Nanoscale junctions for simultaneous electronic and optical measurements of single molecules
Sponsored Links
This presentation is the property of its rightful owner.
1 / 52

Nanoscale junctions for simultaneous electronic and optical measurements of single molecules PowerPoint PPT Presentation


  • 109 Views
  • Uploaded on
  • Presentation posted in: General

Douglas Natelson Department of Physics & Astronomy and Electrical & Computer Engineering. Nanoscale junctions for simultaneous electronic and optical measurements of single molecules. Atomic and molecular scale junctions. Organic semiconductor devices.

Download Presentation

Nanoscale junctions for simultaneous electronic and optical measurements of single molecules

An Image/Link below is provided (as is) to download presentation

Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author.While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server.


- - - - - - - - - - - - - - - - - - - - - - - - - - E N D - - - - - - - - - - - - - - - - - - - - - - - - - -

Presentation Transcript


Douglas Natelson

Department of Physics & Astronomy and Electrical & Computer Engineering

Nanoscale junctions for simultaneous electronic and optical measurements of single molecules

  • Atomic and molecular scale junctions

  • Organic semiconductor devices

  • Strongly correlated nanoelectronics

JSC, 5/18/10


Single-molecule electronic properties

Electron transfer (chemistry)

Conduction (physics)

Many techniques and measurements over the last decade.

Cui et al., Science 294, 571 (2001).

Dadosh et al., Nature 436, 677 (2005).

Liang et al., Nature 417, 725 (2002).

Champagne et al., NL 5, 305 (2005).

Why is this interesting?

  • Transport through discrete quantum states

  • Fundamental physical chemistry problems, novel spectroscopies, potential sensor applications

  • Tunable model systems for strongly correlated materials

  • Relevant to any molecular-scale electronics.


  • “Hot” electrons in the metal leads.

  • Electronic transitions in the molecule

  • Chemical reactions

Quantum transport: A complicated, nonequilibrium problem!

Apply bias….

  • Molecular vibrations.

  • Vibrations in the electrodes.

  • Quantum entanglement.


Quantum transport: A complicated, nonequilibrium problem!

  • “Hot” electrons in the metal leads.

  • Electronic transitions in the molecule

  • Chemical reactions

Apply bias….

  • Molecular vibrations.

  • Vibrations in the electrodes.

  • Quantum entanglement.


Quantum transport: A complicated, nonequilibrium problem!

Apply bias….

Big questions:

  • How important are electron-electron interactions and quantum coherence effects?

  • How does dissipation work at these scales?


Nonequilibrium, beyond conductance: “shot” noise

Conductance: tells us average current under certain voltage bias.

If charge was continuous, that would be the end of the story.

However, charge comes in discrete lumps….

Theorist fantasy: ordered list of arrival times for each electron.

16:07:23.0000315

16:07:23.0000319

16:07:23.0000371

16:07:23.0000389

16:07:23.0000400

16:07:23.0000422

16:07:23.0000430

16:07:23.0000463

.

.

.

Now we can compute á I ñ, as before, as well as á (I – á I ñ )2ñ (within some bandwidth)


Shot noise

http://www.geocities.com/bioelectrochemistry/schottky.htm

Classical: Schottky (1918)

Noninteracting electrons

Arrivals as Poisson process.

What if e- arrivals are not independent? More generally:

F º Fano factor


Shot noise

SI

F ~ 2

F ~1

4kBT

F ~0

R

I

  • Fano factor tells you about correlations between electron arrivals.

  • e.g., F = 2 for Poissonian arrival of pairs, as in SC tunnel junction.

  • In this sense, F tells you about effective charge of excitations.

  • F → 0 in macroscopic systems at moderate temperatures – inelastic scattering effectively smears out the conductance channels.


Shot noise – quantum case

ti = (Quantum) transmission probability through junction

T = 0

T¹ 0

If conductance quantization really comes from Landauer physics, expect suppression of noise whenever ti~ 1.


Shot noise suppression

Painful, laborious, low freq. measurements, 4.2 K.

Shot noise suppression is observed!

van den Brom and van Ruitenbeek, PRL 82, 1526 (1999)


Room temperature shot noise measurements

  • Break junction setup (Nick King, Jeff Russom senior theses)

  • High-frequency detection method – fast!

  • Lock-in approach: only look at V-driven noise.


