Unusual behavior of water soluble polyelectrolyte macromolecules shown by QELS study: is it a property of hydrophobic backbones ?

HAL Id: jpa-00212531

https://hal.archives-ouvertes.fr/jpa-00212531

Submitted on 1 Jan 1990

HAL is a multi-disciplinary open access

archive for the deposit and dissemination of

sci-entific research documents, whether they are

pub-lished or not. The documents may come from

teaching and research institutions in France or

abroad, or from public or private research centers.

L’archive ouverte pluridisciplinaire HAL, est

destinée au dépôt et à la diffusion de documents

scientifiques de niveau recherche, publiés ou non,

émanant des établissements d’enseignement et de

recherche français ou étrangers, des laboratoires

publics ou privés.

Unusual behavior of water soluble polyelectrolyte

macromolecules shown by QELS study: is it a property

of hydrophobic backbones ?

Hedi Mattoussi

To cite this version:

Unusual behavior of

water

soluble

_{polyelectrolyte}

macromolecules shown

_{by QELS study:}

is it

a

_property

of

_hydrophobic

backbones ?

Hedi Mattoussi

_(*)

Polymer

Science and

Engineering Department, University

of Massachusetts, Amherst, MA _01003, U.S.A.

(Received

2

_May

1990,

accepted

19 June

₁₉₉₀₎

Résumé. 2014 Nous avons étudié les aspects

dynamiques

d’un

_{polyélectrolyte}

en solution aqueuse, en utilisant la

_technique

de la diffusion

_{quasi-élastique}

de la lumière. Deux conclusions

_majeures

ont été atteintes _pour le cas des solutions sans contre-ions extérieurs et le cas où la force

_ionique

devient

importante.

Pour le cas sans sel, la

_dynamique

du

système

est caractérisée par deux modes

de fluctuations : un mode

_coopératif

et des fluctuations à

_longue

distance. Ceci reflète une

conformation étendue des solutes macromoléculaires. Au contraire,

l’augmentation

de la force

ionique

est

_accompagnée

_par une

_importante

réduction de la conformation des macromolécules.

Un seul mode de fluctuations

_émerge.

Le coefficient de diffusion

_{collectif, qui}

a aussi été _mesuré, ne

_dépend

ni de la concentration du

soluté,

ni de la force

_{ionique quand}

cette dernière

_dépasse

une valeur

_critique

faible. Nous attribuons ce

_phénomène

à la

_{prédominance}

d’interactions

hydrophobes,

étant donné que les effets

d’écrantage

réduisent la

_portée

des forces de Coulomb

entre les

polyions,

et à cause de la structure

_{d’hydrocarbure}

du

_squelette

de

_polymère.

Abstract. 2014

Using QELS technique

we studied the

_dynamic

_aspects of a

_synthetic

flexible

polyelectrolyte compound

in aqueous solutions.

By

monitoring

the ionic

strength

from very small to

_high

_values, we reached two

_major

conclusions. First, in the absence of added

_{electrolyte(s),}

the

_polyions

are in a «

_{highly » expanded}

_{conformation ;} the

dynamics

are characterized

by

two

fluctuation modes :

cooperative

and

long

range fluctuations. The addition of NaCl introduces a

«

_{sharp »}

and a _priori unusual

_change.

The solute macromolecules

_experience

a non

_gradual

change

in their conformation, which is

_directly

reflected in their

_dynamics.

The mutual diffusion

coefficient,

deduced from the

_{only existing decay}

rate for the correlation

_function,

does not reflect any observable variation with

polymer

or salt concentration

(cP

and _cS

_{respectively),}

for the values

we scanned. We attribute this

_phenomenon

to the

_rising

of

hydrophobic interactions,

once the

shielding

of the electrostatic interactions becomes effective at

_sufficiently

_high

ionic

strengths,

owing

to the

hydrocarbon

nature of the

_polyions.

Classification

Physics

Abstracts 78.00 - 46.60 - 47.50

Introduction.

It has been known that

_{polyelectrolyte}

solutions

_experience

different behavior

_depending

on

whether or not one adds

_{simple electrolyte(s)}

ions to the

_system

_[1-8].

For solutions without

extemally

added

counterions,

the

_long

_{range Coulomb}

_repulsive

interactions between

neighboring charges along

one

_polyion

chain

_impose

a local

_stiffening

and,

consequently,

an

expanded

conformation to the

_polyion.

