Living Radical Polymerization - American Chemical Society

C h a p t e r 23

Synthesis of Polystyrene-Polyacrylate Block Copolymers by Nitroxide-Mediated Radical Polymerization

Downloaded by UNIV OF CALIFORNIA SAN DIEGO on November 18, 2014 | http://pubs.acs.org Publication Date: August 15, 2000 | doi: 10.1021/bk-2000-0768.ch023

Sophie Robin and Yves Gnanou

1

Laboratoire de Chimie des Polymères Organiques, ENSCPB-CNRS-Université Bourdeaux 1, Avenue Pey Berland, BP 108, 33402 Talence Cedex, France

This paper discusses the conditions the best suited to the preparation of well-defined polystyrene (PS) / poly(n-butyl acrylate) (PBuA), using a βhydrogen containing phosphonylated nitroxide. Using kinetic data to compute the rates of cross-addition and propagation of the second monomer, it is demonstrated that polymerizing first n-butyl acrylate (BuA) should give rise to well-defined PBuA-b-PS copolymer samples. This prediction was experimentally confirmed subsequently. When styrene was polymerized first, the copolymer formed was contaminated with a substantial amount of residual PS macroinitiator : the difference between the rates of cross-addition and propagation of BuA resulted in a fast growth of those of the PBuA blocks that were initiated, causing the medium to partition in mesophases with the residual PS precursor entrapped in the monomer-poor phase.

Introduction Nitroxide-Mediated Radical Polymerization currently enjoys renewed impetus after it was discovered that several families of vinylic monomers can be polymerized under living/controlled conditions, provided the nitroxide used to control propagation is adequately chosen (1,2,3,4,5). For instance, (N-ter^butyl-N-(l-diethylphosphono-2,2dimethyl)propylnitroxyl) (DEPN) is a β-hydrogen bearing nitroxide that was found to bring about the controlled polymerization of both styrene and alkyl acrylates when associated with AIBN as initiator (6). Other β-hydrogen containing nitroxides were recently shown suitable for the controlled radical polymerization of various vinylic monomers, including styrene, alkyl systems lies in the frequency of the deactivation/activation process that depends both on the rate of coupling of propagating 'Corresponding author.

334

© 2000 American Chemical Society

In Controlled/Living Radical Polymerization; Matyjaszewski, K.; ACS Symposium Series; American Chemical Society: Washington, DC, 2000.

335 radicals with nitroxides and on the acrylates, methacrylates, and 1,3 dienes (7,8,9). The key to the control of such rate of homolysis of the C—ON bond formed. When used to polymerize styrene or alkyl acrylates, the AIBN/DEPN system not only provides high equilibrium constants (K=k /k ) -reflected in a fast propagation of the two monomersbut it also sets three decade difference between the rate constants of recombination and propagation, affording samples of particularly narrow polydispersity. There lies the beauty of polymerizations mediated by DEPN.


d

rec

Notwithstanding the fact that the same initiator (AIBN)/nitroxide (DEPN) system induces totally different kinetics with styrene and η-butyl acrylate, the controlled character of their respective polymerization prompted us to investigate their block copolymerization by sequential addition (10). This paper thoroughly discusses the constraints associated with such synthesis and discloses the conditions the best suited to obtain well-defined PS/PBuA block copolymer samples. In particular, it demonstrates that the order of addition of monomers matters to achieve high blocking efficiency.

