Key Points (2020-02-05)What we learned• Nanochemistry at solid-water

interfaces Surface Charge and zeta

potential (pHzpc vs. pHiep) Adsorption (inner sphere and

outer sphere) vs. precipitation

• Nanochemistry at solid-water interfaces Precipitation vs. Dissolution Homogeneous vs. heterogeneous nucleation Lattice mismatch Euhedral shape Different growth modes depending on

interfacial energy relationship Homo- vs. hetero- epitaxy

What we will learn today• Nanochemistry at air-water

interfacesIons at air-water interfaceCarbon dioxide-water interface at conditions of gas hydrate formation

• Nanochemistry at solid-water interfaces1. Small Angle X-ray Scattering2. Stability of nanoparticles in aqueous systems

“Stable cluster formation in aqueous suspensions of iron oxyhydroxide nanoparticles”Journal of Colloid and Interface Science, Volume 313, Issue 1, 2007, Pages 152-159Benjamin Gilbert, Guoping Lu and Christopher S. Kim

• Nanochemistry at air-water interfacesIons at air-water interface

Science, Volume 303, 2004, Pages 1146-1147Bruce C. Garrett

Take Home Message

http://classes.engineering.wustl.edu/eece534/

1147

tions with surface vibrational spectra.They show that the depth of the interfa-cial region that contributes to the VSFSsignal is larger for the bromide and iodidesolutions than for neat water. This in-crease in interfacial depth supports theobservation from simulations that bro-mide and iodide segregate near the sur-face with a subsequent enhancement ofsodium below the anion layer. In addition,Saykally and co-workers have obtainedstrong evidence for the presence of thelarger halide anions at the liquid water

surface from resonant SHG experiments(10).

The results qualitatively support the com-putational predictions that chloride, bro-mide, and iodide are present in the surfaceregion. Testing the quantitative predictionsabout the concentrations of ions and theirproximity to the top of the surface regionwill require careful modeling and theoreticalanalysis of the surface vibrational spectra.

The combined evidence from simula-tions and experiments for the presence ofions at the air/water interface is com-

pelling. So what is wrong with the thermo-dynamic analysis that led to the conven-tional picture? The thermodynamic analy-sis assumes a binary system of water andsalt, with salt treated as one component.The simulations show that cations and an-ions behave differently and should be treat-ed as separate components. Furthermore, itis possible to have ions at the interface andstill have a net depletion in the surface re-gion, because ions may be depleted in thesecond layer. Finally, the same simulationsthat predict ions in the surface layer alsoreproduce the observed variations in sur-face tension with ion concentration (4),which are the basis of the conventional pic-ture of the interface.

The quantitative discrepancies betweenmolecular simulations and experiments re-main to be resolved. Other direct experi-mental probes of molecular structure at theliquid interface, such as extended x-ray ad-sorption fine structure (EXAFS) spectra(13), may provide new insight. Molecularsimulations must also be refined by in-cluding the direct use of electronic struc-ture methods. It is already clear, however,that the surface layers of many salt solu-tions contain ions that can participate inchemical processes.

The potential impact of interfacial ionson atmospheric chemistry is only begin-ning to be realized. Seinfeld (14) hashighlighted the importance of heteroge-neous chemistry in the atmosphere andthe role that surface ions may play. For ex-ample, there is now strong evidence thatreactions of surface chloride ions withOH radicals are a significant source ofmolecular chlorine (2), the precursor forchloride atoms that are important oxi-dants in marine environments.

References and Notes1. J. H. Hu et al., J. Phys. Chem. 99, 8768 (1995).

2. E. M. Knipping et al., Science 288, 301 (2000).

3. A. Laskin et al., Science 301, 340 (2003).

4. P. Jungwirth, D. J. Tobias, J. Phys. Chem. B 105, 10468

(2001).

5. P. Jungwirth, D. J. Tobias, J. Phys. Chem. B 106, 6361

(2002).

6. L. X. Dang, T. M. Chang, J. Phys. Chem. B 106, 235

(2002).

7. P. Ayotte et al., J. Phys. Chem. A 103, 10665 (1999).

8. E. A. Raymond, G. L. Richmond, J. Phys. Chem. B, in

press.

9. D. Liu, G. Ma, L. M. Levering, H. C. Allen, J. Phys. Chem.B 108, 2252 (2004).

10. R. Saykally, personal communication.

11. G. L. Richmond, Chem. Rev. 102, 2693 (2002).

12. E. A. Raymond, T. L. Tarbuck, M. G. Brown, G. L.

Richmond, J. Phys. Chem. B 107, 546 (2003).

13. K. R. Wilson et al., J. Chem. Phys. 117, 7738 (2002).

14. J. H. Seinfeld, Science 288, 285 (2000).

15. I thank G. Richmond, H. Allen, P. Jungwirth, D. Tobias,

and B. Finlayson-Pitts for helpful discussions; G.

Richmond and H. Allen for preprints of their unpub-

lished work; P. Jungwirth for providing the figure; and

the Office of Basic Energy Sciences, U.S. Department

of Energy (DOE), for support. PNNL is operated for

DOE by Battelle.

0.0 0.5 1.0 1.5 2.0 2.5 3.0

ρ (z) /ρb

z (Å

)z

(Å)

z (Å

)z

(Å)

D H

C G

B F

A E

0

4

8

12

16

0

4

8

12

16

0

4

8

12

16

0

4

8

12

16

NaF

NaBr

NaCl

NaI

The air/liquid interface of sodium halide solutions. (A to D) Snapshots from molecular dynamics

simulations of NaF, NaCl, NaBr, and NaI solutions. Large spheres: halide ions; small green spheres:

sodium cations. (E to H) Average density profiles of the corresponding solutions, ρ(z), relative to the

bulk value, ρb, of the water oxygen atom (blue), sodium cation (green), and halide ions (F, black; Cl,

yellow; Br, orange; I, magenta) as a function of the distance z normal to the plane of the interface.CR

ED

IT:A

DA

PT

ED

FR

OM

(4

)

www.sciencemag.org SCIENCE VOL 303 20 FEBRUARY 2004

P E R S P E C T I V E S

on

Janu

ary

28, 2

008

ww

w.s

cien

cem

ag.o

rgD

ownl

oade

d fr

om

1146

(thought to function as an internal model)(24) performed like normal subjects underboth the I- and M-conditions (16), suggest-ing that certain internal models may not beaccessible to awareness. If there exist sever-al internal models in the brain, what is theneural difference that makes some accessi-ble to awareness and others not?

