Inisurfs : surface-active initiators : their synthesis and application in emulsion polymerization Kusters, J.M.H.
DOI: 10.6100/IR410004 Published: 01/01/1994
Document Version Publisher’s PDF, also known as Version of Record (includes final page, issue and volume numbers) Please check the document version of this publication: • A submitted manuscript is the author’s version of the article upon submission and before peer-review. There can be important differences between the submitted version and the official published version of record. People interested in the research are advised to contact the author for the final version of the publication, or visit the DOI to the publisher’s website. • The final author version and the galley proof are versions of the publication after peer review. • The final published version features the final layout of the paper including the volume, issue and page numbers.
Link to publication
Citation for published version (APA): Kusters, J. M. H. (1994). Inisurfs : surface-active initiators : their synthesis and application in emulsion polymerization Eindhoven: Technische Universiteit Eindhoven DOI: 10.6100/IR410004
General rights Copyright and moral rights for the publications made accessible in the public portal are retained by the authors and/or other copyright owners and it is a condition of accessing publications that users recognise and abide by the legal requirements associated with these rights. • Users may download and print one copy of any publication from the public portal for the purpose of private study or research. • You may not further distribute the material or use it for any profit-making activity or commercial gain • You may freely distribute the URL identifying the publication in the public portal ? Take down policy If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim.
Download date: 28. Jan. 2017
INISURFS: SURFACE-ACTIVE INITIATORS THEIR SYNTHESIS AND APPLICATION IN EMULSION POLYMERIZATION
PROEFSCHRIFT
ter verkrijging van de graad van doctor aan de Technische Universiteit Eindhoven, op gezag van de Rector Magnificus, prof. dr. J.H. van Lint, voor een commissie aangewezen door het College van Dekanen in het openbaar te verdedigen op dinsdag 25 januari 1994 om 16.00 uur door
JOSEPH MARIA HUBERTUS KUSTERS geboren te Nieuwenhagen
druk: wîbro dissortatiodrukkcrij, he!mond
Dit proefschrift is goedgekeurd door
de promotoren:
prof. dr. ir. A.L. Oerman prof. dr. R.G. Gilbert
en de copromotor:
dr. J.J.G.S. van Es
------- - - - - -
INISURFS: SURFACE-ACTIVE INITIATORS THEIR SYNTHESIS AND APPLICATION IN EMULSION POLYMERIZATION
CIP-DATA KONINKLIJKE BIBLIOTIIEEK, DEN HAAG
Kusters, Joseph Maria Hubertus
Inisurfs: Surface-Active Initiators Their Synthesis and Application in Emulsion Polymerization I Joseph Maria Hubertus Kusters. Thesis Eindhoven. - With ref. With a summary in Dutch. ISBN 90-386-0153-0 Subject heading: emulsion polymerization.
Aan mijn ouders, To Jola
Summary
VII
Summary
Synthetic latices produced via emulsion polymerization find a broad range of applications in for example paper, textile, adhesive, and paintlcoating industries. The basic recipe to make these latices usually contains one or more surfactants of possibly different types. These surfactants can cause difficulties in the application of a latex. The stability of the latex can change over time due to desorption and adsorption of the surfactants. When applying the latex, which quite often means forming a polymer film on a substrate, the surfactant may migrate to the surface and or to the interface of the polymer film, which leads to poor mechanica} properties, low water resistance and also to blooming and blushing.
The main
solution to these problems is fixation of the surfactant onto the surface of the polymer latex particle.
There are three different ways to achieve this fixation:
1) use of a surface-active monomer (surfmer), 2) use of a surface-active chain transfer agent (transsurf), and 3) use of a surface-active initiator (inisurf). This thesis describes the results of the investigations on the synthesis of inisurfs and on the application of these inisurfs in the emulsion polymerization of styrene.
Hereby, the capability of the inisurfs to act as initiator as well as
surfactant could be determined. Moreover, such a study yields more insight into the parameters governing the formation of (surface-active) radicals at the surface of the polymer particles, and on the ways exit and entry are influenced by these radicals.
Chapter 1 gives a general introduetion to polymers and polymerization and
it presents a background for this investigation, as well as a bistorical overview of the obtained results in the field of inisurfs. stated.
The aims of the investigation are
VliJ
Summary
In Chapter 2 the choice of the initiator and surfactant constituent parts of the inisurf and the metbod of linking them is corroborated.
The initiator moiety
was of the azo-type, with one or two carboxyl groups. The surfactant moiety was a nonionic poly(ethylene oxide), having one hydroxyl terminal group and optionally
either a methyl or a nonylphenyl ether end-group. The results of the
esterification experiments, as well as the synthesis of an asymmetrically substituted azo-initiator are described.
In Chapter 3 a brief overview of free radical polymerization is given. Next,
qualitative and quantitative roodels for emulsion polymerization are
discussed.
In Chapter 4 the various experimental and analytica! techniques used
in performing an emulsion polymerization and characterizing the formed latex are given.
A survey of the performance of the various synthesized inisurfs in the ab
initio emulsion polymerization of styrene is given in Chapter 5.
The colloidal
stability of the reaction mixtures was positively influenced by lengthening the ethylene oxide chain and by the presence of a nonylphenyl end group in the inisurf. The results also indicated that the recombination of newly formed free radicals was of great influence on the reaction rate and the nucleation of particles. This effect was more pronounced for symmetrical inisurfs.
From the range of synthesized inisurfs two inisurfs were selected, a symmetrical one (SC0-880) and an asymmetrical one (AC0-880), and these were studied in more detail, as described in Chapter 6.
The dissociation rates of the
inisurfs were measured (kti (323 K): 6.3 10'6 s· 1 and 9.3 10·7 s·', respectively) and showed that there was only a minor effect due to the attachment of the surfactant. The measured CMCs (2.0 10·5 mol dm· 3 and 6.3 10·4 mol dm·l, respectively) were similar to those reported for nonionic surfactants.
The reaction rates of these
inisurf systems and the PSDs at the end of the reaction deviated from those obtained for conventional systems.
This was due to differences in stahilizing
Summary
IX
ability and to the occurrence of geminate recombination.
The above mentioned inisurfs have also been used in seeded experiments, as discussed in Chapter 7.
Both inisurfs showed similar behaviour.
Due to the
presence of a steric stabilizer at the surface of the polymer partiele the exit rate of a radical from the partiele (k: 7.4 10·5 s· 1 and 5.0 104 s· 1, respectively) was smaller than in systems with anionic stabilization. The entry rate of a radical (p: 2 10·5 s· 1 and 2
w-s s·
1 ,
respectively) was very low as was the efficiency (j: 2
w-4 and 1 10"3 ,
respectively). This could be readily explained qualitatively and quantitatively by comparison of the time scale for geminate recombination of the two free radicals, formed by inisurf decomposition, with the time scale for escape by diffusion of one of the two free radicals from the vicinity of the partiele.