This

_phenomenon

is at the

_origin

of _{the appearance} of an « ordered »

_phase,

as was found

_by

small

_angle

neutron and

_X-ray

_scattering

studies

_[4-8].

Conversely,

the counterion excess in solutions with

_high

ionic

_strength

screens out these

repulsive

interactions. The

_screening

effects manifest themselves in a

_{gradual change}

in the

solution

_properties

as function of the ionic

_{strength [4-7].}

For

instance,

the « ordered »

_phase

disappears

for

_{sufficiently high}

counterion excess

_[4].

This is

_{accompanied by}

a reduction in the chain

_expansion

for flexible

_polymers.

This

_change

in

_polyion

conformation with the

addition of ions to the solution has been observed to occur for many

_{polyelectrolyte}

compounds.

Nevertheless,

its onset and

_development

_vary for different

_compounds.

_Many

parameters

contribute :

_polyion

nature and

_{charge density along}

the

backbone,

for instance.

QELS technique

has been used

_by

many authors to

_probe

_{polyelectrolyte}

solutions

properties [9-19].

The diffusion coefficient and its

_dependence

on

polymer

concentration,

molecular

_weight,

macromolecules

valence,

ionic

_strength,

etc., have been measured. Two

diffusion processes were often found for low ionic

_strength

even at small concentrations. Above an ionic

_strength

_threshold,

_only

one diffusion coefficient was measured. This coefficient has different

_types

of variation with the ionic

_{strength, depending}

on the intrinsic

properties

of the

_polymer.

On one

hand,

the value measured at zero

_polymer

concentration,

Do,

is constant for some

_biological

materials :

_DNA,

BSA,

which are

_ordinarily

assumed not to

_undergo

_any conformational

_change

as the ionic

_strength

of the solution increases

_[10-13,

18].

However,

for many flexible

_{macromolecules,}

_Do

does

_change

_{progressively}

and

noticeably

with the increase in counterion excess

_{[18, 29].}

On the other

hand,

for finite

polymer

concentration the diffusion coefficient is found to decrease with

_increasing

cs for BSA and

DNA,

but it increases with cs for K-carrageenan

[12-16].

We

present

a

_{QELS study,}

undertaken on a water soluble

_{synthetic polymer compound.}

The

_change

in solution

_properties

between the salt-free case and the case with added salt is found to occur in an « unusual » manner when

_compared

to

_previous

results

_reported

on

polyelectrolyte

materials. The

_present

set of data is

_compared

to

_previous

work

_reported

on different

_compounds

as well as other measurements undertaken on this

_system

in a different solvent

_{(methanol) [29].}

Expérimental

section.

It is well known that the time

_dependent

structure factor

_{S(q, t )}

is non-zero for non

homogeneous

media,

e.g.

polymer

solutions,

in which concentration fluctuations 8 c occur

[20, 21] :

where D is the translational diffusion

coefficient, q

is the

_scattering

wavevector which

depends

on the

_{scattering angle}

_{O : q}

=

(4

7r/,k )

ns sin

(0/2) ;

ns is the solvent refractive

index and A is the

_wavelength

of the incident

_light.

For concentrated

_{polymer solutions,}

macromolecules

_{interpenetrate}

and fluctuations become also a

_{cooperative phenomenon.}

Therefore,

S(q, t )

could be written as a combination of the two contributions

_{[21] :}

where a and

₃

are two constants which reflect the relative contributions of these two processes, and

obey

the condition a

_{+ f3 =}

1.

_Ds

and

_Dc

are,

_{respectively,}

the

_large

scale and

the

_cooperative

diffusion coefficients. These two different coefficients are related to two characteristic sizes

_through

the Stokes relation :

_{Dc, s}

=

kT/6

7T17s

03BEH’

SH,

where 17s is the

solvent

_viscosity

coefficient,

SH

is the

_large

scale size fluctuation

_(often

attributed to the

« overall » size of the macromolecule for

_{high concentrations)}

and

_eH

is the

_hydrodynamic

correlation

_length

or the

_dynamic

blob size

_[21].

The

_predominance

of one _process or the

other

_depends

on the time scale measurement : for t small

_rc

is

_{predominant ;}

a combination

of both processes is observed over

_longer

time scale.

The

_QELS

_{experiment provides}

the normalized

_intensity

autocorrelation function

G (2)( t),

at a

_{given scattering}

vector _{q :}

G

_{(2)( t)}

also reads as :

where a is the

_{background signal, b}

is a constant

_accounting

for the

_{detecting optics,}

and

g(1)(t)

is the normalized field correlation function

_directly

related to

_{S(q, t ) (Eqs. (1)}

and

(2)).