The finding that there are distinguish-able brain states between the I- and M-con-ditions reinvigorates discussion about in-tentionality, but the final interpretation ofthese results lies in a thicket of furtherquestions. Most broadly, it remains to beunderstood how the neural events are relat-ed to the phenomenal experience that “I”was the author of an action. The internalmodel hypothesis suggests this relationshipmay be due to matching the consequencesof a movement against its internally pre-dicted effects. But predictability cannot bethe complete story, because people judge

the time of their own actions and the ac-tions of others equally well—but strangely,the “actions” of a nonbiological machineare judged quite differently, even whenthey are visually identical and equally pre-dictable (25). This suggests that intention-ality might even be judged retrospectively,an illusion arising from watching yourself(or another agent) make actions (26). Thisis consistent with the idea that you repre-sent the actions of others by analogy withyour own, inferring their intentions bywatching their actions. Thus, in contrast toMontaigne’s belief, it may be that action isthe judge of intention.

References1. B. Libet et al., Brain 106, 623 (1983).

2. P. Haggard, M. Eimer, Exp. Brain Res. 126, 128 (1999).

3. H. H. Kornhuber, L. Deecke, Pflügers Arch. 284, 1

(1965).

4. D. M. Eagleman, A. O. Holcombe, Trends Cogn. Sci. 6,

323 (2002).

5. B. Libet et al., Brain 106, 623 (1983).

6. H. C. Lau et al., Science 303, 1208 (2004).

7. S. Kastner, L. G. Ungerleider, Annu. Rev. Neurosci. 23,

315 (2000).

8. I. C. Griffin, A. C. Nobre, J. Cogn. Neurosci. 15, 1176

(2003).

9. F. Binkofski et al., J. Neurophysiol. 88, 514 (2002).

10. J. Rowe et al., Brain 125, 276 (2002).

11. R. Cunnington et al., Neuroimage 20, 404 (2003).

12. R. Cunnington et al., Neuroimage 15, 373 (2002).

13. D. Thaler et al., Exp. Brain Res. 102, 445 (1995).

14. M. Lotze et al., J. Cogn. Neurosci. 11, 491 (1999).

15. I. Fried et al., J. Neurosci. 11, 3656 (991).

16. A. Sirigu et al., Nature Neurosci. 7, 80 (2004).

17. J. B. Pochon et al., Cereb. Cortex 11, 260 (2001).18. T. Shallice, P. W. Burgess, Brain 114, 727 (1991).19. I. Toni et al., J. Cogn. Neurosci. 14, 769 (2002).20. M. Desmurget, S. Grafton, Trends Cogn. Sci. 4, 423

(2000).21. A. Sirigu et al., Brain 122, 1867 (1999).22. C. D. Frith et al., Brain Res. Brain Res. Rev. 31, 357

(2000).

23. D. M. Wolpert et al., Science 269, 1880 (1995).

24. S. J. Blakemore et al., Neuroreport 12, 1879 (2001).

25. A. Wohlschlager et al., Conscious Cogn. 12, 708

(2003).

26. E. van den Bos, M. Jeannerod, Cognition 85, 177

(2002).

In the conventional picture of simple saltsolutions, atomic ions shun the air/waterinterface and are more likely to be found

in the bulk of the liquid. Hence, simple inor-ganic salts such as sodium halides should berepelled from the water surface. However,recent computational and experimental stud-ies show that atomic ions such as halides canbe present in the surface region, in some cas-es even at enhanced concentrations. Halideions at the surfaces of atmospheric aerosolparticles may play an important role in con-trolling oxidant levels in the marine bound-ary layer of the atmosphere.

The conventional picture of the inter-face of simple aqueous salt solutions isbased on thermodynamic analysis of thevariation of surface tension with composi-tion of the liquid. Hu et al. were among thefirst to challenge this view (1). They ar-gued that chloride and bromide ions mustbe present at the air/water interface to ex-plain measurements of the uptake of Cl2and Br2 gases by aqueous salt solutions.More recent studies of reactions of oxi-dants with concentrated aqueous NaCl so-lutions support this view (2, 3). In thesestudies, surface reactions of ionic specieshad to be included to bring modeling andexperimental results into agreement.

Molecular simulations also support thispicture. In simulations of a 6 M aqueousNaCl solution, 10 to 15% of the accessiblesurface area was occupied by chloride ions,whereas sodium was effectively excludedfrom the topmost liquid layer (2). In theirsimulations of sodium halide solutions,Jungwirth and Tobias (4, 5) observed anincrease in surface concentration with in-creasing size and polarizability of thehalide ion. Thus, fluoride is depleted at theinterface, whereas bromide and iodide con-centrations are enhanced (see the figure).Calculations of the free energy of adsorp-tion also predict enhanced iodide concen-trations at the air/water interface (5, 6).

This picture is consistent with observa-tions of hydrogen bonding in aqueous ionicclusters. Cations form hydrated clusters inwhich the ion binds to water oxygen atoms.The water molecules are distributed fairlysymmetrically around the ion. In contrast,anions bind to water hydrogen atoms. Thewater molecules are arranged asymmetri-cally around the ion, enabling hydrogenbonding between them. This behavior isseen for the larger anions Cl–, Br–, and I–

(7). Hence, sodium cations should preferthe homogeneous environment in the bulkliquid, whereas large anions should formasymmetric structures near the interface—as predicted by the simulations.

Molecular simulations thus present apicture of ions at the interface that is con-sistent with cluster studies. However, the

simulations are sensitive to the descriptionof the molecular interactions they employ(6). Therefore, direct experimental obser-vations of molecular structure and energet-ics of ions in the interfacial region areneeded to corroborate the simulations.Such experiments are difficult to performbecause the liquid interface is disordered,dynamic, and small (typically only a fewmolecules wide) relative to the bulk.

Recent results from two laboratoriesshed light on this important issue (8–10).The authors have studied sodium andpotassium halide solutions with nonlinearspectroscopic techniques, such as secondharmonic generation (SHG) and vibra-tional sum-frequency spectroscopy (VSFS)(11). These techniques sample the surfaceregion of the liquid where isotropic sym-metry is broken. VSFS is a direct probe ofthe hydrogen-bonding environment in thesurface region, but only an indirect probeof the halide ions. SHG provides an esti-mate of the free energy of adsorption.

