WATER AT POLYMER

INTERFACES Bryan D. Vogt

University of Akron

Polymer Engineering

Environment impacts adhesive strength

RH = 78 % RH = 19 % Adhesive

Joint

No problems Joint failure

What are the origins of this adhesion failure at a molecular level?

Can we engineer strategies to overcome this?

Why worry about adhesives?

More than 50,000 rivets to hold together

Reduce rivets by replacing with adhesive

Decrease maintenance costs!

Changing environmental conditions – how does this effect the adhesive strength?

Interfaces and composite systems

Highly filled composite polymer

systems intrinsically high

interfacial area

Problem: disordered system with

multiple length scales

How to obtain fundamental

understanding of moisture in these

systems?

Idealized interface = thin film

filler

binder

Adhesion measurements using shaft loaded blister test

4

16

a

wEhG

displacement-based equation (w)

P

w

2a

Eh: Film Stiffness (modulus, E, · thickness, h)

But this is a bulk measurement, how to understand mechanisms?

Moisture influence: Bulk versus interface contributions

Dry

Moisture exposure Bulk swelling

Interfacial water accumulation

Bulk swelling

Decreased cohesive strength

Stress development in polymer

Interfacial moisture

Decreased contact area

Stress development in polymer

Which effect controls adhesive failure?

Contrast control through isotopic substitution

Neutron Reflectivity (NR)

• Isotopic sensitivity (1H vs 2H)

• Measure water distribution within

film

Water ‘looks’ like polymer

(similar density)

Water visible

(Heavy water, D2O)

X-ray Reflectivity (XR)

• Measure thickness change due to

moisture absorption

• Mass density profile

Use NR to directly observe water distribution in film

substrate

film

X-rays or

neutrons

r

D

roughness

-5

-4

-3

-2

-1

0

log (

R)

0.200.150.100.05

Q (Å-1

)

Qc2 r

D @ 2p/DQ

fringe persistence surface roughness

Information:

• thickness D

• density profile, Qc2 , r(z)

• interfacial roughness

Q

Air or D2O

Quantifying moisture distribution: X-ray and neutron reflectivity

Examining water at polymer interface Thick polyimide on silicon wafer

Si

Silicon oxide

Polymer

OH

OH

OH

OH

OH

OH

OH

OH

OH

OH

OH

polymer sili

con

Wu et al. Polymer Engineering and Science, 1995, 35, 1000.

D2O Qc2 = 3.21 x 10-4 Å-2

Qc2 = 1.57 x 10-4 Å-2

Qc2 = 1.06 x 10-4 Å-2

Qc2 ≅ 1.5 x 10-4 Å-2

h = 2 mm

h = 2 nm

h= 2 mm

No well defined interferences due to relative thicknesses

Shift in critical edge and change in decay only differences

Fit suggests D2O accumulates at interface

Polyhydroxystryrene on silicon wafer

-8

-6

-4

-2

0

log

R

0.200.150.100.05Q (Å

-1)

200

150

100

50

0

Qc

2 (

Å-2

x10

-6 )

200150100500Distance (Å)

Si

Silicon oxide

Polymer

OH

OH

OH

OH

OH

OH

OH

OH

OH

OH

OH

Qc2 = 0.79 x 10-4 Å-2

h = 80 nm

Qc2 = 1.6 x 10-4 Å-2

h = 2 nm

Qc2 = 1.06 x 10-4 Å-2

h= 2 mm

Dry

wet

Minima in reflectivity not fit

without excess at interface

Improved sensitivity to

interface by decreasing film

thickness

Vogt et al., Langmuir 2004, 20, 5285.

Effect of polymer on interfacial concentration

0.30

0.25

0.20

0.15

0.10

0.05

0.00

Vo

lum

e F

ractio

n D

2O

12080400

Distance from Silicon Substrate (Å)

PHOSt

PBOCStConcentration profile near substrate

independent of polymer

J. Mater. Sci. 1996 31 927-927

Langmuir 2004 20(4) 1453-1458

J. Microlitho. Microfab. 2005 4(1) 013003

Native oxide Convert SLD profile to vol. frac.

ox ide/polym er interface di ffi cul t, but the qual i tati vepresence of the interfacial w ater can be explained ratio-nal l y . I n the case of si l i con oxide, the substrate is attractivetow ard w ater. T he chem ical potential of the system canbe w ri tten as a sum mation of external potential s (surface)and internal potentials (intr insic).