For solutions with

_homodisperse

solute macromolecules in the dilute

_regime

_(isolated

macromolecules),

the autocorrelation function has a

simple

form :

where r =

Def q2

and

Def

is the coil mutual diffusion coefficient also

accounting

for the

objects

interactions in the solution.

_However,

in

_polymer

solutions macromolecules are

always subject

to size

_dispersion,

and

_{single exponential}

form does not

_provide

a

good

description

for the relaxation form of G

_(2)(t).

In

_practice,

cumulants

_analysis

is often used to

fit the correlation function :

where,

A, B, C,...

are

_{respectively,}

related to the first second etc., cumulants

[30].

A = - 2 r and

provides

us with

_{D,,f. B}

is related to the variance and so reflects the deviation from a

_{single exponential}

form. The ratio

_B/A

defines the

_{polydispersity}

of the measurement. The cumulants

_expansion

has

_provided

a

_good

fit for the correlation function G

_(2)(t)

for the set of solutions with «

_{high »}

ionic

_strength.

The

_{polydispersity}

factor

_B/A

for these

_systems

was found to be about 0.15-0.25 and is considered

_satisfactory

because of the molecular

weight

distribution

_{[22, 26].}

However,

for the case of salt-free solutions

_(or

solutions with very low ionic

strength),

the

dynamics

occur

_{(Fig. 2b).}

A double

_{exponential expression}

_{is necessary} to describe the

intensity

correlation function

_decay

with time :

Tc,

s are introduced above

(Eq. (5)).

In

practice,

the two

decay

rates are

_actually

extracted

through

the use of the

_Laplace

inversion of the normalized distribution of

_decay

rates G

_{(r) :}

:

A

_{plot G ( T )}

vs.

T (or Log r)

reflects the correlation function

_behavior,

_e.g.,

_{G (r)}

shows one or more

_{peaks depending}

on whether or not the field correlation function

_{g (1) (t)}

has one or more

_decay

rates. The

_Laplace

inversion was done

_{numerically using}

a

_computer

Vax program : Contin

analysis, provided by

the

computer

center at the

_University

of Mas-sachusetts. This

_analysis

has been

_{mostly applied}

to salt-free solutions.

_Nonetheless,

we

checked that for solutions with

_high

ionic

_{strength, G (F)}

exhibits one

_peak

and that the

corresponding Fav (average decay rate) provides

a value

_Dav

for the diffusion coefficient which agrees with

Def

extracted from cumulants

_analysis.

We used a well known

_photon

correlation

_{spectroscopy set-up,}

_{i.e., QELS,}

which has been

described

_previously

[30, 31].

The normalized

_intensity

autocorrelation function

_{G (2)(t)}

is recorded

_using

a

_{photocorrelator Langley}

Ford DM 1096 with 256 channels and 16 channels

delayed.

We used a He-Ne laser

_{source, À}

= 6 328

Á,

and an index

matching

bath for the

sample

tube

_[30,

_31].

In the

_present

_work,

the

_{polyelectrolyte compound}

used is the

_{poly(xylydene}

tetrahyd-rothiophenium chloride),

described

_previously,

and often identified as the

_{poly (p-phenylene}

vinylene),

PPV,

precursor

[23-26].

It has the

following

chemical formula :

This

_compound

has been

_synthesized

in an « unusual manner ». The

_polyions

were obtained from anionic

_{polymerization,}

in water, of the monomer

_(xylydene

tetrahyd-rothiophenium chloride)

ions. The

_polymer

thus obtained is afterwards

_precipitated

in

acetone _{and washed many} times with clean deionized water, to eliminate extra free ions. It is then dried under

_{nitrogen atmosphere}

for several hours. The

_resulting

material is then

dissolved in clean double deionized water and used for measurements _{purpose. To check the}

purity

of salt-free solutions from residual

ions,

we measured the

_{conductivity :}

this was found

very small : a = 0.04-0.1

ms/cm

for _{Cp 1}

_{g /1.}

_However,

the

_conductivity

reaches a value

o- = 13

_ms/cm

at _{cs = 0.01 M} and _Cp = 0.3

g/1.