Raymond et al. (12) have used isotopicmixtures of water in VSFS studies to sepa-rate contributions from various vibrationalmodes. Similar studies on sodium halidesolutions (8) indicate that anions are pres-ent in the surface region but not at en-hanced concentrations. These halide ionsexhibit the same water structure-makingand structure-breaking behavior in the sur-face region as in the bulk. However, theydo not alter the hydrogen bonding of thewater in the topmost surface layer, nor dothey create the type of water structure in-dicative of a double layer formed by anion-cation separation.

Allen and co-workers (9) compareRaman and infrared spectra for bulk solu-

C H E M I S T RY

Ions at the Air/Water InterfaceBruce C. Garrett

The author is in the Chemical Sciences Division,Pacific Northwest National Laboratory (PNNL),Richland, WA 99352, USA. E-mail: bruce.garrett@pnl.gov

20 FEBRUARY 2004 VOL 303 SCIENCE www.sciencemag.org

P E R S P E C T I V E S

on

Feb

ruar

y 8,

201

1w

ww

.sci

ence

mag

.org

Dow

nloa

ded

from

1147

tions with surface vibrational spectra.They show that the depth of the interfa-cial region that contributes to the VSFSsignal is larger for the bromide and iodidesolutions than for neat water. This in-crease in interfacial depth supports theobservation from simulations that bro-mide and iodide segregate near the sur-face with a subsequent enhancement ofsodium below the anion layer. In addition,Saykally and co-workers have obtainedstrong evidence for the presence of thelarger halide anions at the liquid water

surface from resonant SHG experiments(10).

The results qualitatively support the com-putational predictions that chloride, bro-mide, and iodide are present in the surfaceregion. Testing the quantitative predictionsabout the concentrations of ions and theirproximity to the top of the surface regionwill require careful modeling and theoreticalanalysis of the surface vibrational spectra.

The combined evidence from simula-tions and experiments for the presence ofions at the air/water interface is com-

pelling. So what is wrong with the thermo-dynamic analysis that led to the conven-tional picture? The thermodynamic analy-sis assumes a binary system of water andsalt, with salt treated as one component.The simulations show that cations and an-ions behave differently and should be treat-ed as separate components. Furthermore, itis possible to have ions at the interface andstill have a net depletion in the surface re-gion, because ions may be depleted in thesecond layer. Finally, the same simulationsthat predict ions in the surface layer alsoreproduce the observed variations in sur-face tension with ion concentration (4),which are the basis of the conventional pic-ture of the interface.

The quantitative discrepancies betweenmolecular simulations and experiments re-main to be resolved. Other direct experi-mental probes of molecular structure at theliquid interface, such as extended x-ray ad-sorption fine structure (EXAFS) spectra(13), may provide new insight. Molecularsimulations must also be refined by in-cluding the direct use of electronic struc-ture methods. It is already clear, however,that the surface layers of many salt solu-tions contain ions that can participate inchemical processes.

The potential impact of interfacial ionson atmospheric chemistry is only begin-ning to be realized. Seinfeld (14) hashighlighted the importance of heteroge-neous chemistry in the atmosphere andthe role that surface ions may play. For ex-ample, there is now strong evidence thatreactions of surface chloride ions withOH radicals are a significant source ofmolecular chlorine (2), the precursor forchloride atoms that are important oxi-dants in marine environments.

References and Notes1. J. H. Hu et al., J. Phys. Chem. 99, 8768 (1995).

2. E. M. Knipping et al., Science 288, 301 (2000).

3. A. Laskin et al., Science 301, 340 (2003).

4. P. Jungwirth, D. J. Tobias, J. Phys. Chem. B 105, 10468

(2001).

5. P. Jungwirth, D. J. Tobias, J. Phys. Chem. B 106, 6361

(2002).

6. L. X. Dang, T. M. Chang, J. Phys. Chem. B 106, 235

(2002).

7. P. Ayotte et al., J. Phys. Chem. A 103, 10665 (1999).

8. E. A. Raymond, G. L. Richmond, J. Phys. Chem. B, in

press.

9. D. Liu, G. Ma, L. M. Levering, H. C. Allen, J. Phys. Chem.B 108, 2252 (2004).

10. R. Saykally, personal communication.

11. G. L. Richmond, Chem. Rev. 102, 2693 (2002).

12. E. A. Raymond, T. L. Tarbuck, M. G. Brown, G. L.

Richmond, J. Phys. Chem. B 107, 546 (2003).

13. K. R. Wilson et al., J. Chem. Phys. 117, 7738 (2002).

14. J. H. Seinfeld, Science 288, 285 (2000).

15. I thank G. Richmond, H. Allen, P. Jungwirth, D. Tobias,

and B. Finlayson-Pitts for helpful discussions; G.

Richmond and H. Allen for preprints of their unpub-

lished work; P. Jungwirth for providing the figure; and

the Office of Basic Energy Sciences, U.S. Department

of Energy (DOE), for support. PNNL is operated for

DOE by Battelle.

0.0 0.5 1.0 1.5 2.0 2.5 3.0

ρ (z) /ρb

z (Å

)z

(Å)

z (Å

)z

(Å)

D H

C G

B F

A E

0

4

8

12

16

0

4

8

12

16

0

4

8

12

16

0

4

8

12

16

NaF

NaBr

NaCl

NaI

The air/liquid interface of sodium halide solutions. (A to D) Snapshots from molecular dynamics

simulations of NaF, NaCl, NaBr, and NaI solutions. Large spheres: halide ions; small green spheres:

sodium cations. (E to H) Average density profiles of the corresponding solutions, ρ(z), relative to the

bulk value, ρb, of the water oxygen atom (blue), sodium cation (green), and halide ions (F, black; Cl,

yellow; Br, orange; I, magenta) as a function of the distance z normal to the plane of the interface.CR

ED

IT:A

DA

PT

ED

FR

OM

(4

)

www.sciencemag.org SCIENCE VOL 303 20 FEBRUARY 2004

P E R S P E C T I V E S

on

Feb

ruar

y 8,

201

1w

ww

.sci

ence

mag

.org

Dow

nloa

ded

from

The Carbon Dioxide-Water Interface at Conditions of GasHydrate Formation

Felix Lehmkuhler,*,† Michael Paulus,† Christian Sternemann,† Daniela Lietz,†

Federica Venturini,‡ Christian Gutt,§ and Metin Tolan†

Fakultat Physik/DELTA, Technische UniVersitat Dortmund, Maria-Goeppert-Mayer-Strasse 2,44221 Dortmund, Germany, ESRF, BP 220, 38043 Grenoble, France, and HAYSLAB at DESY,

Notkestrasse 85, 22607 Hamburg, Germany

Received August 7, 2008; E-mail: Felix.Lehmkuehler@tu-dortmund.de

Abstract: The structure of the carbon dioxide-water interface was analyzed by X-ray diffraction andreflectivity at temperature and pressure conditions which allow the formation of gas hydrate. Thewater-gaseous CO2 and the water-liquid CO2 interface were examined. The two interfaces show a verydifferent behavior with respect to the formation of gas hydrate. While the liquid-gas interface exhibits theformation of thin liquid CO2 layers on the water surface, the formation of small clusters of gas hydrate wasobserved at the liquid-liquid interface. The data obtained from both interfaces points to a gas hydrateformation process which may be explained by the so-called local structuring hypothesis.