Far from the surface, the external potential contr ibutionis negl i gi ble. T his resul ts in bulk l i ke sw el l ing of thepolym er fi lm , as the total chem ical potential i s internal .At the interface, addi tional contr i butions to the chem icalpotential from the substrate are im portant. T hese includecontr i butions from van der Waals and electrostaticinteractions, al though other forces such as hydration arealso im portant. T he attracti ve nature of si l i con ox ide forw ater resul ts in a posi tive external contribution. T his leadsto an apparent increase in the chem ical potential nearthe substrate, w hich should mani fest i tsel f as an increasein w ater concentration. T he attracti ve force (tow ard thew ater) is at a maximum near the interface and then decaysas the distance (D ) from the interface is increased. T hedistance dependence on the decay is dependent upon theexact com bination of forces acting on the system (i .e., 1/Dfor van der Waals forces, 1/D 2 for electrostatic forces, ande- D for hydration forces).29 T he decay in chemical potentialas a resul t of these forces w i l l cause a w ater concentrationgradient to occur.

D ue to the di ffi cul t i es described prev iousl y , i t w ould beuseful to experim ental l y determ ine the excess w aterconcentration. Prev iousl y , Wu and co-w orkers used neu-tron refl ecti v i ty to character i ze the excess w ater concen-tration at a si l i con/pol y im ide interface.4 T he w ater dis-tr i bution can be ascertained by isotopic substi tution,nam ely , by using perdeuterated w ater. T he neutronscatter ing length densi ty for D 2O (3.3 × 10 - 4 Å- 2) i ssigni fi cantl y higher than that for any other com ponent inthe sam ple, including the si l i con (1.06 × 10 - 4 Å- 2) and thepol ymers (8.19 × 10 - 5 Å- 2 and 7.86 × 10- 5 Å- 2 for PB OCStand PH OSt, respecti vel y ). H ere, w e take a sim i lar

approach and di rectl y m easure the w ater distr i bution inthe PB OCSt and PH OSt fi lms.

T he neutron refl ecti v i ty profi l es and correspondingscatter ing length densi ty profi l es are show n in Figure 5for both PB OCSt and PH OSt. T he Q c

2 (scatter ing lengthdensi ty) profi l es read l i k e the X-ray data in Figures 1 and3 except that Q c

2 now reflects the neutron scattering lengthdensi ty profi l es. For the N R experim ents, Q c

2 depends onboth the composi tion and the mass densi ty . U pon exposureto D 2O, there are changes in the Q c

2 profi l e that w ere notobserved by XR due to the large scatter ing length densi tyof D 2O. T he fi lm thick ness increases, as observed in theXR m easurem ents, but the Q c

2 profi l e of the pol ym er fi lm scan increase appreciabl y and there is an especial l y largeincrease in Q c

2 near the si l i con/photoresist interface. Forthe PB OCSt fi lm , the thick ness increases from 104 ( 3Å to 120 ( 3 Å w i th an increase in Q c

2 from 8.19 × 10 - 5

Å- 2 to 8.55 × 10 - 5 Å- 2. T he PH OSt fi lm show s a morepronounced change in the N R profi l e than that observedfor PB OCSt, w i th an expansion from 110 ( 3 Å to 124 (3 Å and an increase in Q c

2 from 9.06 × 10- 5 Å- 2 to 1.50× 10 - 4 Å- 2. N otice the larger increase (besides theinterfacial region) in Q c

2 upon D 2O exposure for PH OStin com parison toPB OCSt. T his is because the labi l e - OHproton in PH OSt readi l y exchanges w i th deuterium fromD 2O, resul ting in a greater increase in Q c

2 than w hat simpleabsorption, w i thout exchange, w ould predict. H ow ever,this can be accounted for by assum ing al l of the hydroxy lsin PH OSt are replaced w i th - OD .

T he increase in Q c2 can be correlated to the D 2O

concentration in the fi lm . T he w ater concentration profi l ecan be determ ined from the scatter ing length densi typrofi l e determ ined from the best fi t of the neutronrefl ecti v i ty profi l e.

w here Q c2(x) i s the scatter ing length densi ty at a posi tion

x from the substrate, Q c2(poly) i s the scattering length densi ty

of the pure polym er, and Q c2

(D 2O ) i s the scatter ing lengthdensi ty of pure D 2O. For the PH OSt sam ples, Q c

2(pol y ) i s

taken as the scatter ing length densi ty for a com pletel y(31) K inlock , A. J. Adhesion and Adhesi ves Scienceand T echnology;

Chapm an and H al l : L ondon, 1987.