Results and Discussion

Kinetic Predictions as to the Order of Polymerization of Monomers For a successful synthesis of diblock copolymers by controlled radical polymerization, it is essential to retain the chain end functionality of the first block and achieve an efficient cross-propagation. With respect to the first factor, that concerns the production of precursors containing a minimal amount of dead chains, all precautions should be taken to optimize propagation over termination or transfer reactions. Since the ratio of the probability of termination to that of propagation (k [P*]/k [M]) increases with the monomer conversion, it is crucial to discontinue the polymerization of the first monomer before its total consumption so as to prevent too high a concentration of dead "precursor" chains. As to the blocking efficiency, it not only depends on the rate constants of crossaddition and on the propagation rate constants of the two monomers, but it is also subordinated to the equilibrium constants (K=k /k ) between their respective dormant and active species (Scheme 1). In contrast to K, the reactivity ratios of these two monomers and their rate constants of propagation can be found in the literature (11,12,13). Though important, the knowledge of these values has only an indicative value that can well be used to choose the monomer to polymerize first. t

d

rec


p

336 Indeed, polymerizing the monomers in the order of decreasing rate constants of propagation does not necessarily entail a fast initiation of the second block by the first one. To bring about a complete cross-addition and thus achieve high blocking efficiency, one has to also consider the equilibrium constants for the two polymerizing systems and the concentration of polymeric radicals [P*] they imply for a given concentration of dormant species.


Dead Chains

PVl—CH2—CH-OL

PV-PB^-C^-CH-a Β

Dead Chains

Scheme 1 : Reactions occurring while synthesizing block copolymers by nitroxidemediated polymerization (A,B= Ρh or C02Bu). From the knowledge of [P*] as a function [P-DEPN], the initial rates of crossaddition and propagation of the 2 block can be easily deduced. At the very onset of the polymerization of the second monomer, the concentration of the two kinds of radicals should be quite contrasted, so that initiation should be initially favored over propagation. However, this situation cannot prevail for ever ; as conversion increases, the concentration of the first block is bound to continuously decrease -and alternately that of the 2 block to progressively increase-, entailing a constant variation of the concentration of the two kinds of polymeric radicals ([PS*], [PBuA*]). Given the profiles of the concentration of the two species, the rate of cross-addition is bound to be overtaken by the rate of propagation but the question is whether this outrunning occurs soon or late in the conversion of the 2 monomer. Should the rate of propagation overshadows that of cross-addition while much of the first block is not yet to decompose, nd

nd

nd


337 the resulting copolymers are likely to be contaminated with a substantial amount of residual precursor chains. To construct the [P-DEPN] versus [P*] curve, we referred to our previous work on the kinetics of DEPN-mediated polymerizations of styrene and η-butyl acrylate (14). Using PS and PBuA oligomers fitted with end-standing DEPN-based alkoxyamines, the polymerization of styrene and η-butyl acrylate was triggered at 120°C and their conversion followed as a function of time. From the ln([M] /[M]) versus time plot, the corresponding [P*] could be easily deduced, using k values found in the literature (77) and the relation proposed by Fischer (75) for the slope of the linear variation. In that work, Κ values could also be determined upon concomitantly following the conversion in monomer and the evolution of free [DEPN], Κ being equal to K=[P*] *[DEPN] /[PDEPN] as demonstrated by Fischer (75). Based on this methodology, the concentration of polymeric radicals ([PS*], [PBuA*]) were computed as a function of the corresponding alkoxyamines and the curves subsequently drawn for the two systems. Because Κ dramatically varies from one monomer to another one ( K = 3 . 5 1 0 " mol.l" , K A 1 - 2 - 1 0 " mol.l" (16)), a same concentration of dormant species is found to afford quite different values of [PS*] and [PBuA*] as expected (Figure 1). 0


p

t

t

0

9

1

=

S

1 0

BU

1

•

0

0.01

0.02

0.03

0.04

0.05

[PS-DEPN] (mol/l)

0

0.005

0.01

0.015

0.02

0.025

[PABu-DEPN] (mol/1)

0

0

e

Figure 1: Evolution of[P ] with the corresponding initial concentration of [P-DEPN]o, in the case ofPS-DEPN (a) andPBuA-DEPN (b) as macroinit

All data are thus available to compute the rate of cross-addition (RCA) as a function of [P-DEPN] and compare the value obtained with that of propagation (R ) for a same experiment (Figure 2). p