Introduction

Beyond the three classical states of matter, water can forman additional solid phase called clathrate hydrate with theauxiliary presence of gas molecules, typically at low tempera-tures and high pressures.1,2 In these crystalline structures thegas molecules are trapped in a hydrogen bond water cagenetwork. During the past years, hydrates have become veryimportant materials since they may become essential for futureenergy recovery or hydrogen and CO2 storage.2-6

The hydrate formation process is well understood from amacroscopic thermodynamical point of view.7 Nevertheless, theformation on a microscopic level is still not clear. In theliterature, three competing formation models are presented.Within the framework of the cluster nucleation theory by Sloan,1

gas molecules dissolve in water, and labile clusters are formedand agglomerate, preferentially next to the water surface. Aftera critical cluster size is achieved, the macroscopic nucleationsets in. A different model resulting from molecular dynamics(MD) simulations of the water-liquid CO2 interface is the localstructuring hypothesis.8 Here the water and gas moleculesarrange stochastically until an arrangement similar to the hydratephase is reached. This stochastically achieved network is

stabilized after exceeding a certain size and hydrate crystals startgrowing. In the third model introduced by Rodger9 a surface-driven formation is proposed. Gas molecules adsorb on the watersurface and are trapped in the center of partially completed watercavities. Kvamme10,11 extended this model based on results ofMD simulations and predicted the initial nucleation at thewater-CO2 interface. Due to wave motion of the water surfacethe mixing of water and gas molecules is supported, and hydratefragments can be formed. The cluster nucleation theory and thesurface-driven model predict the appearance of hydrate prestagesat the water surface. In contrast, the local structuring hypothesispredicts a spontaneous formation without any precursor clusters.A comparison of the local structuring hypothesis and the clusternucleation theory is presented in Figure 1. Due to their similarity,cluster nucleation theory and surface-driven model are merged.The contrast between presence and absence of prehydratestructures is clearly visible.

Several MD-simulations which have been performed duringthe past years have not been able to prove one of these theoriesexplicitly. Some of the simulations support the local structuringhypothesis12-14 while others find a labile cluster formation,15,16

or at least an indication of hydrate precursor.17 It is the aim ofthis paper to solve this question from the experimental point ofview.† Technische Universitat Dortmund.

‡ ESRF, Grenoble.§ HASYLAB, Hamburg.

(1) Sloan, E. D.; Koh, C. A. Clathrate Hydrates of Natural Gases; CRCPress Inc.: Boca Ranton, 2007.

(2) Sloan, E. D. Nature 2003, 426, 353–359.(3) Florusse, L. J.; Peters, C. J.; Schoonman, J.; Hester, K. C.; Koh, C. A.;

Dec, S. F.; Marsh, K. N.; Sloan, E. D. Science 2004, 306, 469–471.(4) Lee, H.; Lee, J.; Kim, D. Y.; Park, J.; Seo, Y.; Zeng, H.; Moudrakovski,

I. L.; Ratcliffe, C. I.; Ripmeester, J. A. Nature 2005, 434, 743–746.(5) Brewer, P. G.; Friederich, G.; Peltzer, E. T.; Orr, F. M, Jr Science

1999, 284, 943–945.(6) Kang, S. P.; Lee, H. EnViron. Sci. Technol. 2000, 34, 4397–4400.(7) van der Waals, J. H.; Platteeuw, J. C. AdV. Chem. Phys. 1959, 2, 1–

57.(8) Radhakrishnan, R.; Trout, B. L. J. Chem. Phys. 2002, 117, 1786–

1796.

(9) Rodger, P. M. J. Phys. Chem. 1990, 94, 6080–6089.(10) Kvamme, B. Ann. N.Y. Acad. Sci. 2000, 912, 496–501.(11) Kvamme, B. Initiation and growth of hydrate from nucleation theory;

Proceedings of the International Symposium on Deep Sea Sequestra-tion of CO2, 2000, 1-1-1.

(12) Hirai, S.; Okazaki, K.; Tabe, Y.; Kawamura, K. Energy ConVers.Manage. 1997, 38, S301–S306.

(13) Moon, C.; Taylor, P. C.; Rodger, P. M. J. Am. Chem. Soc. 2003, 125,4706–4707.

(14) Guo, G. J.; Zhang, Y. G.; Liu, H. J. Phys. Chem. C 2007, 111, 2595–2606.

(15) Long, J.; Sloan, E. D. Mol. Simul. 1993, 11, 145–161.(16) Guo, G. J.; Zhang, Y. G.; Zhao, Y. J.; Refson, K.; Shan, G. H. J. Chem.

Phys. 2004, 121, 1542–1547.

Published on Web 12/23/2008

10.1021/ja806211r CCC: $40.75 2009 American Chemical Society J. AM. CHEM. SOC. 2009, 131, 585–589 9 585

8

An introduction to the science and energy potential of a unique resource

Energy Resource Potential of Methane Hydrate

Ice That Burns

What is Methane Hydrate?A clathrate is a chemical compound in which molecules of one material (the “host”) form a solid lattice that encloses molecules of another material (the “guest”). Methane hydrate is a naturally-occurring clathrate in which a host lattice of water-ice encloses guest molecules of methane. Methane, made of one carbon atom and four hydrogen atoms, is the simplest hydrocarbon molecule and the primary component of natural gas.

In methane hydrate, the gas molecules are not chemically bound to the water molecules but instead are trapped within their crystalline lattice. The resulting substance looks remarkably like white ice, but it does not behave like ice. When methane hydrate is “melted,” or exposed to pressure and temperature conditions outside those where it is stable, the solid crystalline lattice turns to liquid water, and the enclosed methane molecules are released as gas. This process, called dissociation, can be demonstrated by lighting a match next to a piece of methane hydrate; the heat from the match will cause the hydrate to dissociate, and the methane molecules will be ignited as they are released. This results in the curious spectacle of what appears to be burning ice.