Fi gu r e 5. Scatter ing l ength densi ty profi l es (Q c2) for (a) PH OSt and (b) PB OCSt fi lm s before (dashed l ine) and after exposure (sol i d

l ine) to saturated D 2O vapor. T he reflecti v i ty profi l es w i th fi ts corresponding to the densi ty profi l es are show n in parts c and dfor PH OSt and PB OCSt, respecti vel y . T he refl ect i v i ty for the exposed fi lm s is offset by tw o decades for clar i ty .

µ ) åµexternal + åµi nternal (1)

φw (x) )Q c

2(x) - Q c

2(pol y )

Q c2

(D 2O) - Q c2

(pol y )

(2)

5288 L angm ui r , Vol . 20, N o. 13, 2004 Vogt et al .

This can be extended to

multiple phase systems

Can we further increase sensitivity?

Vogt et al., Langmuir 2004, 20, 5285.

Crosslinked polyacrylate on sputtered alumina

10-6

10-5

10-4

10-3

10-2

10-1

100

Reflectiv

ity

0.120.100.080.060.040.02

q (Å-1

)

XR

NR

Dq

2.0

1.5

1.0

Qc2

(Å-2

x10

-4 )

590058005700

Distance (Å)

Polymer

Alu

min

a

D2O

acc

um

ula

tio

n

SiO

x

Sili

con

Si

Alumina

Polymer

OH

OH

OH

OH

OH

OH

OH

OH

OH

OH

OH

Qc2 = 3.21 x 10-4 Å-2

Qc2 = 0.79 x 10-4 Å-2

h = 1.5 mm

Qc2 = 2.4 x 10-4 Å-2

h = 35 nm

Qc2 = 1.06 x 10-4 Å-2

h= 2 mm

Interference fringes from thin alumina

Shift due to contrast change from

D2O sorption

Improved sensitivity to interface

Vogt et al., J. Appl. Phys. 2005, 97, 114509

Moisture accumulation for polyacrylate on alumina

67

0.1

2

3

4

5

67

1

Reflectivity

0.0350.0300.0250.020q (Å

-1)

0.7 % Swelling

6.5 % Swelling

XR – measure film swelling

Vary cure condition to change H2O solubility

NR

NR

Bulk solubility does not influence moisture content at alumina / polymer interface

Al2O3

Al2O3

H2O

H2O

Vogt et al., J. Appl. Phys. 2005, 97, 114509

Design of system to maximize sensitivity ★ Relatively thin polymer coating (< 150 nm) ★ High contrast oxide layer

(10-30 nm) ★ D2O as probe ★

10-6

10-5

10-4

10-3

10-2

10-1

100

Refle

citiv

ity

0.200.150.100.05

q (Å-1

)

2

4

6

810

-4

2

4

6

8

Reflecitiv

ity

75x10-3

7065605550

q (Å-1

)

shifts

Si

Alumina

Polymer

OH

OH

OH

OH

OH

OH

OH

OH

OH

OH

OH

Qc2 = 0.79 x 10-4 Å-2

h = 85 nm

Qc2 = 2.4 x 10-4 Å-2

h = 35 nm

Qc2 = 1.06 x 10-4 Å-2

h= 2 mm

Multiple interferences yields

added sensitivity to buried

interface contrast

Vogt et al., Langmuir 2005, 21, 2460

Water accumulates at interface

• Does this accumulation directly impact

adhesion?

• Can this accumulation be controlled?

• What are critical factors?

Critical relative humidity

-0.25

0.25

0.75

1.25

1.75

2.25

500150025003500

Wavenumber (cm-1

)

Ab

sorb

an

ce

B-O stretching

Si-O-Si stretching

-0.05

0.15

0.35

0.55

0.75

500150025003500

Wavenumber (cm-1

)

Ab

sorb

an

ce

C-O-C stretching

C=O stretching

Si

Silicon oxide

Polymer

OH

OH

OH

OH

OH

OH

OH

OH

OH

OH

OH

Qc2 = 0.79 x 10-4 Å-2

h = 85 nm

Qc2 = 1.6 x 10-4 Å-2

h = 15 nm

Qc2 = 1.06 x 10-4 Å-2

h= 2 mm

Cohesive failure

Adhesive failure

Is the interface to blame for

failure at high humidity?

Tan et al., Langmuir 2008, 24, 9189

Impact of humidity on interfacial properties

Increasing partial

pressure

What’s driving the failure?

