C&DiCE 2015

Solvation & SoluteDynamics in Ionic Liquids

Mark Maroncelli, Penn State

1. Translation & Rotation

2. Solvation Dynamics

3. Reaction Kinetics & Heterogeneity

4. Effect of Nanostructure?

S1

S0

t

Volume Ratio VU / VV

0.01 0.1 1

Fric

tion

Rat

io

obs/

SE

10-3

10-2

10-1

100

101

SE

Li+NH4

+

neutralsolutes

ionicsolutes

CH4

C&DiCE 2015

(Room Temperature) Ionic Liquids

Lowering Tf: charge delocalization size mismatch low symmetry conformational disorder

imidazoliumImn1

+pyrrolidinium

Prn1+

phosphoniumP1234

+

BF4-, PF6

-, CF3SO3-,

Cation Families:

Common Anions:

Tf2N-DCA-

ammoniumN1234

+

Our Interests: Distinctive solvation/chemistry? New lessons to be learned?

[Pr41][Tf2N] Tf=3 C

Low Volatility High Thermal Stability Intrinsically Conductive Wide Electrochemical Window Solubility/Phase Control

Typical Properties:

C&DiCE 2015

1. The Simplest Dynamics:Solute Translation & Rotation

3

Anne Kaintz Chris RumbleJuan Carlos Araque

C&DiCE 2015

DSE / m2 s-110-12 10-11 10-10

Dob

s / m

2 s-1

10-12

10-11

10-10

Viscosity / cP10 100 1000

obs

/ S

E

0.2

0.4

0.6

0.8

1.0

Ion Self-Diffusion Coefficients & Friction

4

1:1

cations

anions

.5.2

data from Watanabe, Matsumoto, Martinelli, ...

SERTkD B

SE 6

Stokes-Einstein (Sutherland)Hydrodynamic Prediction:

friction on sphere of radius R

obs

SE

SE

obs

DD

Friction Ratio:measure of the coupling between the diffusor and solvent bath

SE predictions are reasonably accurate for estimating ion self diffusion in ILs

nearly identical situation exists in conventional solvents

Ion D in 32 ILs (298 K)

C&DiCE 2015

Diffusion of Aromatic Hydrocarbons in ILs

Volume Ratio VU / VV

0.4 0.6 0.8 1.0 1.2

Fric

tion

Rat

io

obs/

SE

0.0

0.1

0.2

0.3

0.4

0.5

0.6

[Prn1][Tf2N] n=3, 4, 6, 8, 10

Kaintz et al., J. Phys. Chem. B 117, 11697 (2013).

PFG-NMR measurements of simple aromatic hydrocarbons

in a homologous series of ILs

departure from SE predictions increases with decreasing solute and increasing solvent size

shape is also a factor

n=10

n=3

Bz>Na>Py>An>Bpoblate ~prolate

obs/SE vs. Solute/Solvent Size

solute solvent

C&DiCE 2015

Neutral Solutes Diffusing in ILs

6lit. data from: Noble, Baltus, Scovazzo, Kimura, Hardacre, Compton, License, Halpiot, Lagroste, Hussey, ...

nonpolar solutes

polar solutes

literature data:323 solute/solvent pairs, 37 solutes, 56 ILs small gaseous solutes:

CO, CO2, C2H4, C4H10, ... large solutes:

NF

F

F

N

=5 D

N

N

Fe

Volume Ratio VU / VV

0.1 1

Fric

tion

Rat

io

obs/

SE

10-3

10-2

10-1

100

pVU VVa )/(1

nonpolarpolargasother

SE

>100

marked departures from SE predictions primary determinant is U/V size ratio

C&DiCE 2015

Volume Ratio VU / VV

0.01 0.1 1

Fric

tion

Rat

io

obs/

SE

10-3

10-2

10-1

100

101

SE

Li+NH4

+

neutralsolutes

ionicsolutes

CH4

Charged (1e) Solute Diffusion in ILs

7

obs/SE~1 for ions (helpful for interpretations) large difference in obs/SE for small

ions compared to neutral solutes the “solventberg” limit well-

documented for Li+

(self diffusion in neat ILs)

analogous but weaker trends also seen in conventional solvents

mismatch between solute and solvent in size or interactions responsible for marked departures from SE predictions

C&DiCE 2015

Transport Mechanisms from MD

8Araque, Margulis et al., J. Phys. Chem. B 119, 7015 (2015).