Methane hydrate is a material very much tied to its environment—it requires very specific conditions to form and remain stable. Pressure, temperature, and availability of sufficient quantities of water and methane are the primary factors controlling methane hydrate formation and stability, although geochemistry and the type of sediment also play a part. If the pressure and temperature are just right, free methane gas and water will form solid methane hydrate.

The adjacent graph is a phase diagram showing the pressure and temperature ranges where methane hydrate is stable. The horizontal axis shows temperature, increasing from left to right, and the vertical axis shows depth of burial, increasing from top to bottom. Because fluid pressure increases with depth below the surface of the earth or the ocean, depth serves as a proxy for fluid pressure in hydrate phase diagrams. The curved line between the blue and yellow areas is the methane hydrate phase boundary. Above this boundary, temperatures are too warm, and pressures are too low for methane hydrate to form, so methane can only be present as a gas. Below this boundary, solid methane hydrate is able to form and remain stable, because temperatures are sufficiently low, and fluid pressures are sufficiently high to sustain the solid phase.

Methane hydrate dissociating with the methane ignited – “burning ice.” (Courtesy of National Research Council Canada)

Methane hydrate stability diagram.

Model of a methane molecule enclosed in water-molecule “cage.”

9

An introduction to the science and energy potential of a unique resource

Energy Resource Potential of Methane Hydrate

Other factors can affect the stability of methane hydrate. For example, higher salt content in pore water within sediments can restrict hydrate formation, just like road salt keeps ice from forming on highways. Elevated salinity has the effect of shifting the hydrate phase boundary to the left—essentially requiring colder temperatures to form hydrate. Similarly, the presence of small amounts of gases other than methane, such as carbon dioxide (CO2), hydrogen sulfide (H2S), or heavier hydrocarbons such as ethane (C2H6) can act to shift the phase boundary to the right, so that hydrate can form at higher temperatures.

Because methane hydrate can only remain solid at low temperatures and high pressures, it is difficult to recover methane hydrate samples intact, whether the samples are collected from the seafloor or from deeply buried sediments. As soon as a sample is brought to the Earth’s surface, it will follow a pressure-temperature path on the graph that is upwards and towards the right across the phase boundary. Dissociation of the hydrate into water and methane will occur, unless the sample is maintained during retrieval, or quickly pressurized or refrigerated after retrieval to keep it within the hydrate stability envelope. Methane hydrate is a fairly concentrated form of natural gas. When dissociated at normal surface temperature and pressure, one cubic foot of solid methane hydrate will release about 164 cubic feet of methane gas. This is one of the reasons people are interested in methane hydrate as a potential source of methane for energy supply.

Where is Methane Hydrate Found?Methane hydrate is known to occur in both terrestrial and marine environments. Terrestrial deposits have been found in polar regions, hosted in sediments within and beneath the permafrost, while marine occurrences have been found mainly in sediments of the Earth’s outer continental margins. These are the natural settings where methane and water are present, and where pressure and temperature conditions are suitable to form and sustain hydrate.

Location of sampled and inferred gas hydrate occurrences worldwide. (Map courtesy of Timothy S. Collett, USGS)

Cl th t h d t f ti d l

• Cluster Nucleation theory

Clathrate hydrate formation models

y

• Local structuring hypothesis

• Surface-driven formation

The cluster nucleation theory and the surface-driven model predict the appearance of hydrate prestages at the water surface. In contrast, the local structuring hypothesis predicts a spontaneous formation without any precursor clusters.

http://classes.engineering.wustl.edu/eece534/

Classical Nucleation Theory

Nucleation rate (Jn assuming spherical particles) is expressed as:

gAJ

Interfacial free energy Bulk free energy interfacial energy

A

TkgAexpJ

Bn

r 2 s

IAP supersaturationK

volume per molecule

g=4r2-4/3r3(/)

rlnc

Bk T

3BAe pJ

c

A kinetic factorr critical nucleus sizeT temperature

g rc

n 2AexpJ

B

B

k Boltzmann constantk Tln

r

Our experimental approach allows measuring statistical averages of nucleation andgrowth rates from mineral surfaces on the order of 1.0 cm2. In order to obtaininterfacial energies (), the molar volume (; included in B) can be determinedi d d tl b id tif i th t f th l t d h i AFM Xindependently by identifying the nature of the nucleated phase using AFM, X-rayabsorption, X-ray and Neutron Pair Distribution Function analysis, andspectroscopic techniques such as Raman spectroscopy.

A lot of experimental attempts to study the hydrate formationprocess have been performed in the past. Within most of thesestudies the hydrate formation is observed as an interfacialphenomenon. However, the methods used, such as opticalobservation or calorimetric measurements, only give informationabout the formation after the initial nucleation has already takenplace. Takeya et al. reported a formation on a macroscopic levelat a liquid-liquid interface where the initial nucleation startsat the water surface.18 According to this study, needle-likehydrate clusters grow toward the water bulk. The hydrateformation from water and gaseous CO2 was reported by Morganand co-workers who observed the growing hydrate crystalsoptically and by means of calorimetric methods.19 Here, thegaseous CO2 was bubbled through the water bulk for severalminutes to induce the hydrate formation. A different approachhas been followed by Ohmura et al.20 In order to observemethane hydrate formation the authors used water spraying and,by optical observation, found a hydrate formation that starts atthe water surface. As stated before, neither hydrate prestructureformation nor the initial nucleation can be observed with theuse of these methods. Koh and co-workers showed time-resolvedmacroscopic hydrate formation in the stirred water-CO2 bulkmixture by means of X-ray diffraction.21 The authors were notable to favor one hydrate formation model due to measurementsof a disturbed system which offers seeds for macroscopichydrate formation, but they proposed a dynamic intermediatephase during the hydrate formation and growth process. Ingeneral, the quantity measured in all of these experiments isthe induction time ti for hydrate nucleation. This induction timedepends on the kind of guest molecules, the degree ofsupercooling, and the history of the water-gas system.1,18 ForCO2 the induction time is on the order of ti ≈ 2h. Thus, so farthere are no data regarding the gas hydrate formation process

at nanometer length scales for the early state before themacroscopic formation begins. In this work the first study ofthe CO2-water interface on molecular length scales is presented.