In the case of the
symmetrical inisurf there was hardly any effect of surface coverage of inisurf on efficiency, where as in the case of the asymmetrical inisurf, the efficiency increased with increasing surface coverage. This investigation bas shown that the efficiency of inisurfs will be low, as long as it is not possible to separate the two newly formed free radicals at the surface of the polymer particles in an effective manner.
This thesis concludes with an Epilogue, in which the most important conclusions are summarized and some incentives are presented for further research.
x
Samenvatting
Samenvatting
Synthetische latices gevormd met behulp van emulsiepolymerisatie vinden een brede toepassing in bijvoorbeeld de papier, textiel, lijm en verf industrie. Het basisrecept voor deze latices bevat gewoonlijk één of meer emulgatoren van mogelijkerwijs verschillend type. De aanwezigheid van deze emulgatoren kan tot problemen leiden bij het gebruik van de latex. De stabiliteit van de latex kan in de tijd veranderen doordat de emulgatoren adsorberen en desorberen. Bij toepassing van de latex, wat meestal gebeurt in de vorm van een polymere film op een substraat, zal de emulgator naar het oppervlak en/of het grensvlak diffunderen, hetgeen leidt tot slechte mechanische eigenschappen, lage water resistentie en ook tot blooming en blushing. Een oplossing voor dit probleem is het binden van de emulgator aan het oppervlak van de polymeerdeeltjes.
Deze fixatie kan op drie
manieren bereikt worden: 1) via oppervlakte-actief monomeer (surfmer), 2) via oppervlakte-actief chain transfer agent (transsurf) en 3) via oppervlakte-actieve initiator (inisurf). Dit proefschrift beschrijft de resultaten van een onderzoek naar de synthese van inisurfs en de toepassing van deze inisurfs in de emulsiepolymerisatie van styreen, om zo het gedrag van inisurfs als initiator en als emulgator vast te kunnen stellen. Bovendien geeft zo'n studie meer inzicht in de factoren die de vorming van (oppervlakte-actieve) radicalen aan het oppervlak van deeltjes beïnvloeden en de manier waarop exit en entry hierdoor beïnvloed worden.
Hoofdstuk 1 geeft een algemene inleiding in polymeren en polymerisatie. De achtergrond van dit onderzoek, alsmede een historisch overzicht van de verkregen resultaten op het gebied van inisurfs worden gegeven en de doelen van het huidige onderzoek worden geformuleerd.
XI
Samenvatting
In Hoofdstuk 2 wordt de keuze van het initiator- en emulgatordeel van de inisurf en de methode van koppeling beargumenteerd. azo initiator met één of twee carboxyl groepen.
Het initiatordeel was een
Het emulgatordeel was een
nonionisch polyetheenoxide met een primaire hydroxyl groep en met als eindgroep De resultaten van de esterificatie
een methyl of een nonylphenyl groep.
experimenten en van de synthese van een asymmetrisch gesubstitueerde azoinitiator zijn eveneens beschreven.
In Hoofdstuk 3 wordt een overzicht gegeven van vrije radicaal polymerisatie. Daarnaast worden ook qualitatieve en quantitatieve modellen voor emulsiepolymerisatie besproken.
In Hoofdstuk 4 worden de verschillende
experimentele en analytische technieken besproken welke zijn gebruikt voor het uitvoeren van de emulsiepolymerisaties en het karakteriseren van de gevormde latex.
Een overzicht van het gedrag van de verschillende inisurfs in de ab initio emulsiepolymerisatie van styreen wordt in Hoofdstuk 5 gegeven.
Door het
gebruik van een langere polyetheenoxide keten kon de colloïdale stabiliteit van het reactiemengsel positief beïnvloed worden.
De resultaten lieten ook zien dat
recombinatie van nieuw gevormde vrije radicalen van belang is voor de reactiesnelheid en voor het vormen van nieuwe deeltjes.
Dit effect was meer
uitgesproken in het geval van de symmetrische inisurfs.
Van de groep van gesynthetiseerde inisurfs werden er twee gekozen, een symmetrische (SC0-880) en een asymmetrische (AC0-880), die in meer detail onderzocht zijn, zoals beschreven in Hoofdstuk 6. De dissociatiesnelheid van de inisurfs is gemeten (kd (323 K) is respectievelijk 6.3 10-6 s- 1 en 9.3 I0-7 s· 1) en liet zien dat er slecht een gering effect was van de aanhechting van de emulgator. De gemeten CMC's (respectievelijk, 2.0
w-s
mol dm- 3 en 6.3 I0-4 mol dm' 3) bleken
ongeveer gelijk te zijn aan die welke in de literatuur worden gevonden voor niet-ionische emulgatoren.
De reactiesnelheid in deze inisurfsystemen en de
XII
Samenvatting
deeltjesgrootteverdeling aan het einde van de reactie verschilden van die verkregen met conventionele systemen. Dit verschil werd toegeschreven aan een verschil in stabiliserende werking en aan het optreden van geminate recombinatie.
De eerder genoemde inisurfs werden ook gebruikt in seeded experimenten, zoals beschreven in Hoofdstuk 7.
Beide inisurfs vertoonden soortgelijk gedrag.
Ten gevolge van de aanwezigheid van een sterische stabilisator op het oppervlak van de polymeerdeeltjes was de snelheid van uittreden van een radicaal uit deze deeltjes (respectievelijk, k: 7.4 10"5 s·l en 5.0 10"4 s" 1) kleiner dan in systemen met een anionische stabilisator.
De snelheid voor het intreden van een radicaal
1
(respectievelijk, p: 2 10· s· en 2 10·5 s· 1) was erg klein en dientengevolge ook de 5
efficiëntie (respectievelijk, f 2 10·4 en 1 10-3). Dit kon qualitatief en quantitatief verklaard worden door de tijdschaal voor geminate recombinatie van de twee vrije radicalen gevormd door dissociatie van de inisurf te vergelijken met de tijdschaal voor ontsnapping door diffusie van één van de twee radicalen. In het geval van de symmetrisch inisurf had een toename van de bedekkingsgraad met inisurf bijna geen effect, terwijl in het geval van de symmetrische inisurf wel een toename van de efficiëntie gezien werd met toenemende bedekkingsgraad.
Het uitgevoerde
onderzoek heeft laten zien dat de efficiëntie van de inisurfs laag zal blijven zolang de twee pasgevormde vrije radicalen aan het oppervlak van de polymeerdeeltjes niet op een effectieve manier gescheiden worden.
Dit proefschrift wordt afgesloten met een Epiloog. Hierin worden de belangrijkste conclusies samengevat en worden enige aanzetten gegeven voor verder onderzoek.