0.15

0.10

0.05

0.00

D

2O

2520151050Distance from SiOx interface (Å)

PD2O (bar)

0.0077 0.0096 0.0118 0.0135 0.0154 0.0165 0.0176 0.0187 0.0200 0.0213

PMMA

0.12

0.08

0.04

0.00

D

2O

403020100-10Distance (Å)

1.0

0.8

0.6

0.4

0.2

0.0

S

iO2

Water

SiO2

bulk

Exce

ss

Tan et al., Langmuir 2008, 24, 9189

Comparison of interface and bulk

Critical humidity

No discontinuity at interface

So is moisture at interface important?

0.20

0.15

0.10

0.05

0.00

Pe

ak in

terf

acia

l co

nce

ntr

atio

n (

vo

l fr

ac)

0.0200.0150.0100.0050.000Partial pressure D2O (bar)

0.035

0.030

0.025

0.020

0.015

0.010

0.005

0.000

Bu

lk s

olu

bility

(vo

l frac)

0.0200.0150.0100.0050.000Partial pressure D2O (bar)

Tan et al., Langmuir 2008, 24, 9189

1.0

0.8

0.6

0.4

0.2

0.0

100806040200Distance (Å)

phenylphosphonic acid treatment

Influence of substrate on moisture accumulation

1.0

0.8

0.6

0.4

0.2

0.0

100806040200Distance (Å)

Bare Al2O3

1.0

0.8

0.6

0.4

0.2

0.0

100806040200Distance (Å)

t-butylphosphonic acid treatment

0.30

0.25

0.20

0.15

0.10

0.05

0.00

[wate

r]m

ax

100806040200

Water contact angle (º)

SiOx

Al2O3

t-butyl

phenyl octyl Moisture accumulation at interface

dependent upon surface chemistry

Incomplete coverage (74°)

90° for monolayer O

O

P

O O

P

O O O

P

O O

OH

OH

O’Brien et al., Adv. Eng. Mater. 2006, 8, 114

Correlating interfacial moisture and adhesion for PMMA

0.01

0.1

1

10

100

1000

G (

J/m

2)

Surface ChemistrySiOx

Al2O3 t-butyl phenyl octyl

Dry

Wet

160

120

80

40

0

DG

(J/

m2)

0.300.200.100.00

[Max Water] (vol. fraction)

(a)

20

10

0

G (

J/m

2)

0.300.200.100.00

[Max Water] (vol. fraction)

(b)

t-butyl

Interplay between dry adhesion and moisture accumulation

Intermediate surface energy for best wet adhesion

Interfacial water matters!

O’Brien et al., Adv. Eng. Mater. 2006, 8, 114

Change in the interface at the critical

humidity 0.10

0.08

0.06

0.04

0.02

0.00

D

2O

2520151050Distance from SiOx interface (Å)

½ FWHM

Critical humidity

Interfacial moisture broadens at the critical humidity

Is this responsible for the failure?

Tan et al., Langmuir 2008, 24, 9189

How could this change in the profile impact adhesion?

Only the average in plane distribution of water is determined from neutron reflectivity

Distribution is not uniform!

Dry surface

Ultrathin film (15 nm)

Water drop on

surface

Hydrophobic interface

(minimal adhesion loss)

Hydrophilic interface

(catastrophic adhesion loss)

Proposed model for critical relative humidity

Increasing humidity

Critical humidity

1. Initial accumulation at the interface only mildly perturbs the contact area – adhesive energy still

greater than cohesive energy

2. This grows, but does not significantly stress the interface

3. When the bulk concentration increases significantly at the critical humidity, it causes shear stress

accumulation near the interface due to the decreased contact area

4. This stress forces the additional water accumulating at the interface to be pushed into the film

5. This additional water at the interface causes a large stress normal to the film surface leading to

adhesive failure

How to test model? • Stress accumulation should be important

Water causes normal stress in adhesive?

Ef= 3.15 ± 0.07 GPa Ef=1.91 ± 0.1 GPa Ef=1.20 ± 0.06

GPa

PMMA PEMA PnPMA PnBMA

Ef < 100 MPa

Increasing chain length

Tg = 105 ºC Tg = 65 ºC Tg = 36 ºC Tg = 15 ºC

n=0

Model: poly(n-alkyl methacrylate)

n=1 n=2 n=3

Torres et al., ACS Nano 2009, 3, 2677

Impact on changing the model adhesive

250

200

150

100

50

0

G (

J/m

2)

100806040200

Relative humidity (%)

PMMA

Glassy

100

80

60

40

20

0

G (

J/m

2)

100806040200

Relative humidity (%)

PnBMA

Rubbery

250

200

150

100

50

0

G (

J/m

2)

100806040200

Relative humidity (%)

PEMA

Glassy

Low modulus PnBMA does not exhibit critical relative humidity

What about moisture uptake?