Experiment and Discussion

Liquid-Gas Interface. The prestage formation at or near thesurface predicted by the cluster nucleation theory and thesurface-driven model will cause an increased surface roughnessand the appearance of layers of incomplete hydrate cages atthe surface. These changes in surface roughness and theformation of thin layers are observable by X-ray reflectivitymeasurements.

X-ray reflectivity is a powerful tool to investigate theformation of very thin layers on solid and liquid surfaces withångstrom resolution.22 Its great advantage compared to othermethods is the possibility to identify in situ surface roughnesschanges of molecular thin films from the reflectivity measure-ments. In such experiments the wave vector transfer has onlyone component perpendicular to the surface given by

qz ) (4π) ⁄ (λ) sin(θ)

where θ denotes the angle between the sample surface and theX-ray beam, and λ the wavelength of the incident beam.

The scattered intensity R is given by22,23

R(qz))RF(qz) · | 1F∞∫ dFe(z)

dzeiqzz dz|2 (1)

with RF the Fresnel reflectivity of a smooth surface and F∞ theaverage density of the entire sample. Thus, X-ray reflectivityyields the laterally averaged density profile perpendicular to thesample’s surface.

Pressure-dependent X-ray reflectivity measurements of thewater-CO2 interface were performed at the European Synchro-tron Radiation Facility (ESRF) using the high-energy setup forliquid surfaces at beamline ID15A.24,25 CO2 and water formthe cubic structure I (sI) hydrate with a lattice constant of 12Å.2 At T ) 0 °C a minimum pressure of p ) 12.5 bar is

(17) Zhang, J.; Hawtin, R. W.; Yang, Y.; Nakagava, E.; Rivero, M.; Choi,S. K.; Rodger, P. M. J. Phys. Chem. B 2008, 112, 10608–10618.

(18) Takeya, S.; Hori, A.; Hondoh, T.; Uchida, T. J. Phys. Chem. B 2000,104, 4164–4168.

(19) Morgan, J. J.; Blackwell, V. R.; Johnson, D. E.; Spencer, D. F.; North,W. J. EnViron. Sci. Technol. 1999, 33, 1448–1452.

(20) Ohmura, R.; Kashiwazaki, S.; Shiota, S.; Tsuji, H.; Mori, Y. H. EnergyFuels 2002, 16, 1141–1147.

(21) Koh, C. A.; Savidge, J. L.; Tang, C. C. J. Phys. Chem. 1996, 100,6412–6414.

(22) Tolan, M. X-ray Scattering from Soft Matter Thin Film; Springer:Berlin, 1999.

(23) Als-Nielsen, J.; McMorrow, D. Elements of Modern X-Ray Physics;John Wiley & Sons: New York, 2001.

Figure 1. Comparison of the cluster nucleation theory (a-b-c-e) and the local structuring hypothesis (a-d-e). (a) Water without dissolved gas molecules(initial condition). (b) Cluster form immediately after dissolution of gas molecules. (c) Cluster prestages agglomerate by sharing faces. These agglomeratedclusters may be unstable (step back to b is possible). (d) No cluster formation after dissolution of gas molecules. (e) Hydrate nucleation.

586 J. AM. CHEM. SOC. 9 VOL. 131, NO. 2, 2009

A R T I C L E S Lehmkuhler et al.

necessary for a stable CO2 hydrate structure.1 Ultrapure waterfiltered by a Millipore apparatus, and CO2 purchased from AirLiquide with a purity of 99.998% was employed. The photonenergy was E ) 72.5 keV. A pressure cell with aluminumwindows was utilized to measure at gas pressures up to 35 bar.The temperature was chosen to be 0.15 °C, therefore avoidingthe formation of ice and controlled with an accuracy of 0.01°C. A few mm thick water film was prepared in the samplecell, forming a well-defined meniscus. CO2 was allowed to fillthe cell via an inlet. Following each reflectivity measurement,the detector was moved to an out-of-plane position in order tomeasure the diffusely scattered intensity that is necessary foran accurate background subtraction. Measurements for differentgas pressures between 1 bar and the condensation pressure of35 bar were carried out. For pressures above 33 bar the timedependence of the reflectivity curves was investigated. For agiven pressure, this was done by performing several measure-ments within a time interval of more than 8 h. During this periodthe sample was not stirred or disturbed in other ways to favorhydrate formation. Thus, it should be possible to detect hydrateprestructures without offering nucleation seeds. Also surface-sensitive diffraction measurements were performed to detectBragg reflections originating from possible CO2 hydratecrystallites.

Reflectivities for different gas pressures are presented inFigure 2. The oscillations indicate the formation of thin layerson the water surface. The curves are fitted using the effectivedensity model22 based on Parratt’s algorithm.26 Refined electrondensity profiles are presented in Figure 3. The measurement at1 bar CO2 pressure yields a roughness of the water surface ofσ ) (3.2 ( 0.1) Å, which is in good agreement with capillarywave theory.27 The measured water roughness does not changewith increasing gas pressure. Most importantly, this is also thecase when the pressure range where gas hydrate formationbecomes possible was reached. An electron density of FL )(0.282 ( 0.008) e/Å3 was found for the observed layers andfits very well with the tabulated value FCO2 ) 0.279 e/Å3 forliquid CO2. This suggests the adsorption of gas moleculesinstead of hydrate formation because the electron density of

CO2 hydrate is approximately 30% higher compared to that ofliquid CO2.

28 The roughness of this CO2 layer ranges from 3 Åat low gas pressures to 11 Å at pressures near the condensationpressure. In order to explain the data, capillary wave andadsorption theory can be applied. The adsorption of gasmolecules on the liquid surface as a function of gas pressure isa result of van der Waals interactions between the liquid andthe vapor phase. It is usually described by an adsorption isothermin the framework of the Frenkel-Halsey-Hill (FHH) theory.29

The thickness lm of the adsorbed layer can be calculated viathe free energy F of the liquid-gas system yielding29

lm ) ( Aeff

6π∆FkBTln(p ⁄ p0))1

3 (2)

Here, Aeff denotes the effective Hamaker constant, ∆F is thedensity difference between the adsorbed layer and the gas phase,and p0 is the condensation pressure of the gas at the fixedtemperature T.

