Biomolecule-Material Interface at the Nanoscale: atomic structure, electronic properties, and energy applications Sheng Meng Department of Physics and.

Biomolecule-Material Interface at the Nanoscale:atomic structure, electronic properties, and energy applications

Sheng Meng

Department of Physics and School of Engineering and Applied Sciences,

Harvard University

Colloquium Department of Physics and Astronomy, University of Mississippi

Jan. 31, 2008

Biology is naturally nanoscale

Spinach aquaporin

We need tools to manipulate …

• Substrates for investigating biomolecules• Biosensor for recognition and diagnosis • Implants for medical operation & recovery• Drug delivery• Building up bio-chips• Bio-nanotechnology

A hybrid bio-nano machine?

Schwegler, LLNL

Bring Materials to Life

In Contact With a Cell

A closer look: BIO|materials interfaceat nanoscale

Kasemo, Surf. Sci. (2002).

Microscopic Understanding of BIO|materials Interface

OUTLINE

1. Water-surface interaction and a molecular view of hydrophilicity

2. DNA-carbon nanotube interaction and identification

3. Melanin structure and implications in phototechnology

I. Water at surfaces

Our strategy for water/surface

BondingH2O/Pt(111)Structure: monomer→multilayer

Vibration

H2O/metal

H2O/non-metal

Structure→Properties

Properties→Structure

Superhydrophilic

Wetting

Density functional theory (DFT)

)...,,,( 21 Nrrr

Many-electron Schrödinger equation Single-particle Kohn-Sham equation2

21 )...,,,( Nrrr

)()(]['

)(')(

22

2

rErVrr

rdrrV

m iiixcexternal

Theorem II.

Theorem I.

)...,,,( 21 Nrrr

)(r

Calculation details:

• Plane-waves

• Pseudopotential(USPP,PAW)+GGA(PBE)

• Ab initio molecular dynamics (MD)

• Nudged Elastic Band for reaction barriers W. KohnNobel prize, 1998

•Advantages: 1. First-principles (=“parameter free”) Input: atomic numbers

2. Practicality. N ~ 1000, accuracy ~ 0.01 eV.

•Drawback: Unknown Vxc? Approximations: LDA

GGA

))(()( rVrV xcxc

))(),(()( rrVrV xcxc

Time-dependent DFT (TDDFT) for electron dynamics

Our implementation:

• Real time

• Local bases: numeric atomic orbitals

• Parallelizable

• Order (N)

Biomolecules, nanomaterials

)(),,(ˆ)(),0(Given tttrH

t

ti

TDDFT:

Applications:

• Optical absorption (linear response)

• Excited state dynamics

• Chemical reactions

• Atom collision

• Quantum control (strong laser field)

Multiscale modelling

Time: 10-18 10-15 10-12 10-9 10-6 10-3 100 s

Length: 0.1 1 10 100 1000 109 nmThe scale we are working on:

Adapted from

DOE (2006)

TDDFTAb initio MD

Classic MDNon-atomistic models

Results

• Nearly atop, flat (14º), free rotation• dOH (0.98Å) elongated, HOH (105.

6º) broadened• Electron transfer (O→Pt) 0.02e• Diffusion barrier: 0.13 eV

Metal surfaces:

1.Simple structure.

2.Wide applications.

Single water adsorption

A Model System: H2O/Pt(111)

304 meV/H2O

Water clusters

433

359

520

H-bond energy in dimer: >260 meV > 240 meV (free).

H-bonds enhanced upon adsorption.

2D bilayers to 3D multilayers

H-up H-down 2 BL

• H-up: 512 meV

• H-down: 524 meV

• Half-dissociated: 291 meV No dissociation on Pt(111).

bottomup

Nature of water-Pt bonds

H-up H-down

Singlemolecule Dimer

• Chemical bonding

• Lone pair-surface d states

• Localized at bottom layer

d-orbital

lone-pair

Vibrational recognition

0 100 200 300 400 5000.00

0.02

0.04 H-down bilayer

438424

384

202196

91

6957

3416

6

In

tensi

ty (

arb

. un

its)

Vibrational Energy (meV)

0.00

0.04

0.08

53

H-up bilayer

467432

388

198

8769

32184

Meng et al., Phys. Rev. Lett. (2002).