338 ^CA

— ^Α,Β

R

=k ρ,Β [ ρ ; ] [Β ]

n


Obviously, the two rates are set to vary oppositely and thus to cross each other, this particular point being not only determined by the respective Κ and [Ρ'] values of the two systems, but by the initial concentration of dormant species [P-DEPN] as well. Figure 2 and Figure 3 show the evolution of the rates of cross-addition and propagation with the concentration of residual precursor when PBuA and PS-based alkoxyamines were used to polymerize styrene and «-butyl acrylate, respectively. Several remarks can be inferred from these two figures. To prepare diblock copolymers based on PS and PBuA, it is recommended first to polymerize nBuA and subsequently grow the PS block from the formed PBuA. For instance, the use of an initial concentration of 7.8· 10" mol.l" for the [PBuA-DEPN] precursor implies that only 4% of this macroinitiator will be present in the medium at the time the rate of propagation outruns that of cross-addition (Figure 2). Not only is this order of polymerization of monomers particularly favorable to the obtainment of copolymers with low compositional heterogeneity, but it also allows to grow rather short PS blocks from PBuA blocks. 3

1

Quite different is the situation that prevails when [PS-DEPN] is used to grow PBuA as second block (Figure 3). Notwithstanding a much higher concentration of polystyryl radicals (Figure 1), the sluggishness of the latter species to addrc-butylacrylate is the major cause for the slow rate of initiation observed. Whatever the initial concentration of PS-DEPN macroinitiator, the rate of propagation of w-butyl acrylate indeed outruns that of cross-addition while about 50% of PS precursor chains are still present in the reaction medium. This order of polymerization of monomers with PS as first block therefore appears not suited to the production of samples that are homogeneous in size and composition.

! »

» Rca/[M]

2

ι Rp/[M] ; [PBuA-DEPN]o=0.0039 mol/1

, Rp/[M] ; [PBuA-DEPN]o=0.0078 mol/1

ίο

Ï

, Rp/[M] ; [PBuA-DEPN]o=0.0019 mol/1

0

0.001

0.002

0.003

0.004

0.005

0.006

0.007

0.008

0.009

0.01

[PABu-DEPN] (mol/1)

Figure 2 : Evolution of the rates of cross-addition (RCA/[M]) and propagatio ofstyrene as a function of [PBuA-DEPN] , used as macroinitiator. 0


339 However, it can be seen from Figure 3 that the differences between the rates of cross-addition and propagation tend to narrow with low initial concentrations of PS macroinitiator, meaning that PBuA with a large size might be grown with little amount of PS-DEPN precursor left aside. ο Rca/[M] B

Rp/[M] ; [PS-DEPN]o=0.0088 mol/1

A

Rp/[M] ; [PS-DEPN]o=0.0044 moll


• Rp/[M] ; [PS-DEPN]o=0.0022 mol/1

a

3

I"a &

2

0

0.001

0.002

0.003

0.004

0.005

[PS-DEPN]

0.006

0.007

0.008

0.009

0.01

1

(mol.l*)

Figure 3 : Evolution of the rates of cross-addition (RCA/[M]) and propagatio (R ,BUA/[M]) of η-butyl acrylate as a function of [PS-DEPN], used as macroin 0

P

To our knowledge, it is the first time that such an attempt is made to predict the outcome of a block copolymer synthesis mediated by nitroxides. However, one cannot disregard the fact that the model used to describe the kinetics of initiation and propagation may be oversimplified. In our reasoning, we have assumed the two operating systems -PS*, PBuA* and DEPN in equilibrium with their corresponding dormant species- to behave independently one from the other which might not be the reality. Since each system sets its own level of nitroxides in excess, the active species that is known to release more free nitroxide -here PS*- may force the other kind of radicals to recede to lower concentrations than they would have taken in simple homopolymerizations. Since the direct measurement of the concentration of the two kinds of radicals in such copolymer synthesis is not possible, we had no other option than to rely on our previously described reasoning to account for the results experimentally obtained.

