16
Why Don't Alcohol and Water Mix Very Well? “The molecular structure of Alcohol-water mixtures” J.-H. Guo Advanced Light Source, Lawrence Berkeley National Laboratory

Why Don't Alcohol and Water Mix Very Well? “The molecular structure of Alcohol-water mixtures”

Embed Size (px)

DESCRIPTION

Why Don't Alcohol and Water Mix Very Well? “The molecular structure of Alcohol-water mixtures”. J.-H. Guo Advanced Light Source, Lawrence Berkeley National Laboratory. Collaborators. Y. Luo, S. Kashtanov, & Hans Ågren ( Theoretical Chemistry, Royal Institute of Technology ) - PowerPoint PPT Presentation

Citation preview

Page 1: Why Don't Alcohol and Water Mix Very Well? “The molecular structure of Alcohol-water mixtures”

Why Don't Alcohol and Water Mix Very Well?

“The molecular structure of Alcohol-water mixtures”

J.-H. GuoAdvanced Light Source, Lawrence Berkeley National Laboratory

Page 2: Why Don't Alcohol and Water Mix Very Well? “The molecular structure of Alcohol-water mixtures”

Collaborators

Y. Luo, S. Kashtanov, & Hans Ågren (Theoretical Chemistry, Royal Institute of Technology)

A. Augustsson, J.-E. Rubensson, & J. Nordgren (Department of Physics, Uppsala University)

D. K. Shuh (Chemical Sciences Division, Lawrence Berkeley National Laboratory)

Page 3: Why Don't Alcohol and Water Mix Very Well? “The molecular structure of Alcohol-water mixtures”

Molecules in gas and liquid phases

Small molecules can be studied in great detail in the gaseous phase, where molecular interactions can be neglected, and consequently both their geometric structure and electronic structure are often well known.

When the molecules interact in the liquid phase, our knowledge about these fundamental properties is very limited. The arrangement of the molecules change on a fast time scale: the geometry and the electronic structure of the molecules themselves vary, i.e., the properties of the individual molecules are constantly changing. Thus, from this perspective it is not surprising that there is still much to learn about common and simple liquids.

540535530525520515510Energy (eV)

1g-1

1u-1

3u-1

4g-1

CO2O-K emission

resonant O1s - * 535.0 eV

-1.0 eV

-1.5 eV

-1.75 eV

-2.0 eV

-2.3 eV

non-resonant551 eV

Skytt et al., Phys. Rev. Lett. 77, 5035 (1996).

Page 4: Why Don't Alcohol and Water Mix Very Well? “The molecular structure of Alcohol-water mixtures”

Microscopic Mixing of Liquids

Solvation in aqueous solution has been a most intensive studied problem in chemical physics. The observed entropy increase upon mixing alcohol and water is much smaller than expected for an ideal solution.

The anomalously small entropy increase upon solution of alcohols in water is traditionally explained in terms of a hydrophobic interaction with the apolar alcohol headgroups, which induces an ice-like structure in the surrounding water [Frank et al., J. Chem. Phys. 13, 507 (1945).]

Early neutron diffraction data provides structure information of water cages around hydrophobic headgroups in solution [Soper et al., Phys. Rev. Lett. 71, 4346 (1993); Tsai et al., J. Chem. Phys. 104, 9417 (1996); Dixit et al., J. Phys. Condens. Matter 12, L323 (1999); Wakisaka et al., J. Mol. Liq. 90, 175–184 (2001).]

New neutron diffraction data demonstrated that incomplete mixing at the molecular level is essential to explain the negative excess entropy observed when methanol dissolves in water [Dixit et al., Nature 416, 829 (2002).]

We examine the influence of the intermolecular interaction on the local electronic structure of liquid methanol, water, and their solutions. Our study determines the molecular structure of both liquid methanol and water-methanol solutions in unprecedented detail, thus, forms the final leg of the "iceberg, cluster, molecular structure" story pertaining to alcohol-water solutions.