Contents
XIII
Contents
Summary
VII
x
Samenvatting
XIII
Contents
Chapter 1 1.1
1.2 1.3 1.4 1.5
General Introduetion Background of this Investigation Ristorical Overview Aim of this Investigation Survey of this Thesis References
Chapter 2 2.1 2.2
2.3
2.4 2.5
3.4 3.5
4
6 8 10 11
Synthesis
Introduetion Choice of Intitiator and Surfactant 2.2.1 Initiator 2.2.2 Surfactant Results and Discussion 2.3.1 Asymmetrical Azo-Initiator 2.3.2 lnisurf Conclusions Experimental Section References
Chapter 3 3.1 3.2 3.3
Introduetion
13 14 14
17 19 20
21 25 25
29
Theory of Emulsion Polymerization
Introduetion Free-Radical Polymerization Model of Harkins The Rate of Polymerization The Smith and Ewart Theory 3.5.1 The Recurrence Relationships
31 32 35 38 40 40
XIV
3.6 3.7
3.8
Contents
3.5.2 Solution to the Recurrence Relationships 3.5.3 Partiele Number Nucleation Models Zero-One System 3.7.1 Population Balance Equations 3.7.2 Slope and Intercept Metbod 3.7.3 Interval lil Kinetics The Fate Parameter References
Chapter 4 4.1 4.2
4.3 4.4
4.5
4.6
4.7
5.1 5.2
Experimental Procedures
Introduetion Emulsion Polymerization Reactions 4.2.1 Reactor Design 4.2.1.1 Batch reactor 4.2.1.2 Dilatometer 4.2.1.3 Densimeter 4.2.2 Ah lnitio Reactions 4.2.3 Seeded Reactions Seed Preparation Conversion Measurements 4.4.1 Gravimetry 4.4.2 Di1atometry 4.4.3 Densimetry Partiele Size Measurements 4.5.1 Dynamic Light Scattering 4.5.2 Transmission Electron Microscopy 4.5.3 Disk Centrifuge Photosedimentometry 4.5.4 Partiele Nurnber Concentration Surface Tension Measurements 4.6.1 Du Nouy Ring-Metbod 4.6.2 Maximum Pressure Bubble Tensiometry Initiator Decomposition Measurements References
Chapter 5
44 46 47 49 50 51 53 55 57
59 60 60 60 60 62 64 64 66 68 68 69 70 71 72 72 73 74 75 75 75 76 77
The Application of Inisurfs
Introduetion Syrnmetrica1 Inisurfs 5.2.1 Ah lnitio Reactions with Inisurf SE0-350 (7a) and SE0-550 (7b) 5.2.2 Ah lnitio Reactions with lnisurf SC0-630 (7c)
81 83 83 84
XV
Contents
5.3
5.4
5.2.3 Ab Initio Reaction with Inisurf SC0-880 (7d) Asymmetrical Inisurfs 5.3.1 Ab Initio Reactions with Inisurf AE0-550 (7e) 5.3.2 Ab Initio Reactions with Inisurf AC0-630 (7f) 5.3.3 Ab Initio Reaction with Inisurf AC0-880 (7g) Discussion References
Chapter 6
6.1 6.2 6.3 6.4 6.5
7.1 7.2 7.3 7.4
92 92 95
Inisurfs with Antarox C0-880 as Surfactant Moiety
Introduetion Dissociation Rate Critical Micelle Concentration Ab Initio Reactions with SC0-880 and AC0-880 Molecular Weight References
Chapter 7
88 88 89 89
97 98 100 104 109 110
Kinetics of Partiele Growtb in Seeded Emulsion Polymerization witb Inisurfs
Introduetion Symmetrica1 Inisurf Asymmetrical lnisurf Discussion References
114 116 123 126 134
Epilogue
137
Glossary of Symbols
141
Acknowledgement
145
Curriculum Vitae
147
XVI
Chapter 1
Introduetion
Summary: In this chapter a general introduetion to polymers and polymerization wilt be given. Elaborating on emulsion polymeriza· tion, the various problems associated with the use of Latices wiJl be discussed, as welt as the solutions of!ered. A historica/ overview including the work done by other groups will he given, focusing on the use of surface-active initiators: "inisurf'. Following this, the aim of the investigation wiLt he given. The chapter closes with a survey of the thesis.
1.1 General Introduetion
A polymer is a macromolecule, which is obtained by permanently combining a large number of small organic molecules, called monomers, by covalent bonds. Formation of a polymer, called polymerization, can be carried out in two ways: step-reaction and chain-reaction polymerization.
Step-reaction or condensation polymerization takes place between small, at least bifunctional, molecules, with the elimination of a small molecule such as water. Each polymer chain formed can react further with monomer, oligomerîc or polymerie chains, which means that each polymer chain grows over the whole course of the reaction and that the average molecular weight of the polymer
2
Chapter 1
increases during the whole reaction. Chain-Reaction or addition polymerization involves a growing chain with
one reactive site, which may be an ion or a free radical, and an unsaturated molecule as monomer.
Only these reactive ends can react.
A chain-reaction
polymerization consists of three stages: initiation, the formation of reactive sites and the addition of the first monoroer unit; propagation, the addition of more monoroer to the growing chain end; termination, the disappearance of reactive sites and the formation of "non-living" polymer chains: polymer ebains without a reactive endgroup.
However, new polymer ebains are continuously generaled
throughout the course of the polymerization and a polymer molecule is formed very rapidly on the timescale of the entire polymerization.
The production of polymer can be carried out in different ways. Focusing on radical chain polymerization, four methods can be distinguished.
These
different methods will be outlined briefly, and their main advantages and disadvantages will be included. Bulk Polymerization.
Only monoroer is charged into the reaction vessel
and, when necessary, an initiator is also charged. Normally the polymer formed is soluble in the monomer. The advantages of this metbod is the easy processing, the use of simple equipment, and the formation of a polymer with a minimum of contamination. The disadvantage is that it is difficult to control the reaction due to an increase in the viscosity. As a result, the removal of the heat of polymerization becomes a problem.
Locally the temperature can increase, which causes the
production of heterogeneities. Solution Polymerization.
A solvent is used to dissolve the monomer.
In
general this process is easier to control since the viscosity does not change much during the reaction.
The problem of the heat removal is therefore reduced to a
significant extent. The disadvantages include: the use of an expensive, toxic and often inflammable solvent, the separation of the polymer from the solvent at the end of the reaction, and the possible serious reduction of the molecular weight by chain transfer to solvent.
Introduetion
3
Suspension Polymerization. Water is used as the continuous phase and acts
as a heat-absorbing and -conducting medium. The monomer is suspended in smal! droplets. The initiator is soluble in the organic phase. Polymerization takes place inside the micron size droplets, which behave like little bulk polymerization vessels. In this system the problerns of heat removal and toxicity of the solvent are solved. However, attention bas to be paid to the stability of the dropiets in order to avoid coalescence.
The dropiets are kept in suspension by the generation of
sheer, resulting from agitation, in combination with the use of water-soluble stabilizers. At the end of the reaction, the polymer beads can easily be separated from the continuous (aqueous) phase. The polymer so obtained is in general less pure compared to the polymer formed in bulk or solution polymerization. Emulsion Polymerization.