The smallness of this

value,

in the first _case,

reflects the

_purity

of the salt-free

_{solutions ;}

it has a finite value because of the residual contribution of the ions

_coming

from the macromolecules. The solutions were filtered into the

sample

tube

_through

a

_millipore

filter 0.8 _um. A smaller hole size filter

_{(0.45 um)}

is found u

alter the solutions. The molecular

_weight

was determined

_by

G.P.C.,

and checked

_{by light}

scattering [29] : Mw

=

106

and

(MW/Mn.

=

2) [26].

The

_compound

is the

_subject

of active interest because of the

_high

electrical

_conductivity

of

sulfonium salt at

_high

_temperature.

The

conductivity

reached is about a =

103

ms/cm,

when

doped

with metal atoms such as arsenic

pentafluor (AsFS)

[23-25].

PPV also shows a

_strong

tendency

towards

_crystalline

behavior

[27, 28].

In aqueous

solutions,

the

_{high polarity}

of water molecules is assumed tp be

strong

enough

to dissociate chloride ions from the

_polymer

backbone and thus to induce a

_{polyelectrolyte}

behavior to the solution. Sodium chloride

_{(NaCI), (Fisher}

Scientific

Inc.),

was used as a conventional

_electrolyte.

All measurements were made at a constant room

_temperature

of

T = 25 °C.

Results and discussion.

The main

_point

we wish to

_emphasize

is that the salt-free

_(cs

=

0)

solutions exhibited

fundamentally

different behavior from the case of solutions to which salt had been added

(cs # 0).

The

_change

in solution behavior occurs at different levels.

First,

the static scattered

intensity I (q, 0 )

is very weak for salt-free solutions.

However,

it is

_strongly

increased

_by

the

presence of a small amount of added NaCl. For

_instance,

the ratio between the scattered

intensities _{in the presence and the absence} of salt is about 15 for 0 =

25°,

Cp = 1

g/1

and

cs = 0.10 M (Fig. 1).

This

_important

difference cannot be

_simply

attributed to a

higher

refractive index increment

_{( d n / d c )2}

in the _{presence of added salt. The second}

_change

in solution behavior concerns the

_intensity

_correlation,

_{G (2)( t)}

curves. This

_point

is discussed

separately

for both set of solutions.

Fig.

1. - Static scattered

intensity

I

_{(in arbitrary unit)}

vs. q for different

samples : (0)

pure water,

(o)

salt-free solution at _{Cp ~ 1}

_{g /1,}

(A)

Solution with added salt, cs = 0.02 M, _cp = 1

g /1,

(0)

The same

solution with added but not controlled salt : cs « 0.3 M.

a. SALT-FREE SOLUTIONS. - For solutions with no added

electrolyte,

even

_though

the static

different

_decay

rates

_Ff

and

_{Fs (Figs.}

_2a,

_3).

These

_decay

rates reflect two

_separate

modes of

fluctuations called

_{by QELS}

_users, fast and slow modes

_{respectively.}

The

_{corresponding}

diffusion coefficients

_De

and

_{Ds provide}

two characteristic sizes

_eH

and

_{SH, using}

the Stokes relation

_(Tab.I).

_The

net distinction between the

_magnitude

of

_Ff

and

_{Fs (hence}

Table I. Values

_of

the

_cooperative

and

_large

scale

_{diffusion coefficients, De}

and

Ds

as well as the

_{corresponding}

sizes

_eH

and

_{Su for}

«

_salt-free

» solutions are

_reported

in the

first

line. The evolution

_{of Dc}

and

_Ds

with salt concentration are

_reported

in the other lines.

Fig.

2. -

Semi-logarithmic plot

of the

_intensity

correlation function G

_(2)(t)

vs. channel number. The

time t is the

_product

of the channel number and the

_sample

time

(s.t.) : (a)

case of a salt-free solution,

polymer

conentration cp = 0.3

g/1,

s.t. = 80 _{kts, 0} =

15° ;

it exhibits two

modes ; (b)

solutions with

added sodium chloride : cp = 0.3

g/1,

Fig.

3. - Variation of the normalized function G

(r)

with

_Log

r. Two

_peaks

characterize the existence of two _separate modes of fluctuations :

rf

and

_TS.

eH, SH),

rules out the idea

_{of attributing}

them to a bimodal macromolecular

distribution,

since

the fast mode is about two orders of

_{magnitude larger}

than the slow one.

Let us first discuss the

origin

of the two

_separate

modes. Their existence is a direct

consequence of

repulsive

interactions between

neighboring charges along

the

polymer

chain. The

_{high expansion,}

therefore,

imposed

to the macromolecules

_drops

the

_system

into the

« semi-dilute » case as soon as a small amount of

_polymer

is dissolved into the solution. This is made easier

by

two factors : the

strong

polarity

of the water molecules and the

_high

molecular

weight

of the solute macromolecules.