0.05

0.04

0.03

0.02

0.01

0.00

Vo

l fr

ac H

2O

1086420

n

PMMA

PnBMA Near saturation – less water in

PnBMA than PMMA

Longer alkyl chain leads to more

hydrophobic model adhesive

Is the interface impacted? 0.12

0.10

0.08

0.06

0.04

0.02

0.00

D

2O

20100-10

Distance (Å)

1.0

0.8

0.6

0.4

0.2

0.0

S

iO2

PD2O = 0.0077 bar

PMMA

PBMA

SiO2

0.12

0.10

0.08

0.06

0.04

0.02

0.00

D

2O

20100-10Distance (Å)

1.0

0.8

0.6

0.4

0.2

0.0

S

iO2

PD2O = 0.016 bar

PMMA

PBMA

SiO2

200

150

100

50

0

Fra

ctu

re e

ne

rgy (

J/m

2)

100806040200

Relative humidity (%)

PMMA

PnBMA

Different at

high

humidity?

7

6

5

4

3

2Wid

th o

f in

terf

acia

l [D

2O

]

0.0200.0100.000

Partial pressure (bar)

PMMA

PnBMA

Does low modulus lead to decrease in

moisture?

0.12

0.08

0.04

0.00

D

2O

870860850840

Distance (Å)

PMMA

PBMA

POMA

n=0

n=3

n=7

Low modulus appears to lead to suppression of moisture accumulation at interface

(rubbery polymers)

Karul et al., Polymer 2009, 50, 3234

Excess Swelling

T

wexcess dxxt0

)(

Bulk Swelling

ti tf f

if

eqt

tt

w

Shaded area is excess thickness

Simple Model for Thickness Dependent Swelling

Water

0

0

0

)(1

h

tht

h

h excesseq D

Vogt et al., Langmuir 2005, 21, 2460

Thickness dependent swelling

Model corresponds well with thickness dependent swelling

1.3

1.2

1.1

1.0

h/h

0

4 6 8

1002 4 6 8

10002 4 6 8

Initial film thickness (Å)

Al2O3

SiOx

Vogt et al., Langmuir 2005, 21, 2460

Is there always an excess at the interface?

0.50

0.45

0.40

0.35

0.30

0.25

0.20

Volu

me

Fra

ction W

ate

r

6 8

102 4 6 8

100Initial Film Thickness (nm)

On SiOx

On HMDS

Poly(vinyl pyrrolidone)

No thickness dependence for films on native oxide

HMDS treated substrate leads to decrease in absorption for thin films Vogt et al., Polymer 2005, 46, 1635

What about aqueous solutions?

We obtain similar information about the interfaces immersed in D2O

0.30

0.25

0.20

0.15

0.10

0.05

0.00

h/h

o

160012008004000

Initial Film Thickness (Å)

PHOSt

PBOCSt

Liquid immersion with HMDS treatment

• Hydrophilic interface relative to bulk polymer

interfacial excess & enhanced thin film swelling

• Hydrophobic interface relative to bulk polymer

interfacial depletion & suppressed thin film swelling

3.0x10-4

2.5

2.0

1.5

1.0

0.5

0.0

Qc

2 (

Å-2

)

200150100500

z (Å)

Air D2O

PHOSt on HMDS

Si / SiOx

D2O

Air

D2O depletion near

HMDS surface !

film

swelling

3.0x10-4

2.5

2.0

1.5

1.0

0.5

0.0

Qc

2 (

Å-2

)

4003002001000

Distance (Å)

D2O

AirSi / SiOx

D2O enrichment near

HMDS surface !

film

swelling

Vogt et al., J. Microlith. Microfab. Microsys. 2005, 4, 013003

What about salts?

Epoxy on oxidized aluminium

Addition of salt increases fracture energy for failure

Does this correlate with the interfacial water content?