Water acts here also as the continuous phase.
The monomer is dispersed in micron size dropiets in the continuous aqueous phase and stabilized by an oil-in-water emulsifier.
Usually a water-soluble initiator is
added to start the free-radical addition polymerization. In contrast to suspension polymerization, the loci of polymerization are the smal! micelles. This results in a reaction medium consisting of submicron polymer particles swollen with the monomer and dispersed in the aqueous phase. The final product is a stabie latex: a dispersion of submicron polymer particles in water.
In emulsion polymerization
these particles are much smaller compared to suspension polymerization and are typically 50-500 nm in diameter depending upon recipe and polymerization conditions 1• Advantages of this system include: it retains at solid contents of less than 40% its relatively low viscosity. This results in an efficient heat transfer and therefore good temperature controL Toxic and flammable organic solvents do not have to be used and the reaction can proceed to high conversion at a relatively high speed.
A major advantage is that it is possible to obtain polymer with a high
molecular weight, which can be controlled easily by a chain transfer agent.
A
disadvantage is that the polymer recovered from the latex bas a relatively high proportion of additives, such as surfactant and chain transfer agent. In the remainder of this thesis we will focus only on emulsion polymerization.
4
Chapter 1
1.2 Background of this Investigation
In order to cope with the shortage of natura! rubber prior to and during World War I, some first attempts to develop an artificial product as a substitute were undertaken in Germany, by means of emulsion-polymerization-like processes. Nevertheless, it was not until 1927 when the first reference to a process, which can be regarcled as a true emulsion polymerization, appeared; a patent was granted to Dinsmore 2 working for the Goodyear Tire & Rubber Company.
Luther and
3
Heuck were the first to introduce initiators to facilitate rapid polymerization. With these successful attempts, the development of a complete industry started. World War 11 gave a new impulse to this development. In the United States the Synthetic Rubber Program under the supervision of the Office of Rubber Reserve led to the successful production of synthetic rubbers for general purpose.
Since
then the number of papers that appeared on emulsion polymerization has grown exponentially, and the field is still growing in industrial importance. historica! survey has been given by Blackley
4
A detailed
•
The synthetic rubbers produced by emulsion polymerization are obtained in the form of a latex.
These latices are used in a wide variety of applications,
particularly as coating either by themselves or in formulation with pigments and other additives. Thus, they are used, for example, in paints and floor-polishes, and in coatings applied to, for example, paper, paperboard, and plastic films. However, when conventional surfactants are used in effecting emulsion polymerization, inherent difficulties are encountered.
In the latex state, the surfactant freely
desorbs from and adsorbs onto the surface of the polymer particles. This can lead to a change of stability over time. When the resultant latex is coated and dried on a substrate such as paper, the surfactant, which is only physically bonded, does not remain distributed uniformly between the coalescing polymer particles of the drying latex. Instead, the surfactant molecules exude to the surface or the interface of the coating. This leads to poor mechanica! properties, low water resistance, and also to blooming and blushing5 .
Introduetion
5
A solution to the problems caused by the non-bonded surfactant could be the fixation of the surfactant to the surface of the polymer particle, thîs without Iosing the stahilizing ability of the surfactant molecule. There are three ways to obtain in situ a chemica! bond between the polymer and the surfactant: 1) A monomer with surfactant properties (surfmer) could be added.
In
principle this would mean performing a copolymerization. The incorporation of the surfmer into the polymer depends on the reactivity of monomer and surfmer, and also on the way the polymerization is carried out.
More than one monomeric
surfactant can be built in into a polymer chain. In the extreme case, this would lead to a water-soluble polymer. 2) A surfactant with the properties of a chain-transfer agent (transsurf) could be used. The transsurf and the monomer should be consumed at such rates as to keep their concentration ratio constant.
If the transsurf is consumed more
quickly, a product with a different molecular weight and molecular weight distribution will be formed at the end of the polymerization due to the fact that chains are no Jonger stopped by the chain transfer agent
If the transsurf is
consumed slower there will be free surfactant left after the polymerization 1s completed. The formed polymer chains contain one surfactant unit at most. 3) The last possibility is the application of an initiator with surface-active properties (inisuif *). In this case the initiator dissociates into surface-active free radicals. Every initiation by such a primary radical leads to a surfactant-polymer link.
The surfactant is gradually linked by a chemica} bond to the polymer
throughout the course of the polymerization.
Depending on chain transfer to
monomer and the type of termination, the polymer chain can contain zero, one or two surfactant units.
Instead of surfactant fixation in the course of the polymerization, post
In this thesis the term "inisurf' is used for substances that contain one initiating moiety and at least one emulsifying moiety. Matcrials with more initiating groups in one molecule, whether or not they produce surface-active rad ie als, are referred to as "poly-initiators".
6
Chapter I
polymerization partiele surface modification could be applied to obtain a surface with bonded surfactant. This cumhersome method, which involves a comonomer with an additional, suitably reactive group, ineludes several cleaning and reaction steps and will nol be discussed any further.
From the aforesaid possibilities of solving the surfactant problem the application of an inisurf has our major interest. lt appears that this approach has the least disadvantages. Moreover, little is known about inisurf synthesis and their application in emulsion polymerization.
Besides the above mentioned practical application of an inisurf, the creation of radicals at the surface of a polymer partiele is also of scientific interest. Such a localized production of radicals might influence the formation of new particles, and might have an effect on the partiele size distribution, as well as on the rate of polymerization. Thus a study of inisurf in emulsion polymerization could lead to more insight into the formation of polymer particles.
Possible application areas of inisurfs, other than above mentioned, are the formation of particles with a specific size, the production of well-grafted polymer systems and the production of particles with a well-defined core-shell morphology.
1.3 Bistorical Overview In this historica! overview only the work carried out in the field of inisurfs or surface-aclive poly-initiators will be considered. Until now little work has been done on inisurfs or other compounds which conneet the surfactant to the polymer in the course of the polymerization. Pavljucenko et al. 6 reported in 1978 the modification of non-ionic surfactant by the introduetion of peroxy groups.
The introduetion of the peroxy
Introduetion
7
group on the terminal hydroxyl group had a large influence on the colloid chemica! properties. This led to the condusion that the formed micelles had a more ordered structure.
It was advised to use these initiating surfactauts in combination with
other surfactants, either ionic or non-ionic. In 1981 Ivancev et aC publisbed work on inisurf based on a peroxide initiator and a non-ionic surfactant of the poly(ethylene oxide) type. In this work a low initiation efficiency was mentioned. Moreover, a high molecular weight of the formed polymer was observed, as well as a narrow partiele size distribution (PSD), which was readily explained by the reduced rate of initiation and an extremely low chain terminalion rate.
lt was also found that the polymerization rate was
independent of the concentration of added co-surfactant. This was explained by the high rates of partiele growth and disappearance of micelles compaired with the rate of partiele formation. These authors 8 also found that the dissociation rate of the inisurf was higher than the dissociation rate of the corresponding initiator. lt was suggested that this was due to the localization of inisurf on the surface.