Comparison of Experimental Results with Kinetic Predictions

Synthesis of PBuA-b-PS Copolymer with PBuA as First Block The synthesis of the first PBuA block (PBuA-DEPN) was obtained under conditions previously described (10,14) ; a sample of 50,000g.mol' molar mass was targeted, AIBN being used as initiator and DEPN as stable radical in a 1:2.5 ratio. As previously 1


340 mentioned, the polymerization was discontinued before the total consumption of monomer and a polymer exhibiting a polydispersity of 1.2 and 41,000g.mol" as molar mass was eventually obtained after 9.5h of reaction at 120°C (Table I). The efficiency of the initiation step was found equal to 0.85 ; only a larger excess of DEPN could have afforded an efficiency close to 1 as previously demonstrated by Benoit et al. (14), but it would have unnecessarily slowed down the rate of polymerization. Another sample of PBuA-DEPN was prepared from the following alkoxyamine CH CH(COOCH )-DEPN : polymerization proceeded at much faster rate in this case because a smaller excess of free DEPN (0.07 instead of 0.25 in the case of AIBN as an initiator) was used as compared to the previous case. The same molar mass of 50,000 g.mol" was targeted and like previously the polymerization was discontinued at 70% conversion. Before adding styrene to grow the 2 block, residual nBuA was stripped off in the reaction medium containing the PBuA-DEPN and the mixture was heated to 120°C. 1


3

3

1

nd

Table I : Characteristics of PBuA(40,000g/mol)-PS(50,000g/mol) Diblock Copolymers

Exp.

Step (g/mol) w

1.1 1.2 1.3 1.4

PBuA PBuAPS PBuA PBuAPS (7)

35,500 53,700 34,000 85,000

(1)

Time (h) 9.5 0.75 3 2

AW*

f>

Ρ

Λ

(%) (g/mol) 71 19 68 35

41,700 55,500 31,500 87,600

(2)

0.85 0.97 1.0 0.97

1.2 1.3 1.2 1.2 (3)

Calculatedfromconversion. Conversion measured by gravimetry. Determined by SEC using RI detection and PS standards. Polydispersity index obtained from SEC. Initiation efficiency calculated as Μ / Μ Initiation by AIBN with [DEPN]/[AIBN]=2.5. Initiation by the alkoxyamine mentioned in the text with [DEPN]/[Alkoxyamine]=0.07. (4)

(5)

(6)

ηΛ

η6χρ

( 7 )

The polymerization was purposely discontinued at low conversion in styrene to check whether PS block of short size could be grown in a homogeneous way (Table I). As shown in Figure 4, the size exclusion chromatography traces of the two copolymers obtained are found to be entirely shifted towards the higher molar mass region upon superimposing them with that of the precursor.

Figure 4 : SEC traces ofPBuA-b-PS diblock copolymers (1.1,1.2 and 1.3,1.4, Table I).


341 No tailing could be detected in either SEC traces, confirming that all PBuA precursor chains efficiently initiated the polymerization of styrene. For instance, a PS block of 20,000 g.mol" could be attached with perfect efficiency of the initiation step. Another piece of evidence of the excellent definition of the PBuA-b-PS block copolymer formed is the perfect fit between UV and RI detections : this bears out the fact that these samples exhibit high homogeneity in composition. 1

Synthesis of PS-b-PBuA Copolymer with PS as First Block As we were mainly interested in PS-b-PBuA copolymers with a PS block of short size, a sample of PS-DEPN 20,000g.mol" molar mass was synthesized. Using [DEPN] and [AIBN] in a 2.5 molar ratio, this macroinitiator (PS-DEPN) could be obtained with an excellent efficiency (f=1.0) and a narrow molar mass distribution (I =1.2) (14).