The

_present

set of data should also be discussed in the

_light

of other work. The

_major

features

_registered

for this case : weakness of the

intensity

and the two _{mode processes} for the

fluctuations,

have also been observed _{for the PPV precursor in} methanol solutions

_[29].

Sulfonated

_polystyrene

ionomers with a low

degree

of sulfonation

(4

%-6

%)

in

DMSO,

also

showed two

_decay

rates for salt-free

solutions,

in

_comparison

to neutral

_{polystyrene [17].}

Other

_QELS

work undertaken on

biological

materials

(DNA

and

poly (L-Lysine)

for

instance)

have

_pointed

to _{the presence} of two different relaxation modes

_{accompanied by}

an

important

decrease in the scattered

_intensity

when the ionic

_strength

of the solution was

brought

to _very small values

[9-13].

Nevertheless,

this

_preliminary

_agreement

_involving

the

relaxation _process and the scattered

_intensity

does not screen out some

major

differences,

which come out from a closer

_analysis

of the data. The

hydrodynamic

correlation

length

eH,

which was measured to be about 12

À

for the

_present

case, needs to be

_compared

to its

value in methanol

_eH

=

17 Â

and also to the one for SPS of reference

_{[17] : eH}

= 40

Á.

The smallness of

_03BEH

_{in aqueous solutions when}

_compared

to the case of solutions in

methanol,

could be attributed to the difference in

_{polarity strength}

of the two solvents. This

parameter

govems the electrostatic

interactions,

and therefore the macromolecular expan-sion. In

_fact,

the dielectric constant e is about 78

_(SI)

for _{water ;} it is

_only

32.5

_(SI)

for

methanol. The difference in

_{polarity strength}

between these two solvents could also

_play

an

important

role in

_defining

the effective

_{charge density along}

the

_polyion

backbone,

mainly

within the

_prospect

of a

partial

condensation _process

_[29].

In

fact,

the

dipole

SCI carried

by

be

_only

attributed to the solvent

_{polarity strength}

effect : E

_(DMSO)

=

47 ;

it is

larger

than

the value for

methanol,

but SPS macromolecules _carry a lower

_{charge density along}

their backbones. The combination of these two factors for SPS in DMSO may

provide

a

relatively

large

value for the

_hydrodynamic

correlation

_length

when

_compared

to the PPV precursor

cases. A close

_{investigation}

of the

_data,

to check an eventual

dependence

of

_Dc,

Ds

on different

_parameters

_{cp, cs, etc.,} did not

_provide

useful results

_primarily

because of the very weak scattered

intensity.

This

goal

is also

_beyond

our

_present

_purpose, since

cp is too small to undertake such difficult

_analysis.

The slow coefficient

_Ds

is often attributed to the center of mass motion.

_However,

such attribution is not obvious for the

_present

case. The amount of

_polymer

dissolved is very

small,

and

_using

the semi-dilute

_concepts

is not

_{straightforward.}

Other reasons such as

_plasma

(charges)

waves, with

large

scale

fluctuations,

could also take

place

in similar ionic media.

More detailed data for this case of salt-free solutions are needed in order to make any useful

comparison

with theoretical considerations.

b. SOLUTIONS WITH ADDED COUNTERIONS. - We now discuss the effects induced

_by

the

presence of a small

_quantity

of conventional

_electrolyte

in the solutions.

_First,

there is a fundamental

_change

in the form of the

_intensity

correlation function

_G(2)(t)

when a small

amount of sodium chloride is added to the solution. It shows

_only

one

_decay

_rate,

_{r (Fig. 2b),}

of the same order of

_magnitude

as

_Ts

for the

_previous

case. We did check that the relation

r =

Def q 2

holds for the domain of 0 scanned. The existence of

only

_one

decay

_rate for

G

_(2)(t)

is indicative of a _{diffusion process}

_{governed by}

Brownian motion of « isolated »

objects

in these solutions. The second

_{important point}

concerns the diffusion coefficient

(Dgf)

thus,

deduced as well as

_RH.

These values are

_independent

of both cP and

cs for salt concentrations above a certain threshold

_c*

_(about

0.002

_M)

and for the range of cp scanned

(Fig.

4a,

b)).

Solutions with salt concentrations below

c*

are also characterized

_by

the appearance of two

_modes,

as

_previously

discussed for salt-free solutions.