It is
important to mention that in all this work besides the surface-active initiators other surfactanis were used as well, so as to increase the stability of the system. Tauer et al. 9 reported in 1988 on work done on surface-active initiators, where the initiator moiety is of the azo-type and the surfactani is an a,ro-diol with either a poly(ethylene oxide) or a poly(propylene oxide) centre part. Depending on the diollinitiator ratio, it was possible to produce an inisurf or a poly-initiator with up to 104 initiator groups. A stabie latex could be produced with only monomer, water and surface-active initiator, and with a solid content of up to 40 %.
The
latex produced in this way had a lower level of electrolyte content and a lower foam formation ability.
The final product showed a lower water absorption, an
increase in mechanica! properties, and an increase in heat and light resistance. An added advantage was a decrease of environmental polJution by lowering the surfactant content of the waste water. In 1990, in a second paper, Tauer et al. 10 reported on the increase of the molecular weight of the formed polymer compared to a conventional system. This result wil! be discussed in Chapter 6.
8
Chapter 1
Salamone et al. 11 reported on surface-active initiators with an azo-type initiator moiety and either a poly(ethylene oxide) or a poly(ethylene oxide) methyl ether surfactant moiety. In addition to the synthesis of the inisurf and the surfaceactive
poly-initiators,
the
molecular weight
measurements
of the
formed
components, the determination of the critica! micelle concentration and the micellar molecular weight were reported as well. A patent was granted to Tauer et alY in 1990. lt describes the synthesis of a surface-active initiator with an azo-type initiator moiety and an ionic surfactant moiety, as well as the production of a highly monodisperse latex by emulsion polymerization, where the reaction mixture contains only monomer, water and surface-active initiator.
It was suggested that the major application of the so
formed latex will be as a dispersion binder, or in medica! and biomedical systems. In a series of newsletters of the International Polymer Colloid Group, Blackley 13 reported on the inisurfs of symmetrical azo-initiators and nonionic surfactants of the alkyl poly(ethylene oxide) type.
The synthesis and the
dissociation constants of the formed inisurfs were given.
1.4 Aim of this Investigation
The aim of the study described in this thesis is to synthesize inisurfs, and to investigate their behaviour in emulsion polymerization. This aim can be translated into four sub-tasks: I) synthesis, 2) check of the possiblity of application in emulsion polymerization, 3) physical properties of the inisurf and 4) determination of kinetic parameters in emulsion polymerization.
Synthesis.
A route has to be developed for the preparation of a material
with both surfactant and initiator properties.
An azo-type initiator with carboxy
groups and a nonionic surfactant with one hydroxy group will be tested for use as starting materials. An esterification will be carried out to combine the initiator and
Introduetion
surfactani parts.
9
The molecular structure of the resultant inisurf will be verified
with the aid of proton Nuclear Magnetic B.esonance
eH NMR).
Check of the Possibility of Application in Emulsion Polymerization. After the synthesis, the inisurf will be tested in emulsion polymerization.
The inisurf
must be able to stabilize the monomer suspension and preferably form micelles. It should also form free radicals that can initiale the polymerization. At the end of the polymerization the inisurf must be able to stabilize the latex. In these emulsion polymerizations styrene is chosen because of the substantial knowledge available about this monomer. During the polymerization, the possiblr occurrence of phase separation, as well as the polymerization rate will indicate suitability of the inisurf.
Physical Properties of the lnisurf In order to gain further insight, it is also necessary to delermine certain physical parameters.
Since an inisurf is surface-
aclive it will form micelles and the Çritical Micelle Çoncentration (CMC) wil! be determined by the DuNouy-ring-method 14 or with the Maximum Bubble Pressure Method 15 •
Surface coverage of an inisurf has to be taken into account as wel!,
and thus the adsorption isotherm onto polystyrene particles will be determined. Since the actdition of a surfactani to this initiator might influence its dissociation behaviour, the rate of dissociation wil! be measured by means of UV-measurements as a function of time and temperature.
Kinetic Parameters. Besides the batch reactions, from which reaction rate and partiele diameter are determined, seeded reaelions will be carried out to delermine the parameters that govern the growth of the polymer particles. These seeded reaelions will be carried out in dilatometers, where the drop in volume wil! be monitored as a function of time 16 , to obtain the information on conversion necessary densimetry
for 17
calculating
the
growth
parameters.
will be used in the seeded reactions.
Besides
dilatometry,
In a closed loop the reaction
mixture will be pumped through a density cel! and back into the reactor. change in density will be monitored as a function of time.
The
10
Chapter 1
1.5 Survey of this Thesis
A short outline is now given of the remaining chapters in this thesis.
Chapter 2:
The various possible choices for the initiator part and the
surfactant part will be discussed. The synthesis of initiators and inisurfs will be outlined in detail and the measured 1H NMR spectra will be discussed.
Chapter 3:
Qualitative
polymerization will be discussed.
and
quantitative
models
for
emulsion
A kinetic description will be given of the so
called "zero-one" system in Intervals II and III. In addition, a metbod determining kinetic parameters, such as the entry (p) and exit (k) rate coefficients, will be descri bed.
Chapter 4:
The various experimental and analytica! techniques used in
carrying out the emulsion polymerizations and characterizing the formed latex will be described. Attention will also be paid to the methods used in determining some important characteristics of the synthesized inisurfs.
Chapter 5:
The first results of applying the various inisurfs in the
emulsion polymerization of styrene will be discussed. Only the results of ab initia reaction will be given in this chapter. Applicability of the inisurf will be deduced from the stability of the emulsion polymerization system.
The effect of co-
surfactant, applied in some systems, will be discussed as wel!.
Chapter 6:
The results of detailed studies of two inisurfs: a symmetrical
and an asymmetrical one, will be discussed. wil! be given as a function of time.
Dissociation rates of these inisurfs
The CMC and the adsorption isotherm of
these inisurfs at room temperature, will be discussed. Reaction rates, partiele size distributions, and molecular weight distributions wil! be evaluated.
Introduetion
Chapter 7:
ll
The results of seeded emulsion po1ymerization reactions
employing the two inisurfs as described in Chapter 6, will be given. The growth parameters of a polymer particle, i.e., entry (p) and exit (k), will be discussed. Efficiency of initiation of the primary surface-active free radicals will be given as a function of coverage of the seed particles with inisurf and wil! be compared to conventional systems.
Epilogue:
The conclusions drawn from the results mentioned in the
previous chapters, will be discussed and some suggestions for further research will be given.
Parts of this work have been presented at the IUPAC International Symposium on Macromolecules (Montréal, Canada, July 1990) and the Gordon Research Conference on Polymer Colloids (lrsee, FRG, September 1992).