[Pr41][Tf2N]

CH4 NH4+

C&DiCE 2015

Trajectory-Resolved Structure

9

CH4

NH4+

jumps are associated with low density (and low polarity = “soft”) and caging with high density (and polarity = “stiff”) regions of the liquid

CH4 explores both regions more effectively than NH4+

C&DiCE 2015 10

CH4

NH4+

blue = cation headsgreen = cation tails

(only anionsdepicted)

CH4 exchanges with nonpolar tails -- Coulomb lattice unperturbed NH4

+ exchanges anions in its 1st shell -- Coulomb lattice must rearrange

C&DiCE 2015

/T / (cP K-1)10-3 10-2 10-1 100

Rot

atio

n tim

e

rot

/ s

10-11

10-10

10-9

10-8

10-7

aproticprotic

ImPrNP

Solute & Ion Rotation

11

Jin et al., JPCB 111, 7291 (2007)

Fluorescence Measurements

Cat/An DCA- BF4- PF6

-

Im21+ 220 200

Im41+ 83 310 110

O ON

CF3

slip

stick

TkLLV

B

hydLrot )1(

6)(

VvdW/Vhyd from Dielectric Measurements

rotation of large solutes not unusual Im+ cations and other small molecules

may show more interesting dynamics

VvdW/Vhyd VvdW/Vhyd

65

22

25

4

2

Recent NMR & Fluorescence Data

C&DiCE 2015

2. More Complicated Dynamics:The Solvation Response

12

Min Liang Xin-Xing Zhang

C&DiCE 2015

The Spectral / Solvation Response

13

)(t

)()0()()()(

ttS

Solvation Coordinate

Ener

gy

S1

S0

t

peak frequency

spectral or solvationresponse function

time

“Dynamic Stokes Shift”of Emission Spectra

basic property of polar liquids determines friction on charge motion

C&DiCE 2015

S(t) in Ionic Liquids

14

upconversion (FLUPS) & time-correlated single photon counting (TCSPC)experiments combined to capture the full response over 80 fs – 40 ns

Frequency / 103 cm-116 18 20 22

Rel

ativ

e In

tens

ity

20 ns

100 fs

TCSPC FLUPS

Emission Spectra in [Im41][PF6]

t

Time / s10-13 10-12 10-11 10-10 10-9 10-8

Spec

tral R

espo

nse

S ( t)

0.0

0.2

0.4

0.6

0.8

1.0

~100-fold slower than dipolar solvents ( also 100-fold larger)

strongly bimodal decay broadly distributed in time

S(t) in CH3CN & IL

O ON

CF3

CH3CN

IL

300x

C&DiCE 2015

Time / s

S (t)

Dissecting the Solvent Response

15

})/(exp{)1(}exp{)( 2221

tftftS GGG

Zhang et al., JPCB 117, 4291 (2013).

Viscosity / cP10 100

Slow

Tim

e

str

/ ps

100

1000

[Imn1][Tf2N][Pr

n1][Tf2N]

24

42

43

21

M1

Inverse Reduced Mass mu/0.00 0.01 0.02 0.03 0.04

Gau

ssia

n Fr

eque

ncy

G /

ps-1

0

2

4

6

8

10

12

14

R=.92

IRF

21

41

M1

[Imn1][Tf2N][Prn1][Tf2N]

42 22

mechanism is primarily solvent translation #1: 30-40% inertial motions displacements of 1st shell ions

#2: the rest is overdamped & highly coupled ion motions subtle displacements of ions over <1

#1

#2 #1#2

( )

N=21

C&DiCE 2015

[Im41][DCA]

0.0

0.2

0.4

0.6

0.8

1.0

B1B2

[Im41][PF6]

Time / ps10-1 100 101 102 103

Solv

atio

n R

espo

nse

S(t)

0.0

0.2

0.4

0.6

0.8

1.0

BN*

[N3000][NO3] B*W*

[S333][Tf2N]

10-1 100 101 102 103

BW

Solvation & Dielectric Relaxation

16

Some S(t) Comparisons

Obs.