_However,

the

corresponding decay

rates differ from

_{T s and}

_{rc (for}

_{cs =}

₀₎

and fluctuate from one solution to another.

This set of results is unusual and different from what has been

_reported

for flexible

polyelectrolyte

macromolecules. The most

_striking

fact is the absence of

_dependence

of

Def

on _{cs. In}

_fact,

a variation of the mutual diffusion coefficient and also the

_single

chain one

Def ( Cp ->

0 )

has been recorded for flexible

_{compound :}

sulfonated

_polystyrenes

with a wide

range of

charge density

have

_given

mutual diffusion coefficients that

_depend

on _cs

[18].

A similar behavior has been observed for

_{poly (L-Lysine)}

solutions

_[9].

Solutions of DNA and

BSA have also showed a

_{complex dependence}

_{of Def}

_{(cp # 0)}

on _{cs, when} the ionic

_strength

is

increased.

_Moreover,

the

_present

solute macromolecules showed a different behavior in methanol : a

_complex

variation of

_Def

with concentration of solute

_polyions

and added salt

[29]. Consequently,

the increase of the ionic

_strength

of the solution

_{(cs >}

0.002

_M)

seems to induce a

_sharp

and « definite »

_change

in the conformation of the macromolecules _{in aqueous} solutions.

For dilute

_polymer

solutions,

an

_{expression accounting}

for the solute interactions

(thermodynamic

and

_{hydrodynamic)}

for the mutual diffusion coefficient reads as

_{[32] :}

Fig.

4.

_{- (top)}

Plot of the effective diffusion coefficient

_Def

vs.

_polymer

concentration cp for a

_given

salt

concentration value _cs = 0.008 M.

(bottom)

Plot of the diffusion coefficient

Def

v.s. salt concentration cs for a

_{given polymer}

concentration cp = 0.3

g/1.

cl and c2 are two constants of the order of 1

_[32].

An

_analysis

of the

_experimental

data within the framework of such an

_expression

did

_provide

a

_good

basis to

_explain

the

_complex

variation of

_Def

with both

_polymer

concentration and ionic

_strength

in methanol

_[29].

However,

this

_expression

does not seem to reflect the

_type

of

_dependences

of

Def

on these

_respective

_parameters.

There may be two

_explanations

that one can think of :

(1)

the addition of sodium chloride ions to the solution could have

_brought

the solutions to a 0

_condition,

_{for cs}

_{large enough.}

_Therefore,

the second Virial coefficient is zero and the

dependence

on _cp will

_strongly

weaken,

and could even become

_negative.

_However,

this

suggestion

could not

_{explain why}

the diffusion coefficient

_Def

does not vary with the ionic

strength

of the solution. We tried to scan much smaller values of

_polymer

concentrations in order to check an eventual

dependence

of

_Def

on Cp. It was not

_possible

to extract _any useful

data,

given

the extreme weakness of the scattered

_{intensity. Consequently,}

it would be safer

(2)

the second

attempt

is based on

_entropic

consideration and seems more

_likely

to

_provide

an

explanation

to the _present set of data

_(Fig.

4a,

b).

Within these

_{considerations,}

one needs

to think of the difference between the two solvents

_(water

and

_methanol)

on a molecular level

and from a statistical mechanics

_point

of view. The

_hydrophobic

_{interactions,}

_purely

of

entropic

nature, are

_{special properties experienced by hydrocarbon compounds}

in the

presence of water molecules. A close

_analysis,

at the molecular

level,

of the PPV precursor

macromolecules shows that

_they

are made of two

_{parts :}

_{the side groups}

_(tetra

hyd-rothiophenium chloride)

which carry the

dipoles

SCI,

and

consequently

are _very sensitive to

polar

environment ;

and the

_hydrocarbon

backbone

_{(phenylene vinylene).}

The skeleton is very

likely

to

_{experience hydrophobic}

interactions when

_exposed

to water molecules. The

solvation

_properties

of these

chains,

in water, is made

_{possible mainly}

because of the side group

dipoles

interactions with those of water molecules. Within the framework of this

reasoning,

the presence of extra counterions screens out the Coulomb

_repulsive

interactions,

which were

_responsible

for the

_polymer

chain extension. The

_screening

of the

charge-charge

repulsion along

each

_backbone,

leads the solute molecules to interact with the water

molecules as a

_hydrocarbon

chain.