Parts of this thesis have been publisbed or will be published: the synthesis of symmetrical inisurfs (part of Chapter 2) and the ab initia reactions with these inisurfs as described in Chapter 5 18 ; the seeded study with symmetrical inisurf of Chapter 7 19 ; and the synthesis of asymmetrical inisurf (part of Chapter 2), the ab initia reactions with these asymmetrical inisurfs as described in Chapter 5 and the
seeded study with asymmetrical inisurfs of Chapter 720 •
References: I. Bovey, F.A.; Kolthoff, I.M.; Medalia, AL; Meehan, E.J. Emu/sion Palymerizatian; Interscience Publishers: New York, 1955. 2. Dinsmore, R.P. U.S. Pat. 1,732,795, 1929; Chem. Abstr. 1930, 24, 266. 3. Luther, M.; Heuck, C. U.S. Pat. 1,860,681, 1932; Chem. Abstr. 1932, 26, 3804. 4. Blackley, D.C. Emulsian Palymerisation; Applied Science Publishers Ltd.: London, 1975.
12
Chapter 1
5. Dickstein, J. Polym. Prep.-Am. Chem. Soc., Div. Polym. Chem. 1986, 27, 427. 6. Pavljucenko, V.N.; Ivancev, S.S.; Ro~kova, D.A.; Dikaja, N.N.; Domniceva, N.A.; Budtov, V.P. Kolloid Z. 1978, 40, 64; Eng. Transl.: Colloid J. USSR 1978, 40, 48. 7. Ivancev, S.; Pavljucenko, V.N. Acta Polym. 1981, 32, 407. 8. Ivancev, S.; Pavljucenko, V.N.; Byrdina, N. J. Polym. Sci., Polym. Chem. Ed., Part A-1 1987, 25, 47. 9. Tauer, K.; Goebel, K.-H.; Kosmella, S.; Neelsen, J.; Stähler, K. Plaste Kautsch. 1988, 35, 373. 10. Tauer, K.; Goebel, K.-H.; Kosmella, S.; Stähler, K.; Neelsen, J. Makromol. Chem., MacromoL Symp. 1990, 31, 107. 11. Salamone, J; Liao, W; Watterson, A. Polym. Prep.-Am. Chem. Soc., Div. Polym. Chem. 1988, 29, 275. 12. Tauer, K.; Goebel, K.-H.; Neelsen, J.; Kosmella, S.; Stähler, K.; Schirge, H.; Kaltwasser, H. G.D.R. Pat. 276 877 Al, 1990; Chem. Abstr. 1990, 113, 232262s. 13. (a) Blackley, D. In International Polymer Colloids Group's Newsletter; Napper, D., Ed.; 1989, 20 (1), 4. (b) Blackley, D. In International Polymer Colloids Group's Newsletter; Napper, D., Ed.; 1989, 20 (2), 4. (c) Blackley, D. In International Polymer Colloids Group's Newsletter; Napper, D., Ed.; 1990, 21 (1), 8. 14. DuNouy 1. Gen. Physiol. 1918, 1, 521. 15. Sirnon Ann. Chim. Phys. 1851, 32, 5. 16. Fryling, C.F. Ind. Eng. Chem., Anal. Ed. 1944, 16, 1. 17. Poehlein, G.W.; Dougherty, D.J. Rubber Chem. Techno!. 1977, 50, 601. 18. Kusters, J.M.H.; Leijten, A.M.M.; Van Es, J.J.G.S.; German, AL. in preparation. 19. Kusters, J.M.H.; Napper, D.H.; Gilbert, R.G.; German, A.L. Macromolecules 1992, 25, 7043. 20. Kusters, J.M.H.; Verweerden, T.M.M.; Van den Enden, M.J.W.A.; German, AL.; Gilbert, R.G. in preparation.
Chapter 2
Synthesis
Summary: In this chapter the choice of initiator and surfactant, used in the synthesis of inisurf, wil/ be examined. An outfine of the method used for the synthesis of inisurf will be given. The various inisurfs synthesized wilt be mentioned and their 1H NMR spectra will be discussed. This is foliowed by a description of the experimental procedures of the syntheses.
2.1 Introduetion
An inisurf consists of two parts: an emulsifying moiety and an initiating moiety. The link between the surfactant and the initiator in the inisurfs used in this investigation, was envisioned to result from an esterification. This type of reaction can result in a very high yield, while being performed under comparatively mild conditions. In particular a low reaction temperature is necessary in order to obtain the inisurf, since the initiators used, being similar to the ones normally used in emulsion polymerization, would readily dissociate during the synthesis, if carried out at elevated temperatures. Since not every initiator is compatible with the conditions of esterification, the different types of initiators, as well as the selection procedures used for these
14
Chapter 2
initiators, will be discussed. This will be foliowed by an overview of the available types of surfactants, including the ones selected. Not all the initiators, which were of interest for this investigation, are commercially available.
Therefore, the
discussion will be extended to the synthesis of a non-commercially available initiator. The choice of the metbod of esterification of initiator with surfactant will be elucidated. Results of the syntheses of the initiator and of the various inisurfs will then be discussed, and the chapter will close with the experimental details.
2.2 Choice of Initiator and Surfactant
Different types of initiators and surfactants can be used in emulsion polymerization.
These types of initiators and surfactants will be reviewed, after
which the choice made will be corroborated.
2.2.1 Initiator
An initiator is a chemica! substance or a radiation source, which is capable of producing free radicals.
There are several types of free radical initiators, and
some of these will be discussed now.
Peroxides.
These initiators have the general structure: ROOR' or ROOH,
and dissociate thermally into two free radicals by cleavage of the oxygen-oxygen bond. The dissociation rate of the peroxides is strongly influenced by the groups neighbouring the peroxide-function. This influence can be divided into three major effects: (1) the relative stability of the radicals formed, (2) steric effects, and (3) electronic effects. In addition to these structural effects, some peroxides are also susceptible to radical-induced dissociation, which leads to a loss of efficiency, smce the peroxides then dissociate without adding more radicals to the system. This
radical-induced
dissociation
does
not
occur
in
vinyl-monomer
Synthesis
15
polymerization 1• Be si des the structural effects and the radical promoting effects, the dissociation rate may also be influenced by pH, solvent, and the presence of transition metals and contaminants. Redox Systems.
These systems consist of an oxidizing and a reducing
agent. A wide variety of compounds can be used in these systems. In the case where peroxides are used as one of the two components of the redox system, the dissociation temperature is lower, and the radical efficiency is decreased, when compared with the peroxide system. Azo Compounds.
These initiators have the general structure: RN=NR'.
This type of initiator dissociates thermally and produces two free radicals and a nitrogen molecule.
As with peroxides, the side groups that are attached to the
azo-function have an effect on the dissociation rate. In contrast to the case of most peroxides, the dissociation rate of azo compounds shows only a minor solvent effect and is hardly affected by transition metals, acids, bases, or contaminants. Radiation Sources.