Summary of water/Pt

• Top sites, nearly flatly, various structures• Water doesn’t dissociate

• Local lone pair-dz2,dxz chemical bonds • H-bonds strengthened

• OH stretches for vibration recognition • Two types of H-bonds

Structure

Bonding

Vibration

General trends: water-surface distance

Meng et al., Phys. Rev. B (2004); Phys. Rev. Lett (2003).

cHydrophili

cHydrophobi

E

E

ads

HB

1

1

Wetting order

Wetting order: RuRh Pd Pt Au

0

1

2

3

AuPtPdRhRu

EH

B/E

a

d7s1 d8s1 d9s1 d10s1

ω

hydrophobic

hydrophilic

Ice tessellation on silica

Yang, Meng et al., Phys. Rev. Lett. (2004).

Water/-cristobalite (100)

2.82 Ǻ

3.16 Ǻ

Inverse Design: Properties →Structure

Meng, Zhang, Kaxiras, Phys. Rev. Lett.(2006).

Why diamond?• Lattice match: 2% • Carbon only—biocompatible• Affinity to proteins/DNA: better than Si, Au• Additional merits: hardness, low friction etc• Low cost: Nanocrystalline

Na/K H/F

Applications• Self-cleaning• Scratch and fouling resistant• Heat transfer

A superhydrophilic biocompatible surfaceA superhydrophilic diamond surface

Wettability before and after surface modification

1 2 3 4 5 6 7 8 90.1

1

10

Water Coverage (BL)

1ML H 1/3ML K+2/3ML H 1/3ML Na+2/3ML H 1/3ML Na+2/3ML F

Eb=0.674 eV

>Eice=0.670 eV

hydrophobic

hydrophilic

superhydrophilic

Water/surface: Summary

BondingH2O/Pt(111)Structure: monomer→multilayer

Vibration

H2O/metal

H2O/non-metal

Structure→Properties

Properties→Structure

Superhydrophilic

Wetting

II. DNA nucleoside interaction and identification with carbon nanotubes

Nano Letters 7, 45 (2007).Nano Letters 7, 663 (2007).

Why DNA-carbon nanotube (CNT)?

Similarities:• Prototypical one-dimensional• Conducting properties: metal, semiconductor, or

insulator?

Differences: (single-stranded) DNA: CNT:• extremely flexible stiff• strongly hydrophilic hydrophobic• central in biology central in nanotechnology

Combine!

d(GT)30/CNT

Poly(T)/CNT(10,0)Zheng et al., Science (2003); Nature Materials (2003).

It is also possible to wrap CNT using long genomic ssDNA (>>100 bases).Gigliotti et al., Nano Lett. (2006).

Structure of the single-strand DNA wrapped CNT

Label-free DNA detection by electronic signals

Jeng et al., Nano Lett. (2006).

Ultimate goal:

Ultrafast DNA sequencing

What’s missing?• The nature of DNA-CNT interaction• Its dependence on nucleoside identity

Our objective

• Single nucleoside/CNT interaction• Discriminate nucleosides based on electronic

features

MethodologyConfiguration space search: Force FieldsElectronic features: Density Functional

TheoryNucleoside identification: Artificial Neural

Network

DNA detection by electronic means:Is that possible at single-base resolution?

DNA/CNT molecular dynamics

DNA wraps around CNT in a short time ~ 2 ns.

Proposed experimental setup

Meng, Maragakis, Papaloukas, Kaxiras, Nano Letters 7, 45 (2007).

i) Orientationii) Electronic featureiii) STM image

Most stable configurations

28.4% 27.6% 10.1%

~1000 configurations (local minima)

25.2% 6.8%

4.3% 3.2%

• Interaction through base plane

• van der Waals interaction: 0.7~0.8 eV (LDA~0.4 eV)

• Electric field further stabilizes adsorption

A/CNT

C/CNT

Reduced noises: favorable base orientations

Experiment Hughes et al. (Harvard)

On CNTFree

Our Theory: through very delicate simulations of electron dynamics (TDDFT), all features are reproduced if we align nanotube axis as indicated.

“Measured” vs. calculated orientation

•Red line: “measured” CNT axis

•Atomic models: Calculated

Meng et al., Nano Letters 7, 663 (2007).

• Mutual polarization

• Slight electron transfer from base to CNT

Electronic interaction: charge density

Red: electron depletionBlue: electron accumulation

Electronic features: density of states

Six featuresF1: HOMO

F2: LUMO

F3: Band-gap

F4: Number of occupied peaks

F5: Highest occupied peak

F6: Integral

100% accuracyH

OM

O LUM

O

Identification made easy: STM images

Ultrafast sequencing: currently availabe: 30, 000 bases/day (454 LifeSciences)

20 images/s: 1,728,000 bases/day!

