scientific research. To identify the most suitable

9

molecules and processes for future applications and challenges, we are now focusing our research efforts

10

on achieving in-depth scientific understanding of electron-OIR interactions.

11

Supporting Information

12

Supporting information is available in the online version of this paper:

13

Experimental details regarding sample preparation, experimental setup for cryogenic SEM and TEM lithography and

14

imaging, chemical stability and pattern transfer, diamond nanofabrication and further characterization; comparison of

15

different precursors; TEM analysis of OIR patterns after fabrication.

16

Animation video showing an example of iterative 3D lithography process, used to create the pattern in Figure 4a.

17

Author information

18

Corresponding author

19

Email: [email protected]

20

Author contributions

21

Experimental work on OIR patterning and characterization was carried by W. T., as conceived and discussed together with A. H. and

22

M. B.; A. E. and P. L. performed in-situ TEM experiments and other TEM characterization. H. T. L. performed the work on OIR

23

chemical stability, dry etching and characterization of the etched features. W. T. drafted the manuscript; all authors contributed to

24

revisions and comments and discussed the results.

25

Notes

26

The authors declare no competing financial interests.

11 ACS Paragon Plus Environment

Nano Letters

Page 12 of 13

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

1

Acknowledgments

2

The authors would like to acknowledge J. E. Jørgensen, M. Nimb, T. Feld and S. M. B. Petersen for providing custom parts in the

3

experimental setup. The authors thank for their contribution, assistance and valuable inputs A. Gregersen (for the electronics

4

setup), R. Cork (vacuum setup), J. Michael-Lindhard (SEM instrument), K. H. Rasmussen (diamond etching). We are also grateful to

5

D. Branton for helpful comments. We also acknowledge financial support from the VILLUM foundation.

6

References

7

1.

Jahani, S.; Jacob Z. Nature Nanotech., 2016, 11, 23–36.

8

2.

Aspelmeyer, A.; Kippenberg, T. J.; Marquardt F. Rev. Mod. Phys., 2014, 86, 1391.

9

3.

Sparreboom, W.; van den Berg, A.; Eijkel, J. C. T. Nature Nanotech., 2009, 4, 713–720.

10

4.

Ladd, T. D.; Jelezko, F.; Laflamme, R.; Nakamura, Y.; Monroe, C.; O’Brien J. L. Nature, 2010, 464, 45–53.

11

5.

Geim, A. K.; Grigorieva. I. V. Nature, 2013, 499, 419–425.

12

6.

Huth, M.; Porrati, F.; Schwalb, C.; Winhold, M.; Sachser, R.; Dukic, M.; Adams, J.; Fantner, G. Beilstein J. Nanotech., 2012, 3,

13

597–619.

14

7.

Manfrinato, V. R.; Zhang, L.; Su, D.; Duan, H.; Hobbs, R. G.; Stach, E. A.; Berggren, K. K. Nano Lett., 2013, 13, 1555–1558.

15

8.

Van Oven, J. C.; Berwald, F.; Berggren, K. K.; Kruit, P.; Hagen, C. W. J. Vac. Sci. Technol. B, 2011, 29, 06F305.

16

9.

Esposito, M.; Tasco, V.; Cuscunà, M.; Todisco, F.; Benedetti, A.; Tarantini, I.; De Giorgi, M.; Sanvitto, D.; Passaseo, A. ACS

17

Photonics, 2015, 2, 105–114.

18

10. King, G. M.; Schürmann, G.; Branton, D.; Golovchenko, J. A. Nano Lett., 2005, 5, 1157–1160.

19

11. Han, A.; Kuan, A.; Golovchenko, J. A.; Branton, D. Nano Lett., 2012, 12, 1018–1021.

20

12. Tiddi, W.; Elsukova, A.; Beleggia, M.; Han, A. Micro Electronics Engineering, Submitted.

21

13. Gangnaik, A. S.; Georgiev, Y. M.; Holmes, J. D. Chem. Mater., 2017, 29, 1898–1917.

22

14. Hagen, C. W. Appl. Phys. A, 2014, 117, 1599–1605.

23

15. Zhang, J.; Con, C.; Cui, B. ACS Nano, 2014, 8, 3483–3489.

24

16. Chao, W.; Kim, J.; Rekawa, S.; Fischer, P.; Anderson, E. J. Vac. Sci. Technol. B, 2009, 27, 2606–2611.

25

17. Bresin, M.; Toth, M.; Dunn, K. A. Nanotechnology, 2013, 24, 035301.

26

18. Bret, T.; Mauron, S.; Utke, I.; Hoffmann, P. Microelectron. Eng., 2005, 78, 300–306.

27

19. Sangiao, S.; Martín, S.; Gónzalez-Orive, A.; Magén, C.; Low, P. J.; De Teresa, J. M.; Cea, P. Small, 2017, 13, 1603207.

28

20. Babinec, T. M.; Hausmann, B. J. M.; Khan, M.; Zhang, Y.; Maze, J. R.; Hemmer, P. R.; Lončar, M. Nature Nanotech., 2010, 5,

29

195–199.

12 ACS Paragon Plus Environment

Page 13 of 13 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

Nano Letters

82x39mm (220 x 220 DPI)

ACS Paragon Plus Environment