(a)

sync

sync

(b)


Room temperature shot noise measurements

  • Au contacts, room temperature, acquisition time = hours, V ~ 100 mV

  • Noise suppression clearly observed!

  • Tells us that inelastic suppression still not taking place.


How we make single-molecule devices

Electromigration technique

Park et al., APL75, 301 (1999)

current pulse

  • Analogous to STM:

  • Conduction dominated by tunneling volume ~ 1 molecule.

  • Every device is different!

  • Can’t “see” what’s going on!

  • Vibrational fingerprint?


Vibrational resonances

Drain

Source

35 meV

Vibrational effects

One molecule-specific feature:

E


Vibrational resonances

Drain

Source

35 meV

Vibrational effects

One molecule-specific feature:

E

C60 device, 4.2 K

Ward et al., J. Phys: Cond. Matt. 20, 374118 (2008)


Vibrational effects

Vibrational resonances

Drain

Source

One molecule-specific feature:

E

Qiu et al., PRL 92, 206102 (2004)


Raman spectroscopy

E1

  • Inelastic light scattering.

  • Requires a = a(r).

  • Time-varying r from vibrations (w0) leads to sum + difference frequency generation with incident light (w).

E0

Stokes

anti-Stokes

Rayleigh

  • Cross sections typically 10-29 cm2 - very small!

  • Stokes/anti-Stokes ratio can tell you temperature due to Boltzmann occupancy factor.


Plasmons and surface-enhancement

Plasmons = sound waves in the electron fluid.

metal nanoparticle


Plasmons and surface-enhancement

Plasmons = sound waves in the electron fluid.

metal nanoparticle


Plasmons and surface-enhancement

Plasmons = sound waves in the electron fluid.

metal nanoparticle


Plasmons and surface-enhancement

Plasmons = sound waves in the electron fluid.

metal nanoparticle


Plasmons and surface-enhancement

Plasmons = sound waves in the electron fluid.

metal nanoparticle dimer


Plasmons and surface-enhancement

Plasmons = sound waves in the electron fluid.

metal nanoparticle dimer

  • Local electric field can be much larger (g(w)) than incident field!

  • Raman scattering rate ~ g(w)2g(w’)2

  • If g(w), g(w’) ~ 1000 each, then Raman enhanced by 1012.

  • Single-molecule sensitivity possible in surface-enhanced Raman spectroscopy (SERS).


1 µm

Vibrational Spectroscopy

50 µm

Nanoscale gaps are ideal for surface-enhanced Raman spectroscopy.


Vibrational Spectroscopy

Si

  • EM calculations (FDTD) show that nanometer-scale protrusions can lead readily to SERS enhancements approaching 1012 (!).


Vibrational Spectroscopy


Vibrational Spectroscopy

  • At nanogap, large SERS signal, “blinking”, and spectral diffusion.

  • Simultaneous transport + Raman would open many possibilities.


Vibrational Spectroscopy

A

B

C

1 µm

D

E

Remember, data so far taken in air, at room temperature.


Transport + SERS

  • Enhancement “turns on” as junction is migrated.

  • Inter-electrode plasmon modes form once conductance ~ 10-4 S.


Transport + SERS

  • Raman and transport correlate very strongly in time.

  • Demonstrates single-molecule Raman sensitivity.

  • Demonstrates multifunctional sensing at single-molecule level.


Simultaneous conduction + SERS

D. Ward et al., J. Phys. Condens. Matt. 20, 374118 (2008).


Stokes-AntiStokes and local temperature

  • Ratio of antiStokes to Stokes intensities provides a measure of excited state population, and therefore effective temperature.

  • The exponential makes things challenging.

Ignoring cross-section and enhancement issues,

450 cm-1 mode at 80 K → AS/S = 3.8 × 10-4.

So, 10000 Stokes counts → 3.8 antiStokes counts

  • SERS itself can lead to optical pumping of excited state.

Can we see bias-driven effects?


1625 1/cm

1317 1/cm


50 mV

A

Upper bound on temperature at A is ~220 K


500 mV

D

B

F?

C

A

E

D

E

F

B

A

C

A=560 K

B=520 K

C=508 K

D=513 K

E=470 K

F=382 K


Missing points are below detection threshold


Conclusions

  • We can make atomic- and molecular-scale structures and measure their electronic and optical properties.

  • We can see quantum effects in electronic conduction at room temperature in these gadgets.