Developing system with high sensitivity

Epoxy

Aluminum Silicon oxide

Silicon

Note that SLD for aluminum is very close to silicon

Qc2 = 0.59 x 10-4 Å-2 h = 80 nm

Qc2 = 1.68 x 10-4 Å-2 h = 25 nm

Qc2 = 1.06 x 10-4 Å-2 h= 2 mm

Qc2 = 0.95 x 10-4 Å-2 h= 18 nm

Impact of salt on water distribution Comparison of profiles with saturated NaCl and D2O Bulk behavior

Sharper interface between epoxy and aluminum oxide in case of salt

Less bulk swelling in case of salt as well

Is this due to salt incorporation into the film?

Silicon

Silicon oxide Aluminum

Epoxy

D2O

Examination of NaCl in film

Silicon

Silicon oxide Aluminum

Epoxy

D2O + H2O

Contrast match

(CM)

CM + NaCl

In situ Ex situ

No evidence of salt accumulation within polymer

Qc2 = 1.04 x 10-4 Å-2 for NaCl

(Wang and Schaefer, Langmuir 2010, 26, 234)

Can we directly visualize salt ion distribution?

Prabhu et al., Langmuir 2005, 21, 6647

1.0

0.8

0.6

0.4

0.2

0.0

Volu

me fra

ction P

NB

HF

A

500400300200100

Z [Å] from Si

10-5

10-4

10-3

10-2

10-1

100

Refle

ctivity

0.160.120.080.04

Q [Å-1

]

QQcc22 r r

DD @ pD@ pDQQ

persistence persistence roughnessroughness

10-5

10-4

10-3

10-2

10-1

100

Refle

ctivity

0.160.120.080.04

Q [Å-1

]

QQcc22 r r

DD @ pD@ pDQQDD @ pD@ pDQQ

persistence persistence roughnessroughness

substratesubstrate

filmfilm

xx--raysrays

oror

neutronsneutrons

r

D

roughnessroughness

Q

substratesubstrate

filmfilm

xx--raysrays

oror

neutronsneutrons

r

D

roughnessroughness

Q

Direct measurement of Developer Profile

PNBHFA Film

ContrastContrast--Matched SolutionMatched Solution

d12-TMAH

PNBHFA Film

ContrastContrast--Matched SolutionMatched Solution

PNBHFA Film

ContrastContrast--Matched SolutionMatched Solution

PNBHFA Film

ContrastContrast--Matched SolutionMatched Solution

d12-TMAHd12-TMAH

PNBHFA Film

ContrastContrast--Matched SolutionMatched Solution

PNBHFA Film

ContrastContrast--Matched SolutionMatched Solution

• Contrast match solvent to the dry film (not much reflectivity)

• d12-TMA uptake increases contrast (more reflectivity)

• Base uptake observed

H2O (-0.3)

D2O (3.3)

Air (0)

PHOSt (0.8)

PNBHFA (1.0) Si (1.1)

SiO2 (1.9)

d12-TMAH (4.2)

Com

bin

e to

matc

h P

NB

HFA

d12-TMA profile within ultrathin PNBHFA

• Direct measurement of base profile within thin solid polyelectrolyte films

• TMA+ concentration within film enhanced with increasing base concentration

• Non-uniform profile within the film; reduced near the substrate

• Diffuse counterion profile at free surface

Prabhu et al., Langmuir 2005, 21, 6647

Combining contrasts

Silicon

Silicon oxide

PNBHFA

D2O

Silicon

Silicon oxide

PNBHFA

CM

d-TMAH h-TMAH

Polymer is visible Ions are visible

Can measure excess ion concentration that extends into

solution from charging of polymer film

Prabhu et al., Langmuir 2005, 21, 6647

Conclusions

• Water accumulates at polymer/substrate interface

• Concentration at interface

• Independent of bulk solubility

• Dependent on substrate chemistry

• Water accumulation correlates with adhesion loss

• Mechanism for critical relative humidity appears to be

stress concentration due to moisture accumulation

• Salts can increase or decrease water uptake in films

• Contrast match provides facile route to visualize ion

distribution

Acknowledgements

Christopher C. White

Kar Tean Tan

Don Hunston

Jon Martin

Emmett P. O’Brien

Wen-li Wu

Eric K. Lin

Christopher L. Soles

Vivek M. Prabhu

Hae-Jeong Lee

Dean M. DeLongchamp

Ronald L. Jones

Patricia M. McGuiggan

Jack F. Douglas

Polymers

Division

Sushil Satija

Bulent Akgun

Dan Dender

Paul Kienzle

FUNDING

DARPA Contract No. N66001-00C-8083

NSF CMMI-0653989

Jessica Torres

Casey Campbell

Alper Karul