1

p

Table II : Characteristics of PS(20,000g/mol)-PBuA(100,000g/mol) Diblock

Copolymers

Exp. 2.1 2.2

Step W

PS PS-PBuA (1)

Time (g/mol) (h) 18,800 5.5 104,900 6

j (V (g/mol)

AW" (%)

18,900 65,900

94 86

(2)

}

f

h

1.2 2.5

1.0

(3)

Calculated from conversion. Conversion measured by gravimetry. Measured by SEC using RI detection and PS standards. Polydispersity index obtained from SEC. Initiation efficiency calculated as M / M Initiation by AIBN with [DEPN]/[AIBN]=2.5. (4)

(5)

( 6 )

n t h

n e x p

Once the residual styrene was stripped off, the PS-DEPN containing medium was charged withfreshw-butyl acrylate and heated to 120°C. The sample obtained after 6h of reaction and 86% conversion was characterized by SEC (Figure 5).

-•

ml

Figure 5 : SEC traces of PS-b-PBuA diblock copolymer (2.1 and 2.2, Ta function of nBuA conversion.

As anticipated, this copolymer was contaminated with a non negligible amount of residual PS precursor chains (18%), and its molar mass distribution was broad owing to a slow cross-addition compared to the propagation of «-butyl acrylate. Even though this overall behavior was forecast by our kinetic calculations, we were intrigued by the



342 presence of such a large amount of PS homopolymer in the resulting material since a PBuA block of large size was targeted. Using the UV detection of the SEC line, aliquots sampled out at various times were characterized and compared. As a matter of fact, after experiencing a rapid initial decrease, the peak due to residual PS-DEPN was found to remain constant with respect to that of the copolymer, as soon as the conversion in nBuA reaches 30%. Approximately for that conversion monomer, the reaction medium was noticed to become strongly turbid, denoting the occurrence of a phase separation. If the disappearance of PS-DEPN suddenly levelled off after crossing a certain monomer conversion, this means that the medium underwent a strong partition induced by the phase separation, causing w-butyl acrylate to be unevenly distributed in the medium. Because the initiated PBuA blocks grew fast to large size, the incompatibility between PS and PBuA that has resulted might have driven the residual PS-DEPN within PS-rich particles ; actually, the latter did not coalesce because of their steric stabilization by the PBuA blocks of PS-b-PBuA copolymers. Owing to the preference of «BuA to stay in the poly(BuA)-rich phase, residual PS-DEPN chains had therefore little chance to add this monomer. Since it was previously demonstrated that the difference in the rates of crossaddition and propagation tends to narrow with smaller initial concentration of PS-DEPN, poly(BuA) blocks of molar masses much larger than 100,000 g.mol" were targeted (Table III), the polymerization being discontinued at relatively low conversion. As in the previous case, one can notice the presence of residual PS-DEPN in approximately same proportions (16%), confirming that the partition of the medium hindered the access of the monomer to PS-DEPN. 1

Table III : Characteristics of PS(20,000g/mol)-b-PBuA(variable size aimed) Diblock Copolymers using varying [PS-DEPN], Exp

Time (h)

A-W"

(g/mol) 3.1 3.2 3.3

(6)

(

106,000 84,900 155,000 470,000

/>

(%)

6 3 4 12

86 86 27 90

Τ (4) P

AW" (g/mol)

l

2.5 2.0 1.6 2.0

65,900 40,400 74,800 324,000

(2)

18 30 16 16 (3)

^Calculated from conversion. Conversion measured by gravimetry. Measured by SEC using RI detection and PS standards. Polydispersity index . Remaining macroinitiator estimated by SEC using UV detection.** The PS-DEPN macroinitiator was precipitated in methanol. (4)

(5)

6

In an attempt to modify the ratio of the rate of cross-addition to that of propagation, a series of experiments were carried out at temperatures lower than 120°C (Table IV). Because the various reactions involved in such block copolymer synthesis have different energies of activation, one might expect to favor one step (initiation) over the other one (propagation) by decreasing temperature.