Simulation Voltage: +1.4 V

Odom et al., Nature (1998).

Experiment: CNT

III. Melanin structure, flavonoids, and renewable energy applications

Melanin is a ubiquitous pigment

Existence• Human: skin, hair, eye, ear, brain• Animals• Plant• Microorganism

Functions • Photoprotection• Camouflage• Vitamin D synthesis• Antioxidation• Hearing• Parkinson’s disease

Meng & Kaxiras, Biophys. J. (2008).

Molecular structure unknown

300 400 500 600 700 800

Abs

orba

nce

(arb

. uni

ts)

Wavelength (nm)

UV-vis spectra

Monomer units

Chen et al., Pigment Cell Res. (1994).Meredith & Sarna, Pigment Cell Res. (2006).

?

Our model: a porphyrin-like 2nd structure

Kaxiras, Tsolakidis, Zonios, Meng, Phys. Rev. Lett. 97, 128102 (2006).

X-ray

UV-vis

Hybrid melanin/solid structure for photo-technology

“Just for flavor”: Flavonoids as one of natural pigments

Flavonoids Chlorophyll Carotene

TiO2 Dye-sensitized TiO2

Melanin

? A

B

C

Attach the pigment to TiO2 semiconductor

Band structure wavefunction

HOMO

LUMO

Optical absorption

experiment

dye

dye/TiO2

Wongcharee et al. (2007).

Charge injection dynamics

T=350 Kδt=0.02419 fs

excited electron

HOMO

LUMO

Pigment/semiconductor antenna system for solar cells

e

e

ConclusionsConclusions

• Design biocompatible, superphilic surface

• DNA/CNT at different levels -Experimental determination of base orientation -Electronic characteristics in spectroscopy and images: ultrafast sequencing

• Renewable energy applications -Porphorin-like melanin structure -light harvest

BIO|materials contact very promising.

water/surface

pigment/TiO2DNA/CNT

Basics:

Sensor: Energy:

AcknowledgmentAcknowledgmentTheoretical:Prof. E.G. Wang (IoP,CAS)Dr. Jianjun Yang (IoP,CAS)Dr. Yong Yang (IoP,CAS)

Prof. Efthimios KaxirasDr. Weili WangDr. Maria FytaDr. Yina MoDr. P. Maragakis (DE Shaw Co.)Dr. C. Papaloukas (Ioannina U)

Experimental:Prof. Jene A. GolovchenkoProf. Daniel BrantonMary HughesProf. Michael Aziz

Funded by:Prof. Shiwu Gao (GU/Chalmers)Prof. B. I. Lundqvist (Chalmers)

Prof. Zhenyu Zhang (ORNL/UT)Dr. Wenguang Zhu (UT)Yang Lei (U. London)

Prof. Bengt Kasemo (Chalmers)Prof. D. V. Chakarov (Chalmers)

Prof. Martin Wolf (Freie U, Berlin)Dr. Ch. Frischkorn (Freie U, Berlin)

Prof. P. Meredith (U. Queensland)Jennifer Riesz (U. Queensland)

Prof. Z.X. Guo (U. London)

l

alalvv vtvmtC )0()()(

)()())(2/(1)(exp

)()()(

2/1

2/1

kakakRkitki

kikeNmtv

jjjlj

kjj

jaal

kjjB

kjjjjvv

tkiTkV

knktkiV

tC

)(exp)2(

)2/1)()(()(exp)2(

)(

3

3

kj

j kg

))(()( Fourier transformation

Molecular dynamics (MD) simulation

Vibrational spectrum

Basics about waterflexible Water structure according to BFP rule

Ice Liquid water

~0.24 eV

even more flexible

Connection with macroscopic

= – Eads

Werder et al., J. Phys. Chem. B (2003).

θ ∝ -Eabs= -EHB/ ω

Assume EHB =constant, θ ∝ -1/ ω (ω> ω0 )θ = 180° – 108° / ω

Water/Pt vs Water/Au

Meng et al., J. Chem. Phys. (2003).

ω

(III) Hydrogen generation and storage•H2 from water splitting

-thermal precious metal (Pt)/O2 separation

-electrolysis electricity?

-photolysis solar cell?

-biophotolysis hydrogenases (cyanobacteria) ?

Ti

CNT

H2/TiB2 nanotube: 5.5 wt%, 0.2-0.6 eV

Meng, Kaxiras, Zhang, Nano Lett. (2007).