Pred. #1#2

dielectric continuum predictions are qualitatively correct in ILs

but too fast by factors of 3-4 similar results in IL + dipolar

solvent mixtures

5x

Zhang et al., JPCB 117, 4291 (2013).

Solvation Dynamics

Dielectric Response

h

)(E

)(P

2x

4x

3x

C&DiCE 2015

Why Continuum Predictions are too Fast

17

Maroncelli et al., Faraday Discuss.154, 409 (2012).

Charge Density Correlation Function),()(),( tkktk qqqq

Time / ps10-1 100 101 102 103

qq

( k, t)

0.0

0.2

0.4

0.6

0.8

1.0

k=0.1 Å-1

0.8

1.0

0.5

dipole solvation

k

() captures only k=0 (=) dynamics solvation involves entire spectrum k=0- solvent response is slower at k>0 thus

dielectric predictions are too fast

• molecular solvation entails charge redistribution in response to non-uniform fields; basic dynamics captured in qq(k,t) of the neat solvent

(k=2/=ion size.)

C&DiCE 2015

Solvation & Conductivity

18

0

21

0 4)()(

c

solv dttS

Zhang et al., JPC Lett. 4, 1205 (2013).

The D.C. model predicts integral solvation times solv to be inversely proportional to conductivity ~const.

as predicted, solv and 0 are strongly correlated

prediction is an average of 3-fold too fast

( is easy)

Resistivity -1 / m100 101 102

Solv

atio

n Ti

me

s

olv

/ p

s

102

103

104

Im+

Pr+

N+

P+

misc

Tf2N

-

mis

c

continuumprediction

C153 in 34 Ionic Liquids

04)(ˆ)(ˆ icorr

4)(ˆ)(ˆ i

The () 0 Connection:

expts.report

divergenceremoved

conductivity & permittivity reflectidentical dynamics

)(tMM

)(tJJ

C&DiCE 2015

3. Even More Complicated: Reaction Kinetics

19

Min Liang Lillian Li Minako Kondo

C&DiCE 2015

solv / ps10-1 100 101 102 103

et /

ps

100

101

102

103

1:1

1. Intramolecular Charge Transfer

20

TR Emission in [N3111][Tf2N]

Frequency / 103 cm-112 14 16 18 20 22

Rel

ativ

e In

tens

ity t=02.5 ns

Biphenyl Acridinium (BPAc+)

N N+

e+

LE CT

LECT

[Nip311][Tf2N][Im21][Tf2N][Im41][PF6][N3111][Tf2N][P14,666][Tf2N]

reaction in ILs continues trendestablished by dipolar solvents

Reaction and Solvation Times

Li et al., JPCB 115, 6592 (2011).

C&DiCE 2015

Calculated Time clc / ps101 102 103

Obs

erve

d Ti

me o

bs /

ps

101

102

103

21

N

S

N

2. Isomerization

DMASBT

An Empirical Correlation

)()/ln(ln cfTba

[Nip311][Tf2N]

Time / ps0 500 1000 1500 2000

Rel

ativ

e In

tens

ity10-2

10-1

100

298 K

283

279 K

CH3CN

Non-Exponential Kinetics

rates in ILs correlate with those in dipolar solvents

kinetics more complex

Kondo & Maroncelli, JPCB 117, 12224 (2013).

C&DiCE 2015 22

REE & Kinetic Heterogeneity

N

S

N

heterogeneous kinetics should be the norm for ps processes given the slow solvation / structural relaxation times of ILs

Ener

gy

Solvation Coordinate

S1

S0

p()

ex / 103cm-120 22 24 26 28

rxn /

ps

0

100

200

300

400

500

em / 10

3 cm-1

19.2

19.4

19.6

19.8

20.0

abs

em

%50~

Reaction Time vs excRed Edge Excitation

C&DiCE 2015

Wavelength / nm450 500 550 600 650

Rel

ativ

e Em

issi

on In

tens

ity

0.0

0.2

0.4

0.6

0.8

1.0

1.2exc/nm

350370390400410430450470

Other Examples

23

CN

N +

‐LE CT

DMABN

-

+

+

‐

DEAHFN* T*

Wavelength / nm350 400 450 500 550 600

Rel

ativ

e Em

issi

on In

tens

ity

0.0

0.5

1.0

1.5

2.0exc/nm

280285290295300305310315

Samanata & Co., JPCB 114, 1967 (2010).