343 1

1

Table IV : Characteristics of PS(20,000g.mol )-b-PBuA(100,000g.mor ) Diblock Copolymers obtained at various temperatures Time (g/mol) (h) (%) 106,000 6 86

Exp 4.1 4.2 4.3

93,400 42,400

133 253

(JJ

R

(g/mol)

71 20

18 12 14

120 95 80

2.5 1.3 1.2

65,900 67,000 32,900

(1)


Temperature(V CO

j(4)

(2)

(3)

Calculated from conversion. Conversion measured by gravimetry. Measured by SEC using RI detection and PS standards. Polydispersity index . ^ Temperature of polymerization of the second block. Remaining macroinitiator estimated by SEC using UV detection. (4)

5)

{6)

The results obtained were rather disappointing since no significative change could be perceived in the amount of residual DEPN. To better demonstrate the interrelation between the incomplete consumption of PS-DEPN and the incompatibility that arises in the reaction medium because of a too slow cross-addition, PS-DEPN of shorter size shorter than 20,000 g.mol" were used as macroinitiators. 1

1

Table V : Characteristics of PS(variable size aimed)-b-PBuA(20,000g.mor ) Diblock Copolymers J(4) R Μ Exp. Γ block 2" block Time (5) n,exp size size (g/mol) (h) (%) (g/mol) (gmot ) (gmot ) 20,800 1.5 5.1 3,900 20,000 1 93 22,500 1.4 23,600 5.2 90 4,300 20,000 5 22,300 a

μ )

i¥A

1

5.3 5.4

7,000 20,000 (1)

1

20,000 100,000

21,400 104,900

70 86

9 6

28,200 65,900

12 18

1.3 2.5

(2)

(3)

Calculated from conversion. Conversion measured by gravimetry. Measured by SEC using RI detection and PS standards. Polydispersity index . Remaining macroinitiator estimated by SEC using UV detection. (4)

(5)

As the incompatibility between polymeric chains and thus the order-disorder transition of the corresponding block copolymers are governed by the product χ · Ν (17), (χ being the interaction parameter and Ν the overall degree of polymerization), two different polymers, each of them of small size should be less prone to separate in distinct phases. When PS-DEPN of 3,000 to 4,000 g.mol" molar mass were used as macroinitiator, no phase separation could be actually detected and as a result no unreacted macroinitiator remained in the reaction medium after growing the poly(BuA) block. The slow cross-addition step was mirrored in rather large polydispersities (1.41.5), but excellent control over the sample molar mass could be achieved. In contrast, the use of PS-DEPN precursor of slightly higher molar mass (7,000g.mol' ) resulted in the occurrence of phase separation and the presence of unreacted precursor chains at the time the polymerization of BuA was discontinued. 1

1



344

Figure 6 : SEC traces ofPS-b-PBuA using PS-DEPN precursor of short size (5.1 and 5.2, Table V). Another means to demonstrate that the incompatibility between growing poly(BuA) and remaining PS-DEPN prejudices the consumption of the latter was to grow a statistical P(S-co-BuA) block in between PS and PBuA blocks. Instead of thoroughly stripping off residual styrene after the growth of the first PS-DEPN block, w-butyl acrylate was directly introduced in the reaction medium that was then heated to 120°C.

_1

1

Table VI : Characteristics of PS(20,000g.mol )-b-P(S-co-BuA) (lOO^OOg.mol ) Copolymers

Exp. 6.1 6.2 6.3

st

1 block size (g/mol) 7,600 10,100 18,300

Time (h)

(g/mol) 73,200 79,200

(I)

10.5 10.5 9.5

S> (%)

{4)

I (g/mol) 40,700 61,900

64 66 63

(2)

Living Radical Polymerization - American Chemical Society

Recommend Documents