An adsorption isotherm can be fitted to the layer thicknessdetermined from the measured data. A good agreement betweenexperiment and calculation is achieved (see Figure 4). Thissuggests again the absence of gas hydrate structures. Besides,the inset in Figure 4 shows the layer roughness as a function oflayer thickness. The agreement of experimental results andtheory for adsorbed liquid films29 is again very good. Owing tothe good interpretation of the data, only the formation ofadsorbed liquid CO2 layers can be observed, but no indicationof hydrate or hydrate prestage formation after more than 8 hwas found. This is in qualitative agreement with measurementsat the water-propane interface.30 Furthermore, no Braggreflections could be observed by X-ray diffraction at thisinterface. Therefore, the formation of hydrate crystallites mustbe very weak s if there is any.

Liquid-Liquid Interface. As no hydrate formation at thewater-gaseous CO2 interface was visible, the pressure wasraised to a value above the condensation pressure of CO2. Thus,a macroscopic liquid CO2 layer with a thickness of severalmillimeters is formed at the water surface. In order to observe

(24) Reichert, H.; Honkimaki, V; Snigirev, A.; Engemann, S.; Dosch, H.Physica B 2003, 336, 46–55.

(25) Honkimaki, V; Reichert, H.; Okasinski, J. S.; Dosch, H. J. SynchrotronRadiat. 2006, 13, 426–431.

(26) Parratt, L. G. Phys. ReV. Lett. 1954, 95, 359–369.(27) Braslau, A.; Pershan, P. S.; Swislow, G.; Ocko, B. M.; Als-Nielsen,

J. Phys. ReV. A 1988, 38, 2457–2470.

(28) Uchida, T.; Ebinuma, T.; Narita, H. Laboratory studies on the formationand dissociation processes of CO2-hydrate crystals; Proceedings ofthe International Symposium on Deep Sea Sequestration of CO2, 2000,1-4-1.

(29) Paulus, M.; Gutt, C.; Tolan, M. Phys. ReV. E 2005, 72, 061601.(30) Paulus, M.; Gutt, C.; Tolan, M. Surf. Interface Anal. 2008, 40, 1226–

1230.

Figure 2. Reflectivities normalized by the Fresnel reflectivity for selectedCO2 pressures. Solid lines represent fits. The curves are shifted for clarity.

Figure 3. Electron density profiles corresponding to the refinements ofFigure 2 for different gas pressures.

J. AM. CHEM. SOC. 9 VOL. 131, NO. 2, 2009 587

Molecular Level Gas Hydrate Formation at H2O-CO2 Interface A R T I C L E S

the formation of gas hydrate, X-ray diffraction measurementswere performed at this liquid-liquid interface. The studiedinterface area was about 1 mm2. Bragg reflections of CO2

hydrate were observed. These reflections occur and disappearcontinuously. Figure 5 shows a time scan at the detector positionof the CO2 hydrate (321) Bragg reflection. The intensity stronglyvaries with time indicating fluctuations of the hydrate crystallitesat the interface.

In order to investigate this hydrate formation more precisely,additional diffraction patterns were measured using the wide-angle diffraction setup with a MAR345 image plate detector atBL9 of the DELTA synchrotron source.32 A slightly adaptedsample cell with an inner diameter of 8 cm was used toinvestigate the hydrate formation in an interfacial area of about80 mm2. This allows measuring the formation of crystallites atthe interface with spatial resolution. After condensing amacroscopic thick liquid CO2 film, Bragg reflections wereimmediately observable (see Figure 6). More scans after severalminutes show a dynamic behavior of the formed hydrateclusters. The dynamics is still observable after about 100 min,and thus, no macroscopic freezing can be observed on this time

scale. By applying the Scherrer equation, a crystal size ofapproximately 200 Å can be estimated. Thus, the local formationof mobile hydrate crystallites which are moving freely at theinterface was observed. A macroscopic formation, which wouldbe observable by calorimetric measurements18 or would showa freezing in of the dynamics at the surface, was not visible.

Conclusion

Due to the instant formation of hydrate crystallites at theliquid-liquid interface, its stochastic nature, and the absenceof any hydrate prestructures at the gas-liquid interface, the localstructuring hypothesis is favored by our study. The formationon locally limited areas points to a rather stochastic process incontrast to the cluster models proposed by Sloan, Rodger, andKvamme.1,2,9-11 Hydrate layers or predicted prestructures whichshould appear at the water-gaseous CO2 interface are notobservable. The local formation of mobile CO2 hydrate crys-tallites at the liquid-liquid interface and their size of ap-proximately 200 Å suggests that the gas amount at thewater-gas interface is too low for any hydrate formation, evenin presence of an adsorbed CO2 layer with a thickness up to 40Å. Furthermore, these layers are disturbed by capillary wavefluctuations which may inhibit the formation of hydrates. Stablecrystallites are formed in presence of a thick layer where theregion of hydrate formation is less influenced by these fluctua-tions. Owing to the size of the crystallites which is about 1order of magnitude above the expected nucleation size,8 it canbe deduced that an adsorbed layer with a thickness of at least200 Å is necessary for hydrate formation at the water-gasinterface.

Summary

In summary, the water-gas interface shows in comparisonto the liquid-liquid interface a very different behavior withrespect to the gas hydrate formation process. The adsorption ofgas molecules on the water surface leads to a high supply ofCO2 at the water surface but does not trigger the gas hydrateformation process. No hydrate formation could be observed formore than 8 h. In contrast we were able to observe formationof CO2 hydrate at the water-liquid CO2 interface. The presenceof a macroscopic amount of liquid CO2 induces the localformation of mobile CO2 hydrate crystallites. Due to this finding,i.e. the stochastic nature of crystallite formation and the absenceof surface covering prestructures at the interfaces, the local

(31) National Institute for Standards and Technology Chemistry WebBook;http://webbook.nist.gov/chemistry.

(32) Krywka, C.; Sternemann, C.; Paulus, M.; Javid, N.; Winter, R.; Al-Sawalmih, A.; Yi, S.; Raabe, M.; Tolan, M. J. Synchrotron Radiat.2007, 14, 244–251.

Figure 4. Adsorption isotherm, solid line represents the fitted curve, p0 )34.99 bar.31 (Inset) Surface roughness of the adsorbed CO2 layer as functionof layer thickness. Solid line represents a calculation within the so-calledanharmonic approximation.29

Figure 5. Timescan of the (321) Bragg reflection of CO2 hydrate measuredat the water-liquid CO2 interface.

Figure 6. Diffraction images showing the region where the (321) reflectionis observed. The intensity scale (arbitrary units) is presented right. (a) Afterfilling with CO2. (b) After 60 min. (c) 95 min. (d) 97 min. A detailedrepresentation of the time effects on the intensity of the (321) Braggreflection is presented in Figure 5.