Page 5: Why Don't Alcohol and Water Mix Very Well? “The molecular structure of Alcohol-water mixtures”

Methanol crystal: infinite hydrogen-bonded chains; liquid methanol: non consensus after decades of studies

The early hypothesis of cyclic structures by Pauling [L. Pauling, The Nature of the Chemical Bond , Cornell Univ. Press, 1960] has both been supported and contested by neutron diffraction analysis; the competing interpretation being that the majority of liquid molecules are ordered in chains with up to 10 members or linear trimer/tetramer chains

Based on neutron diffraction data, it was found experimentally:

chains up to ten molecules with average of six molecules in liquid methanol [Yamaguchi, K. Hidaka, and A.K. Soper, Molecular Physics 96, 1159 (1999)]; Haughney et al., J. Phys. Chem. 91, 4934 (1987); Svishchev et al., J. Chem. Phys. 100, 5165 (1994) claimed the same findings based on MD calculation;

Sarkar, S. & Joarder, R. N., J. Chem. Phys. 99, 2032 (1993) revealed a different picture of cyclic hexamers (six-rings);

Using the same diffraction data, Tanaka found linear trimer and tetramer chains [Bull. Chem. Soc. Jpn. 58, 270 (1985)]

Molecular structure in liquid methanol

Page 6: Why Don't Alcohol and Water Mix Very Well? “The molecular structure of Alcohol-water mixtures”

Molecular structures in liquid methanolMolecular structures in liquid methanol

3 87654 9

3

87

654

9 10

chainschains

ringsrings

Page 7: Why Don't Alcohol and Water Mix Very Well? “The molecular structure of Alcohol-water mixtures”

Soft-x-ray spectroscopy

.

h

h'

UnoccupiedMOs

OccupiedMOs

O 1s

Inte

nsit

y (a

.u)

540535530525520515Energy (eV)

XASXES4a1

2b2

3a1

1b2

1b1

Primer excitation: 10-18 sDe-excitation: 10-15 s

Page 8: Why Don't Alcohol and Water Mix Very Well? “The molecular structure of Alcohol-water mixtures”

Wet samples in vacuum conditions

100nm Si3N4, ~4 µl liquid volume, 66% transmission @O1s, vacuum pressure < 1 x 10-9 Torr

grating

detector

slit

O-ring

Si3N

4

liquid

SR

PHIOCA: photon-in and photon-out chemical analysis

100

80

60

40

20

0

Tra

nsm

issi

on (

%)

1000800600400200Energy (eV)

O 1s (66%)

Fe 2p (82%)

C 1s (46%)

C_100nm; SiC_100nm Si_100nm; SiN_100nm Al_100nm

Page 9: Why Don't Alcohol and Water Mix Very Well? “The molecular structure of Alcohol-water mixtures”

Inte

nsity

(a.

u.)

530525520515Energy (eV)

(a)

(b)

(c)

(d)

(e)

H2O molecule and liquid water

535530525520515Energy (eV)

water molecule liquid water

Theory

Experiment

1b1

3a1

1b2

1a1

1b1

3a1

1b2

2a1

Molecular orbitals

Hatree-Fock level with Sadlej basis set using the DALTON program on 32 nodes of the T3E computers at the NSC in Linköping

Guo et al., Phys. Rev. Lett. 89, 137402 (2002)

Finding: an electron sharing takes place between water molecules. Such a sharing mainly involves the so-called 3a1 orbital, which is a mixing of oxygen 2p and hydrogen 2s atomic orbitals.

Page 10: Why Don't Alcohol and Water Mix Very Well? “The molecular structure of Alcohol-water mixtures”

Inte

nsit

y (a

.u.)

530525520515Energy (eV)

Resonant Non-resonant

SYM

A-ASYM

D-ASYM

Experiment540536532

O 1s gas water

2b24a1

liquid water

Hydrogen bonding structure in liquid water

4-HB

3-HB-O

3-HB-H

The X-ray emission spectra reveal the influence of hydrogen bonding on the local electronic structure of liquid water

Page 11: Why Don't Alcohol and Water Mix Very Well? “The molecular structure of Alcohol-water mixtures”

Molecular structure in liquid methanol

Finding: The structure of liquid methanol at room temperature is a combination of rings and chains each made up of either 6 or 8 methanol molecules

Guo et al., Phys. Rev. Lett. 91, 157401 (10 Oct. 2003)

Inte

nsit

y (a

.u.)

528526524522520Energy (eV)

Calculated O K-emission of methanol chains

3

4

5

6

7

8

9

10

Inte

nsit

y (a

.u.)