  • We can make nanoscale optical antennas that allow chemical sensing and vibrational pumping at the single-molecule level.

  • Basic science with potential applications in sensing, photonics, cutting-edge nanoelectronics.

Very exciting! We have a lot to do….


Acknowledgments

My group: Daniel Ward, Jeff Worne, Alexandra Fursina, Patrick Wheeler, Kenneth Evans, Amanda Whaley,Ruoyu Chen, HengJi, Dr. Gavin Scott, Dr. Jiang Wei

Collaborators: J.M. Tour, N.J. Halas, P. Nordlander, J. C. Cuevas

D. Ward et al., NanoLett. 7, 1396-1400 (2007).

D. Ward et al., NanoLett. 8, 919-924 (2008).

D. Ward et al., J. Phys. Condens. Mat.20, 374118 (2008).

P. J. Wheeler et al.,NanoLett. 10, 1287-1292(2010).


Large aspect ratio nanogaps by self-alignment

II. Oxidation

I. E-beam lithography First electrode

CrxOy

Cr

IV. Etching away Cr/Cr2O3 layer

Room temperature, 90 sec, Cr-7 etchant

III. Second electrode

A. Fursina et al., Appl. Phys. Lett. 93, 113102 (2008)


Uniquenanofab capabilities

Industrially scalable nanogap fabrication method

20 μm


Optical rectification

Tunneling nonlinearity can lead to DC current from AC bias.

Light induces some Vopt across our gap, oscillating at ~ 1015 Hz.

Rectification (photocurrent) can lead to a means of quantitatively estimating the field enhancement factor! (provided that tunneling is fast!)


Optical rectification

This has been shown to work quantitatively at microwave frequencies.


Validity of classical rectification

In general, correct quantum treatment = photon-assisted tunneling.

Tien-Gordon (perturbative) approach:

First order in alpha:

If a << 1 and nonlinearity varies little (i.e., DOS is “boring”) over

then classical rectification picture is reasonable.

J. C. Cuevas


Validity of classical rectification

Viljas and Cuevas, PRB 75, 075406 (2007)

DFT calculations originally done for Au contacts show that we are likely in luck as far as DOS goes, as long as junctions are clean.


Optical rectification

How would this work?

  • Use low freq measurement (wlow = 2p´ 2 kHz) with known Vac , and use lock-in at 2wlow to measure (d2I/dV2) as a function of Vdc.

  • Simultaneously, measure the photocurrent as a function of Vdc (use a second lock-in and chop the light).

  • Adjust Vac until the two signals (2nd harmonic + photocurrent) are identical. Voila – Vac should now = Vopt.

d2I/dV2

I

V

V


Actual experiment:

  • Apply dc + small ac voltage, V0.

  • Simultaneously measure I, dI/dV, (1/4)d2I/dV2Vo2, and Iphoto as a function of Vdc.


Challenges:

  • Junction stability when illuminated is comparatively worse.

  • Capacitance in cryostat wiring makes low freq measurement not as sensitive as we would prefer.

  • Issues w/ oxide under large bond pads (two distinct problems: mismatch of conductance and d2I/dV2; and/or no spatial dependence of photocurrent).


Cross-checks


Examples:

dI/dV in units of nA/V.

Quantitative agreement between Iphotoand (1/4)d2I/dV2 V02 happens when V0 = 30 mV.

This implies Vac from the optical field is ~ 30 mV!

Measured dI/dV at Vdc = 0 implies an interelectrode gap of ~ 0.14 nm, implying local field = 2.1 x 108 V/m.

Incident intensity = 22.6 kW/cm2

Implied field enhancement ~ 718x


Examples:

dI/dV in units of nA/V.

Quantitative agreement between Iphotoand (1/4)d2I/dV2 V02 happens when V0 = 25 mV.

Measured dI/dV at Vdc = 0 implies an interelectrode gap of ~ 0.044 nm, implying local field = 5.7 x 108 V/m.

Implied field enhancement ~ 1940x (!!)


Examples:

dI/dV in units of nA/V.

Quantitative agreement between Iphotoand (1/4)d2I/dV2 V02 happens when V0 = 32.4 mV.

Measured dI/dV at Vdc = 0 implies an interelectrode gap of ~ 0.092 nm, implying local field = 3.6 x 108 V/m.

Implied field enhancement ~ 1230x


  • Login