High energy radiation, e.g., X-ray, y-ray, or
~-ray,
is
able to fragment molecules into ions, as well a into free radicals. Alternatively, UV -radiation, in combination with a photo-initiator, is capable of producing free radicals. In some cases a photosensitizer may be used.
In choosing an initiator for this work eertaio requirements have to be met. Firstly, it is desirabie to have only one component in the initiating system. As was mentioned in Chapter 1, one of the disadvantages of emulsion polymerization is that the polymer obtained has a relatively high content of non-polymerie materiaL Use of an initiator system, which consists of two or more components (e.g., redox initiators), would increae the number and amount of non-polymerie "contaminants" and could subsequently lead to a lower performance of the final polymer. Secondly, the components should be handled with a minimum of safety hazards. Thirdly, the initiator needs a reactive group, besides the initiating group, that can be used to establish the link between initiator and surfactant.
Formation of this
linkage should not have any adverse influence on the initiating part. Finally, the initiator should not dissociale during the synthesis, but the initiating moiety of the
16
Chapter 2
inisurf should still dissociate at an appreciable rate at elevated temperatures.
Only peroxides and azo-initiators satisfy the first requirement. The second requirement eliminales peroxides as a possibility. In genera!, peroxides have to be handled with extreme care. Thus, an azo-initiator appears to be the only type of initiator, which satisfies the first two requirements. The third requirement is met in a commercially available azo-initiator: 4,4'-b,zobis(4-Çyanofentanoic Acid) (la) (ACPA, see Fig. 2.1)
0
CN CN 0 ' c-CH -CH -cN=N-c-CH · CH -c / 2 2 ' ' 2 2 ' OH HO CH3 CH3
Figure 2.1
,,
"'
Structure of 4,4'-Azobis(4-cyanopentanoic Acid) (la).
The two carboxy groups of this initiator can he utilized in an esterification. Extension of the carbon chain by addition of surfactant has probably only a smal! effect on the dissociation rate2·3·'
The use of this initiator will lead to inisurfs,
which are symmetrical in the azo-function, and these will subsequently he designated as "symmetrical inisurfs". In the course of this investigation, an idea was developed to investigate asymmetrical inisurfs, that will give a surface-active radical and a smal!, water-soluble radical upon thermolysis.
With regard to this, 4-t-!!utylb,zo-
4
4-Çyanopentanoic Acid (lb) (BACA, see Fig. 2.2) was envisaged to meet the above requirement. This compound is not commercially available. The synthesis of this initiator wil! be discussed in Section 2.3.1. The inisurf formed with this initiator, wil! be designated as "asymmetrical inisurf". The last requirement, suitability, wil! be investigated by carrying out ab
Determination of thc dissociation ratc of the various inisurfs synthcsized will be described in the relevant section of Chapter 6.
Synthesis
17
initio emulsion polymerization and will be discussed in Chapter 5.
0
CN
CH
~ ' ' 3 C-CH -CH2-C-N=N-C-CH / 2 ' ' 3 HO CH3 CH3
Figure 2.2
Structure of 4-t-Butylazo-4-cyanopentanoic Acid (lb ).
2.2.2 Surfactant
Surfactauts are substances with a hydrophilic and a hydrophobic part. The hydrophobic part usually is a linear or branched hydrocarbon-chain, which can be
an unsaturated, or, less frequently, a halogenated or oxygenated hydrocarbon, or a siloxane chain. The hydrophilic part is an ionic or other highly polar group. In general, surfactauts are classified by their hydrophilic part: Anionic Suifactants.
This type of surfactant bears a negatively charged
group, e.g., RCOU Na+ and RS0 3- Na+. Cationic Suifactants.
group, e.g., RNH/
er and
This type of surfactant bears a positively charged
RN(CH3) 3+ CL
Zwitterionic Suifactants.
This type of surfactant has both positively and
negatively charged groups in the surface-active part, e.g., RNH/CH 2COo- and RN(CH 3)z +CHzCH2S03-. Nonionic Suifactants. This type of surfactant bears no ionic charge, e.g.,
RCOOCH 2CHOHCH20H and R(OC 2H4)pH. Mainly anionic and nonionic types of surfactants are used in emulsion polymerization.
By applying an anionic surfactant in the synthesis, an anionic
inisurf will be formed. It was envisioned that the use of such an ionic inisurf in emulsion polymerization would create a situation comparable with the situation in emulsifier-free emulsion polymerization.
In emulsifier-free emulsion polymeriza-
18
Chapter 2
tion 5•6•7 , mainly sodium peroxydisulphate is used as initiator.
The amomc free
radicals formed upon dissociation, will oligomerize in the aqueous phase and will become surface active. These in situ formed, surface-active free radicals are very similar to the free radicals which would be formed by dissociation of anionic inisurfs. Thus, it may be reasoned that the effects of anionic inisurfs on emulsion polymerization are probably similar
to
the results obtained by emulsifier-free
emulsion polymerizations. Also from a synthetic point of view, nonionic surfactants are to be preferred in the synthesis of inisurf, because the use of an anionic surfactant would require the selective protectionldeprotection of the anionic group as additional reaction steps.
Due to the presence of the labile azo-group, the synthesis of the inisurf
should be as simple and as mild as possible. Additionally, nonionic surfactants have some advantages over anionic surfactants.
The degree of surface activity of nonionic surfactants can easily be
adjusted by changing the fudrophile-hipophile-,!lalance 8 (HLB) of the surfactant. Nonionic surfactants are less sensitive to the acidity (pH) or ionic strength of the medium, since they primarily work by means of the steric stabilization mechanism9· 10 •
Moreover, nonionic surfactants with a functional group, which
can be applied in the synthesis of inisurf, are more readily available than suitably functionalized anionic surfactants. The nonionic surfactants used, are of the poly(ethylene oxide) type. Two different types of polyether chains have been used. The first type was not a surfactant in itself. Water-soluble monomethyl poly(ethylene oxide)s (2) have been used (see Fig. 2.3).
CH ~o~CH ~eH -O~H 3 L 2 2 Ln Figure 2.3
Structure of Monomethyl Poly(ethylene Oxide)s (2), with 2a: nave :::: 7; 2b: nuve = 12.
We reasoned that the resulting inisurf would be soluble in the aqueous phase like
Synthesis
19
the initiators normally used in emulsion polymerisation.
After dissociation the
poly(ethylene oxide)-based radical would initiate polymerization in the aqueous phase.
After adding some monomer units the oligomer would become surface
active, upon which it would adsorb onto the surface of a polymer particle. The surface-active radical so formed would contribute to the colloidal stability of the polymer particles and would lead to a continuation of the polymerization in the particles, as in the case of peroxydisulphate initiatod. The seeond type of surfactants are "real" surfactants and of the nonylphenyl poly(ethylene oxide) type (2) (see Fig. 2.4).