588 J. AM. CHEM. SOC. 9 VOL. 131, NO. 2, 2009

A R T I C L E S Lehmkuhler et al.

the formation of gas hydrate, X-ray diffraction measurementswere performed at this liquid-liquid interface. The studiedinterface area was about 1 mm2. Bragg reflections of CO2

hydrate were observed. These reflections occur and disappearcontinuously. Figure 5 shows a time scan at the detector positionof the CO2 hydrate (321) Bragg reflection. The intensity stronglyvaries with time indicating fluctuations of the hydrate crystallitesat the interface.

In order to investigate this hydrate formation more precisely,additional diffraction patterns were measured using the wide-angle diffraction setup with a MAR345 image plate detector atBL9 of the DELTA synchrotron source.32 A slightly adaptedsample cell with an inner diameter of 8 cm was used toinvestigate the hydrate formation in an interfacial area of about80 mm2. This allows measuring the formation of crystallites atthe interface with spatial resolution. After condensing amacroscopic thick liquid CO2 film, Bragg reflections wereimmediately observable (see Figure 6). More scans after severalminutes show a dynamic behavior of the formed hydrateclusters. The dynamics is still observable after about 100 min,and thus, no macroscopic freezing can be observed on this time

scale. By applying the Scherrer equation, a crystal size ofapproximately 200 Å can be estimated. Thus, the local formationof mobile hydrate crystallites which are moving freely at theinterface was observed. A macroscopic formation, which wouldbe observable by calorimetric measurements18 or would showa freezing in of the dynamics at the surface, was not visible.

Conclusion

Due to the instant formation of hydrate crystallites at theliquid-liquid interface, its stochastic nature, and the absenceof any hydrate prestructures at the gas-liquid interface, the localstructuring hypothesis is favored by our study. The formationon locally limited areas points to a rather stochastic process incontrast to the cluster models proposed by Sloan, Rodger, andKvamme.1,2,9-11 Hydrate layers or predicted prestructures whichshould appear at the water-gaseous CO2 interface are notobservable. The local formation of mobile CO2 hydrate crys-tallites at the liquid-liquid interface and their size of ap-proximately 200 Å suggests that the gas amount at thewater-gas interface is too low for any hydrate formation, evenin presence of an adsorbed CO2 layer with a thickness up to 40Å. Furthermore, these layers are disturbed by capillary wavefluctuations which may inhibit the formation of hydrates. Stablecrystallites are formed in presence of a thick layer where theregion of hydrate formation is less influenced by these fluctua-tions. Owing to the size of the crystallites which is about 1order of magnitude above the expected nucleation size,8 it canbe deduced that an adsorbed layer with a thickness of at least200 Å is necessary for hydrate formation at the water-gasinterface.

Summary

In summary, the water-gas interface shows in comparisonto the liquid-liquid interface a very different behavior withrespect to the gas hydrate formation process. The adsorption ofgas molecules on the water surface leads to a high supply ofCO2 at the water surface but does not trigger the gas hydrateformation process. No hydrate formation could be observed formore than 8 h. In contrast we were able to observe formationof CO2 hydrate at the water-liquid CO2 interface. The presenceof a macroscopic amount of liquid CO2 induces the localformation of mobile CO2 hydrate crystallites. Due to this finding,i.e. the stochastic nature of crystallite formation and the absenceof surface covering prestructures at the interfaces, the local

(31) National Institute for Standards and Technology Chemistry WebBook;http://webbook.nist.gov/chemistry.

(32) Krywka, C.; Sternemann, C.; Paulus, M.; Javid, N.; Winter, R.; Al-Sawalmih, A.; Yi, S.; Raabe, M.; Tolan, M. J. Synchrotron Radiat.2007, 14, 244–251.

Figure 4. Adsorption isotherm, solid line represents the fitted curve, p0 )34.99 bar.31 (Inset) Surface roughness of the adsorbed CO2 layer as functionof layer thickness. Solid line represents a calculation within the so-calledanharmonic approximation.29

Figure 5. Timescan of the (321) Bragg reflection of CO2 hydrate measuredat the water-liquid CO2 interface.

Figure 6. Diffraction images showing the region where the (321) reflectionis observed. The intensity scale (arbitrary units) is presented right. (a) Afterfilling with CO2. (b) After 60 min. (c) 95 min. (d) 97 min. A detailedrepresentation of the time effects on the intensity of the (321) Braggreflection is presented in Figure 5.

588 J. AM. CHEM. SOC. 9 VOL. 131, NO. 2, 2009

A R T I C L E S Lehmkuhler et al.

Small Angle X-ray Scattering (SAXS): Transmission Mode

4 2sin dqD

π πθλ λ

= =

Beam

Fluid Cell

Pre-Ion Chamber

Post-Ion Chamber

2θ

Detector

Sample

Sample to detector distance (D) defines q range for given λ

d = pixel size

From Lazzari, 2006

P(q) = individual particle scatteringdepends on the shape of the particle

S(q) = interparticle scattering, “Structure Factor”

given by relative arrangement of particles(i.e., FT of the pair correlation function)

I(q) = npS(q)P(q)

Dilute solutions:

I(q) = npP(q)

Concentrated solutions or aggregates:

x-rays x-rays

P(q) = F2(q) F(q) = Form Factor”

np = number density of particles

S(q) =1+NV

4πr2 g(r) −1[ ]sin(qr)qr

dr∫

X-ray Scattering

How is the scattering intensity generated?

Different Particle Sizes: Related to Form Factor

From Kline, 2000

Different Concentrations:Related to Structure Factor

From Kline, 2000

and

: the case of having different particle sizes

Considering hard sphere model, electrostatic repulsion, or fractal structure

How to interpret 2D image from SAXS2D image (dark and blank subtracted)

1.0E+02

1.0E+03

1.0E+04

1.0E+05

0.1 1

q (/nm)

Inte

nsity

1D data

Data reduction:Angular integration

We have two choices: 1. Using Local Monodisperse Approximation, fit this 1D data with fitting program (SAXSConvert) get mean particle size (ro) and σ get size distribution using log-normal, Gaussian, Schulz distribution function.

2. Using Guinier Approximation (the scatters are sufficiently dilute)I(q) ≈I(0)exp(-q2Rg

2/3), Ln I(q) = Ln I(0)- q2Rg2/3, Rg: radius of gyration

For uniform sphere: r2 = 5/3Rg2 Using Svergun’s program (GNOM using IFT method)

for particle size distribution.