TICT Reaction of DMABN

LE CT

N* T*

exc exc

ESPT in DEAHF

Kimura et al., JPCB 114, 11847 (2010);JPCB 117, 6759 (2013).

C&DiCE 2015 24

Lopes & Padua, J. Phys. Chem. B 110, 3330 (2006).

Distinct Local Dynamics Distinct Local Energies Nanoscopic Domains

Hurley et al., Phys. Rev. E52, 1694 (1995)

reaction senses “dynamic heterogeneity” – the fact that local packing effects cause different regions to have different fluidities

different solute-solvent interaction energies affect PES for reaction

solutes are distributed among nanoscopicdomains of different polarity and/or fluidity

Possible Interpretations

S1

C&DiCE 2015

4. Aside: Where’s the Effect of Nanoscopic Structure?

25

Jacob Schesser

C&DiCE 2015

Solvation in [P14,666][X] ILs

26

Viscosity Correlation

Shimizu, J. Mol. Struct. 946, 70 (2010).

blue=+, red=-, grey=alkane

Snapshot of [P14,666][Tf2N] Simulation

Resistivity -1 / m100 101 102

Solv

atio

n Ti

me

s

olv

/ p

s

102

103

104

Im+

Pr+

N+

P+

misc

Tf2N

-

mis

c

continuumprediction

Resistivity CorrelationViscosity / cP

101 102 103

Solv

atio

n Ti

me

s

olv

/ p

s

102

103

104

Im+

Pr+

N+

misc

Tf2N

-

mis

c9x

C&DiCE 2015

Further Recent Attempts

27

- anthracene & bianthryl in [Imn1][Tf2N] (n=2-12), & [P14,666][Tf2N]+n-hexane

Frequency

Abso

rban

ce

Anthracene Absorption SpectraAnthracene Solvatochromic Shifts

Predicted -gas / cm-1900 1000 1100 1200 1300

Obs

erve

d - g

as /

cm-1

900

1000

1100

1200

1300

n-C5

n-C16

Imn1 n=2-12

DMSO

P14,666

+n-C6

MeOH

THF

21

21

21

2

2

2

2

nn

nn

gas

= 4100 cm-1, = 170 cm-1

in 74 organic solvents:

C&DiCE 2015

Bianthryl ET Reaction Times

28

LECT

t

Frequency

Emis

sion

Int

ensi

ty e

+

-LE CT

ET R

eact

ion

Tim

e /

ps

C153 Solvation Time / ps

TR Emission Spectra

ET Reaction Times

ILs+ mixtures[P14,666][Tf2N]+ n-C6 xC6 =0.8

M. Liang, preliminary results

green=n-C6red=P14,666

+

blue=Tf2N-

C&DiCE 2015 29

Summary

D T/ as expected by hydrodynamic models, but Stokes-Einstein predictions can be wrong by factors of 100 or more (at room T)

diffusion of small neutral versus charged solutes is markedly different

rotations of small solutes can be far from hydrodynamic expectations

solvation is highly bimodal: ~1/3 sub-ps and 2/3 relax over 1 ps- 10 ns

unlike dipolar solvents, translational solvent motions dominate

S(t) and () reflect the same underlying molecular dynamics but dielectric continuum predictions too fast by a factor >4

solvation closely related to conductivity (solv1/)

slow solvation/structural relaxation in ILs leads to heterogeneous reaction kinetics for fast reactions (<1 ns)

origins of heterogeneity unclear (especially relevance of nanostructure)

Solute Translation & Rotation:

Solvation Dynamics:

Fast Reaction Kinetics:

~

C&DiCE 2015

Acknowledgements

30

Humboldt University Penn State University

Min LiangXin-Xing Zhang Anne Kaintz

Sergei ArzhantsevHui JinDurba RoyMinako KondoLillian LiJacob Schesser*

Chris Rumble*Brian ConwayMike ShadeckMarissa Saladin

Collaborators:Claudio Margulis, University of IowaEd Castner, Rutgers UniversityGary Baker, University of MissouriNikolaus Ernsting, Humboldt University

Other Contributors

Juan CarlosAraque

Universityof Iowa