532528524520516Energy (eV)

8-ring

6-ring

8-chain

6-chain

Theory

Experiment

O K-emission

B

C

A

560550540530

O 1s absorptionA

BC

Page 12: Why Don't Alcohol and Water Mix Very Well? “The molecular structure of Alcohol-water mixtures”

530525520515Energy (eV)

IM + IW Methanol Water mixture

B

C

Oxygen K-emission

3a1

E1'

E2

E4

E1

1b1

1b2

E3

Incomplete mixing in alcohol-water solution

X-ray emission spectra of an equimolar mixture of methanol and water are compared to the spectra from the pure liquids. We find that a 1-to-1 combination of the pure liquid spectra reproduces the solution spectra to considerable detail. The spectra are very sensitive to the changes in the local electronic structure and this observation indicates incomplete mixing at the microscopic level.

However, there is a significant discrepancy (indicated by arrow) when the excitation energy is selected to emphasize the chain structures (spectrum B). The solution spectrum shows a relative intensity decrease in the E1' region, indicating a depletion of methanol chains in the solution.

difference

Page 13: Why Don't Alcohol and Water Mix Very Well? “The molecular structure of Alcohol-water mixtures”

535530525520Energy (eV)

Oxygen RIXS

530.6 eV

532.5 eV

531.5 eV

533.5 eV

534.5 eV elastic peaks

544540536532528

A' CH3OH mixture H2O

O 1s A

B

RIXS of alcohol-water mixture

The narrow spectral feature shown in the resonant emission spectra holds the key for identifying the structures. Based on the simulations of resonant emission spectra of water and methanol in both gas and pure liquid phases, this narrow spectral feature can only be generated from the emission of a water molecule that is "isolated" from the rest of the water network. We also know that there must be an interaction between water and methanol chains because there is a depletion of methanol chains in the solution. This results in a significant reduction of the number or possible trial structures. The new structures in alcohol-water mixture must be responsible for the spectral profiles.

Page 14: Why Don't Alcohol and Water Mix Very Well? “The molecular structure of Alcohol-water mixtures”

RIXS of liquid methanol and water

550540530520510500Energy (eV)

Prepeak

-1 eV

-1.5 eV

-2.5 eV

-3 eV

-4 eV

-5 eV

-7 eV

+3 eV

Normal

Water

Inte

nsit

y (a

.u.)

535530525520515Energy (eV)

O K-emissionMethanol

560.0 eV

537.0 eV

535.5 eV

534.5 eV

533.5 eV

532.5 eV

The RIXS spectra of liquid methanol and water under detuning conditions are different to their mixture

Page 15: Why Don't Alcohol and Water Mix Very Well? “The molecular structure of Alcohol-water mixtures”

532530528526524522520Energy (eV)

Resonantly excited O K-emission

8-chain + 3-water

A'--

A'-

A'

E1

E3 E2

8-chain + 1-water 8-chain + 2-water

6-chain + 2-water 6-chain + 3-water 6-chain + 4-water

6-chain + 2-water

6-chain + 4-water

6-chain + 3-water

8-chain + 1-water

8-chain + 3-water

8-chain + 2-water

Molecular structure in alcohol-water mixture

Theoretical spectra of a methanol molecule in various methanol-water local structures are computed. The structures presented above must be responsible for the spectral profiles.

Page 16: Why Don't Alcohol and Water Mix Very Well? “The molecular structure of Alcohol-water mixtures”

Summary

In liquid water: A strong involvement of a1-symmetry valence states in

the hydrogen bonding indicates there is electron sharing between water molecules

In liquid methanol: The molecules in the pure methanol liquid predominantly persist as hydrogen-bonded chains with 6 and/or 8 molecules, and as 6- and/or 8-molecule rings.

For an equimolar water-methanol mixture, water and methanol molecules are incomplete mixing at the microscopic level.

In addition, our results indicate a new mechanism for increasing order in the solution: water molecules bridge methanol chains to form rings with 6 and 8 methanol molecules.

Many vital chemical and biological processes take place in aqueous solutions. We believe that the techniques have great general potential to provide new and valuable information in the quest for the microscopic origin of the properties of liquids and solutions; in particular the RIXS, will open new research opportunities in physical chemistry, nanotechnology, as well as in biochemistry and biology.