Charged particle single nanometre manufacturing

Abstract

Following a brief historical summary of the way in which electron beam lithography developed out of the scanning electron microscope, three state-of-the-art charged-particle beam nanopatterning technologies are considered. All three have been the subject of a recently completed European Union Project entitled "Single Nanometre Manufacturing: Beyond CMOS". Scanning helium ion beam lithography has the advantages of virtually zero proximity effect, nanoscale patterning capability and high sensitivity in combination with a novel fullerene resist based on the sub-nanometre C₆₀ molecule. The shot noise-limited minimum linewidth achieved to date is 6 nm. The second technology, focused electron induced processing (FEBIP), uses a nozzle-dispensed precursor gas either to etch or to deposit patterns on the nanometre scale without the need for resist. The process has potential for high throughput enhancement using multiple electron beams and a system employing up to 196 beams is under development based on a commercial SEM platform. Among its potential applications is the manufacture of templates for nanoimprint lithography, NIL. This is also a target application for the third and final charged particle technology, viz. field emission electron scanning probe lithography, FE-eSPL. This has been developed out of scanning tunneling microscopy using lower-energy electrons (tens of electronvolts rather than the tens of kiloelectronvolts of the other techniques). It has the considerable advantage of being employed without the need for a vacuum system, in ambient air and is capable of sub-10 nm patterning using either developable resists or a self-developing mode applicable for many polymeric resists, which is preferred. Like FEBIP it is potentially capable of massive parallelization for applications requiring high throughput.

Keywords: charged particle beams; electron; field emission; ion; nanolithography.

Figure 1

(a) SEM image of ALIS gas field ion source, produced with permission of [24], copyright 2010 Japanese Journal of Applied Physics. (b) Principle of operation of ALIS. (c) W atom trimer emitter at tip providing single atom emitter (ion microscope image obtained by Dr S. A. Boden at Southampton University).

Figure 2

(a) ORION Plus He Ion Microscope from Zeiss AG, located at University of Southampton. (b) Schematic of He⁺ ion focusing column.

Figure 3

Electron and ion beam substrate penetration – Monte Carlo simulations of charged particle paths showing forward and back scattering. Reproduced with permission through Creative Commons Attribution (CC BY) from [30], 2007 AIP Conference Series.

Figure 4

(a) Spin-coating results for fullerene resists HM-01A and HM-01C using anisole and chlorobenzene solvents, respectively. (b) Schematic of the monoadduct methanofullerene molecule used in the HM resist series [40].

Figure 5

Isolated dose-optimised SHIBL experiments on HM01 fullerene resist: 8 nm wide sparse exposed using 30 keV He⁺ ions [41].

Figure 6

Results of SHIBL at 30 keV beam energy: (a) optical micrograph, (b) AFM image, (c) line profile (HM-01 fullerene resist thickness ca. 10 nm). He⁺ doses left to right: 37.9, 56.5, 85.0, 128.0, 192.0 μC·cm⁻², (d) Comparison of the dose response curves for HM-01 in SHIBL and EBL. Sensitivities are 40 µC·cm⁻² and 20 mC·cm⁻², respectively, revealing a 500-fold increased sensitivity in SHIBL [41].

Figure 7

HIM image of dense (1:1) single-pixel features exposed at (a) 0.09 nC·cm⁻¹ and (b) at 0.04 nC·cm⁻¹ in ca. 10 nm thick HM-01A fullerene resist, with line scans of secondary electron intensity providing measurements of linewidth [41].

Figure 8

(a) AFM and (b) corresponding HIM, (c) AFM and (d) corresponding SEM images of doughnuts fabricated using SHIBL and EBL with fixed outer radii (R₂) of 200 nm and 7 μm, respectively, and varied inner radii (R₁). (e) Comparison of the proximity effect for SHIBL and EBL on 20 nm thick PMMA [42].

Figure 9

(a) Schematic of the multibeam EBL system at TU Delft [–46]. (b) Experimental multibeam tool using a Nova Nano SEM from FEI Co as platform.

Figure 10

Schematic of electron beam induced deposition (EBID).

Figure 11

Schematic representation of the evolution of high-resolution electron beam induced deposition as a nanopatterning technique since its first demonstration in 1976. (Clockwise order) Schematic of EBID, 8 nm lines patterned by Broers et al. on a thin membrane [55]; simulation of dot grown on a thin membrane using a zero-diameter electron beam demonstrating the relevance of deposit-generated SEs in EBID [70]; experimental demonstration of dots with average diameter of ca. 1 nm patterned on a thin membrane by EBID [64]; direct patterning of 3 nm dense lines on bulk Si/SiO₂ by EBID [65]. Reprinted with permission from [55] and [70], copyright 2003, 2013 AIP Publishing; reprinted with permission from [65], copyright 2011 American Vacuum Society.