_Hydrophobic

effects could therefore be manifested in a

collapse,

and very

likely

association of the macromolecules in the soltion. Such

_phenomenon

could

_explain

the

_important

increase of the scattered

_intensity.

It

_provides

isolated

_aggregates

in the

medium,

which are also very

weakly

sensitive to the variation of the ionic

_strength

of the _{environment for cs>}

_{cs .}

More,

this process could

provide

an

_explanation

for the lack of

dependence

of

_Def

on _Cp, for

_relatively

small

_polymer

concentrations. Within the framework

of

_{equation (10)}

for

_Def,

the

_collapse

_{and association processes manifest themselves in the} simultaneous effects on both

DO

and

_KD[n].

This could induce a cancellation between the increase in solute

_objects

and their mutual interactions within the solution.

_More,

the

manifestation of

_hydrophobic

interactions could

_explain

the difference in behavior of PPV precursor macromolecules when dissolved in methanol and in water. This

_hypothesis

(collapse

and

_eventually

association of the macromolecules at

_{high enough salt)}

_{agrees with} what occurs at

_higher

ionic

_{strength :}

_{for cs}

_larger

than

0.5 M,

a

_{precipitation}

_process of the

. solute

macromolecules takes

_{place [23-26].}

_Therefore,

one can think of the

_{precipitation}

process as a late state of the

_{collapse-association}

of the solute

_{macromolecules,}

when the salt

concentration becomes

_{high enough.}

The

_{precipitation}

does not occur in methanol for which

the solvation of NaCI levels off _{before any other}

_phenomenon

could occur

_[29].

The lack of

dependence

of the diffusion

coefficient,

for very dilute solutions

_{(cp - 0 )}

on _cs, was also observed for

_biological

macromolecules such as DNA and

_{BSA [15,}

_18,

_19].

This was

primarily

attributed to the absence of a conformational

_change

in these

_naturally

_rigid

macromolecules.

Finally,

it would be useful to _{compare the}

_hydrodynamic

size

_RH

deduced in the

_present

case to the one from methanol solutions at infinite dilution and _very

_high

ionic

_{strength [29] :}

A

_{simple comparison}

between these two sizes would consolidate the

_hypothesis

of association

discussed above. In

fact,

the value of

_RH

measured in methanol verifies the relation

&/RH ew 1.5 (RG

is the radius of

_gyration)

_[29],

and could therefore be attributed to a

_single

chain size. ’

Conclusion

Using

QELS

technique,

we scanned the

dynamics

of a flexible

polyelectrolyte

in solution.

was added to the

_system

or not. For salt-free

solutions,

the

_repulsive

electrostatic interactions

between

_{neighboring charges}

on the same macromolecule

_impart

to them an extended conformation. This was reflected in the existence of two

_independent

modes for the concentration fluctuations. The addition of counterions induces an « unusual

_{change »}

in the

solution

_properties.

The

_dynamics

are no

_longer

characterized

_by

two modes for the fluctuations. The mutual diffusion coefficient is found to

_depend

on neither the solute nor on

the salt

concentrations,

for the range of values we scanned. This

_type

of behavior is attributed

to an eventual manifestation of the

_hydrophobic

interactions,

when the ionic

strength

of the solution is raised

_{sufficiently.}

To check the

_validity

of such a

_hypothesis,

a small

_angle

neutron

scattering experiment

could

_provide

an answer.

_Using

the

_{mixing labeling technique,}

one could have access to the

_single

chain dimension in both cases : very low and

high enough

ionic

strengths. Viscosity

measurements could also

_{provide complementary}

results for the

_present

set of data.

Acknowledgments

We thank

Lark,

Inc. who

_provided

us with the materials we used. We benefited from close

cooperation

and fruitful discussion with Professors F. E.

_Karasz,

K. H.

_Langley,

L.

_Leger

and M. Daoud. This work was

_supported

in

part

by

a

_grant

from

AFOSR,

88-001.

References

[1]

TANFORD _C.,

_{Physical Chemistry}

of Macromolecules

_(Wiley,

New

_York)

1961.

[2]

MORAWETZ M., Macromolecules in

_{Solutions, High Polymers,}

Vol. 21

(Wiley,

New

York)

1975.

[3]

NAGASAWA M. and TAKAHASHI _A.,

_{Light Scattering}

From

_Polymer

_Solutions, Ed. M. B.

_Huglin

(Wiley,

New

_York)

1972, Ch. 16.