Figure 12

Carbonaceous nanowires on bulk SiO₂ such as the one in (a) slimmed by electron beam induced etching in the presence of water in an ESEM resulting in wires of different widths (b–e). Adapted with permission from [94], copyright 2007 American Chemical Society.

Figure 13

Left: SEM micrograph of an EBID mask consisting of 17 nm lines at 50 nm spacing, transferred into the silicon substrate using fluorine etch. Right: AFM profile showing a height ratio before and after etching of 8 [105].

Figure 14

(a) Thermographic scale representation of the electric field between tip and sample calculated by solving the Laplace equation for a voltage of 50 V, a tip–sample distance of 10 nm and a tip diameter of 17 nm. (b) Schematic description of the fundamentals of electron tunneling into air from the tip of a scanning probe.

Figure 15

(a) SEM image of self-sensing and self-actuated cantilever with the thermomechanical actuator and piezoresistive deflection read-out. (b) Close-up SEM image of the tip at the front of the cantilever. Reproduced with permission through Creative Commons Attribution (CC BY) from [146], 2017 AIP Conference Series.

Figure 16

(a) Schematic showing feedback loops combined in FE-eSPL tool enabling lithography using the current feedback loop and measurement of written structures by AFM using the force feedback loop. Reprinted with permission from [157], copyright 2014 American Vacuum Society. (b) FE-eSPL tool developed in the Rangelow Group at TU Ilmenau. Reprinted with permission from [158], copyright 2018 Elsevier.

Figure 17

Example of stitching test showing an AFM image obtained directly after FE-eSPL exposure, showing negative tone features. The same pattern was written four times, indicated by the green lines, to compose the complete pattern. Stitching was achieved using scanner positioning. Before patterning the previously written features were measured with AFM and the new pattern was aligned to these structures, using an overlap of 0.5 µm with previous ones (indicated by the blue dashed lines).

Figure 18

(a) Results of typical exposure dose test for determination of lithographic tone as a function of the exposure dose. AFM image of standard exposure dose test on resist AZ^® BARLi^® acquired directly after FE-eSPL. The dose is varied by changing the writing speed from 1 to 2.5 μm·s⁻¹ (marked with red arrow) and by changing the current from 15 to 30 pA (blue arrow). (b–d) Close-up AFM image on different lithographic tones obtained after etching. (b) Negative tone feature, (c) intermittent tone, which is a combination of positive tone (trenches in the middle of the line) and negative tone (hills at the border of the lines), (d) positive tone features.

Figure 19

AFM images of FE-eSPL patterning of different materials. (a) MoS₂-flake on SiO₂ substrate placed between gold contacts before patterning. (b) MoS₂ flake after FE-eSPL patterning and first development step. (c) Patterning demonstration of novel molecular glass resist UBT7 from University of Bayreuth. (d) Patterning demonstration of HM01 fullerene resist from Irresistible Materials Ltd.

Figure 20

Single-digit nanometre features written by FE-eSPL. (a) SEM image obtained after etching of 9 nm thick AZ^® BARLi^® II resist (AZ Electronic Materials) showing 12.5 nm half pitch structures (upper part), (b) SEM image obtained after etching of 9 nm thick AZ^® BARLi^® II resist showing 5–6 nm thick lines, (c) AFM image after FE-eSPL exposure of P3HT showing regular 25 nm dot structures, (d) AFM image and (e) AFM profile obtained directly after FE-eSPL showing 7–10 nm (FWHM) dots in AZ^® BARLi^® II resist.

Figure 21

(a, b) SEM images of volcano-gated tip. (c, d) Simulation of the electron beam/trajectories for volcano-gated tips with different volcano heights showing the (c) defocusing and (d) focusing effects. Reproduced with permission from [175], copyright 2017 Elsevier.

Figure 22

(a) Electron trajectories for 30 keV EBL exposure of 100 nm calixarene film on a Si sample. (b) Electron trajectories for FE-eSPL with 50V tip bias, 10 nm tip–resist distance and 10 nm thick calixarene resist (note that only one half of the tip and simulation area is shown). (c) Deposited energy distribution (EDD), describing the energy transferred from the electrons into the resist due to inelastic scattering, shown for various depths inside the resist for 30 keV EBL exposure of 100 nm calixarene film on a Si sample, for calculation details see [177]. (d) Energy loss distribution (ELD), describing the energy lost by the electrons due to inelastic scattering events, for 50 eV FE-eSPL exposure of 10 nm thick calixarene resist (note that the calculation differs from the EDD, thus giving different absolute values). (a, c) Images reproduced with permission through Creative Commons Attribution (CC BY) from [177], 2013 SciencePG; (b, d) images reproduced with permission from [174].

Figure 23

SEM image of Quattro cantilever array used for parallel AFM imaging with four cantilevers.