[4]

NIERLICH _M., WILLIAMS C. E., BOUE _F., COTTON J. P., DAOUD _M., JANNINK _G., PICOT _C., MOAN M., WOLFF _C., RINAUDO M. and DE GENNES P. _G., J.

_Phys.

France 40

₍₁₉₇₉₎

701.

[5]

DE GENNES P. G., PINCUS P., VELASCO R. _H., BROCHARD F., J.

_Phys.

France 37

₍₁₉₇₆₎

1461.

[6]

ISE _N., OKUBO T., Acc. Chem. Res. 13

₍₁₉₈₀₎

309.

[7]

ISE N., OKUBO T., KUNUGI S., MATSUOKA _H., YAMAMOTO K., ISHII Y., J. Chem.

Phys.

81

₍₁₉₈₄₎

3294.

[8]

MATSUOKA H., ISE N., OKUBO T., KUNUGI, S., TOMIYAMA H., YOSHIKAWA Y., J. Chem.

Phys.

83

₍₁₉₈₅₎

378.

[9]

LIN S. _C., LEE _W., SCHURR J. M.,

Biopolymers

17

₍₁₉₇₈₎

1041.

[10]

PATKOWSKI A., GULARI E., CHU B., J. Chem.

Phys.

73

₍₁₉₈₀₎

4178.

[11]

FULMER A. W., BENBASAT J. _A., BLOOMFIELD V. _A.,

_Biopolymers

29

₍₁₉₈₁₎

1147.

[12]

ELIAS J. _G., EDEN D.,

Biopolymers

20

₍₁₉₈₁₎

2369.

[13]

SCHMITZ K. S., Lu M.,

Biopolymers

23

₍₁₉₈₄₎

797.

[14]

DRIFFORD M., TIVANT _P., BENCHEIKH-LARBI F., TABTI K., ROCHAS _C., RINAUDO M., J.

_Phys.

Chem. 88

(1984)

1414.

[15]

DOHERTY P., BENEDEK G. B., J. Chem.

_Phys.

61

₍₁₉₇₄₎

5426.

[16]

NEAL D. _G., PURICH _D., CANNEL D., J. Chem.

_Phys.

80

₍₁₉₈₄₎

3469.

[17]

KIM M. K., PEIFFER D. G., J. Chem.

Phys.

83

₍₁₉₈₅₎

4160.

[18]

WANG L., Yu _H., Macromolecules 21

₍₁₉₈₈₎

3498.

[19]

NICOLAI _T., MANDEL M., Macromolecules 22

₍₁₉₈₉₎

1618.

[20]

BERNE B., PECORA R.,

Dynamic Light Scattering (Wiley,

New

_York)

1976.

[21]

DE GENNES P. _G.,

_{Scaling Concepts}

In

_{Polymer Physics (Cornell University}

_{Press, Ithaca,} New

York)

1979.

[23]

MACHADO J. _M., Ph. D. _Thesis, U. of Massachusetts, Amherst

_(1988).

[24]

GAGNON D. R., CAPISTRAN J. D., KARASZ F. E., LENZ R. W.,

Polymer

28

₍₁₉₈₇₎

567.

[25]

GAGNON D. _R., KARASZ F. E., THOMAS E. _L., LENZ R. _W.,

_Synth.

Met. 20

₍₁₉₈₇₎

85.

[26]

MACHADO J. M., DENTON III, F. R., SCHLENOFF J. B., KARASZ F. E., LAHTI P. M., J.

_Polym.

Sci. Part B :

_{Polym. Phys.}

27

₍₁₉₈₉₎

199.

[27]

GRANIER T., THOMAS E. L., GAGNON D. R., KARASZ F. E., LENZ R. W., J.

_Polym.

Sci.,

Polym.

Phys.

Ed. 24

₍₁₉₈₆₎

2793.

[28]

GRANIER T., THOMAS E. L., KARASZ F. E., J.

_Polym.

Sci., Part B :

_{Polym. Phys.}

Ed. 27

₍₁₉₈₉₎

469.

[29]

MATTOUSSI H., KARASZ F. _E., LANGLEY K. H., J. Chem.

_{Phys. (1990)}

in press.

[30]

BHATT _M., JAMIESON A. M., Macromolecules 21

₍₁₉₈₈₎

3015.

[31]

BISHOP M., LANGLEY K. H., KARASZ F. K.,

Phys.

Rev. Lett. 57

₍₁₉₈₆₎

1741.

[32]

BERRY G. C.,

Encyclopedia

of

_polymer

Science and

_Engineering,

Eds. H. Mark, C. G.