H
Figure 2.4
Structure of Nonylphenyl Poly(ethylene Oxide)s (2), with 2c: nav. = 9; 2d: nav• = 30.
We reasoned that with this surfactant, the formed inisurf would already be surface-active. This implies that the radicals will already be formed at the surface of the polymer particles.
2.3 Results and Discussion
Before going into the details of the syntheses of the inisurfs via esterification of the initiator and surfactant, the synthesis of a specific type of azo-initiator will be discussed.
20
Chapter 2
2.3.1 Asymmetrical Azo-Initiator
The asymmetrical azo-initiator BACA (lb) has been prepared in two steps from commercially available levulinic acid (3) and t-butylhydrazine (4) (see Eq. 2.1). The first step, carried out in an aqueous acidified environmene·4, involved condensation of the ketone group in (3) with the hydrazine (4) to give the hydrazone (5), which reacted in situ with hydrogen cyanide to give (6).
0~
~H3
/,/0
+
/c-GI2 -CH:2-C, HO
2
H N-NH-~-GI
CH3 (4)
(3)
3
Gl
3
~0
' 0~
/c
CH:l 1'N··NH-C-GI3
rn2
HO
c~-c, (S)
0~
013
(2.1)
Gl3
~
~~
/ C-CH:2-CH2 -~·-NH- NH-~-GI3 HO
GI3
CH3
(6)
Without purification, (6) is oxidized to (lb) (see Eq. 2.2).
0~
~
/C HO
~3
Oxidation
0~
CH 2 ·CH 2 -~-NH-NH-~-CH3
rn3
GI 3
/ C- CHc HO
(6)
Cl-!:2·
~ N= N
c-- GI3
3
CH 3
CH (lb)
(2.2)
Upon precipitation, BACA (lb) was obtained as a white powder in 70% yield and had a melting point of 353.5 K (lit4 : 353 K).
The 1H and
13
C NMR
spectra (see Section 2.5) are in good agreement with the proposed molecular
Synthesis
21
structure. Experimental details are described in Section 2.5. BACA was used in the esterification without further purification.
2.3.2 Inisurf The choice of initiator and surfactani moieties was governed by the fact that the conceived way of obtaining the link
an esterification -
could be carried out
under comparatively mild conditions, i.e., at, or preferably below, room temperature. This requirement is necessary since the azo-initiator, used as one of the reaction components, dissociates at elevated temperatures.
Within the
boundaries of this requirement there are a number of ways of performing an esterification, the most important being: ( 1) actdition of an excess of one of the reactants, usually the alcohol, and (2) transformation of the carboxylic acid into an acid halide or active ester compound, which readily reacts with a stoichiometrie amount of alcohol.
Method 1: the addition of an excessof the alcohol, which in our case is the surfactant, is not really feasible.
Due to the fact that physical properties, e.g.,
solubility in an organic or aqueous phase, surface activity (foaming), of the formed inisurf and the used surfactani are comparable, it is almost impossible to carry out a separation on a preparalive scale. In the case of expensive starting materials, the addition of an excess of one of the materials is not preferabie either. Method 2: the transformation of a carboxylic acid into its acid halide was tried with la (ACPA) (using phosphorus pentachloride in benzene at 273 K), but only a 40% yield of the dihalide was obtained. The reaction of this bisacylchloride with an alcohol had, in genera!, a yield of about 70%. This leads to an overall yield of only 30%. A better way was found in the use of Q.i_çyclohexylç_arbodiimide (DCC) as the condensing agent. This is a one pot procedure that can generally be carried out at, or significantly below, room temperature. By means of some modifications of
22
Chapter 2
the original procedure, yields of up to 95 % can be achieved, using a wide variety of alcohols and acids as the reaction partners in stoichiometrie amounts (see Eq. 2.3). These modifications and improvements have been described by a number of groups, e.g., Holroberg et a/. 11 , Hassner et al. 12 , and Shinkai et alY.
They
reported that the use of DCC by itself does not give satisfactory results. The yield can be increased by the actdition of a catalytic amount of a strong acid (Jzara!oluene.§.ulfonic in pyridine
11
•
~cid,
pTSA) to the solution of the carboxylic acid and the alcohol
Pyridine also acts as a catalyst; it can be replaced by an .Q.imethyl~mino.t:!Yridine,
aminopyridine (generally
amounts) in toluene as solvent
12
,
DMAP) (added in catalytic
which is preferabie for health reasons. As a side
product gi.f}'clohexyl!J.rea (DCU) is formed.
Further experimental details are
described in Section 2.5. DCC pTSA ij
DMAP
0
+
R 1c,
R20H
DCU
OH (1)
«o R 1c, o-R2
(2)
(2.3)
(7)
For the various inisurfs synthesized, different abbreviations are used. The abbreviation begins with an S or an A indicating whether it is a .§.Ymmetrical or an ~symmetrical
inisurf. The remainder of the abbreviation: two letters and a number,
indicates the type of surfactant and the average length of the ethylene oxide chain in the surfactant, respectively: EO-# for monomethyl poly(l;;.thylene QXide) and CO-# for nonylphenyl poly(ethylene oxide) (Antarox® CO#).
The following inisurfs have been synthesized: SE0-350 (7a) and SE0-550 (7b) (see Fig. 2.5), SC0-630 (7c) and SC0-880 (7d) (see Fig. 2.6), AE0-550 (7e) (see Fig. 2.7), AC0-630 (7f) and AC0-880 (7g) (see Fig. 2.8).
Synthesis
23
Figure 2.5
Ctjll
Structure of Symmetrical Inisurfs SEO-# (7) with a Monomethyl Poly(ethylene Oxide) Chain as Surfactant Moiety, with 7a: nave ::: 7; 7b: nav. = 12.
l
' 0 ,, C-UI ·····CH -C-N CN
~))-o-{rn,-rn,-of,
2
2
rn,
2
Figure 2.6
Structure of Symmetrical Inisurfs SCO-# (7) with a Nonylphenyl Poly(ethylene Oxide) Chain as Surfactant Moiety, with 7c: nave = 9; 7d: nave = 30.
Figure 2.7
Structure of Asymmetrical lnisurf AE0-550 (7e) with a Monomethyl Poly(ethylene Oxide) Chain as Surfactant Moiety (nave ::: 12).
Figure 2.8
Structure of Asymmetrical Inisurfs ACO-# (7) with a Nonylphenyl Poly(ethylene Oxide) Chain as Surfactant Moiety, with 7f: n,we = 9; 7g: na•• = 30.
The esterifications were monitored by 1H NMR. For the surfactant starting materials, all ethylene oxide units appear in the same region, at
o 3.50-3.90 ppm
24
Chapter 2
for monomethyl poly(ethylene oxide) and at
a
3.55-3.75, 3.85, 4.11 for
nonylphenyl poly(ethylene oxide). Upon esterification, the methylene group(s) of the first ethylene oxide unit(s) directly attached to the ester group(s), gave rise to a distinct resonance at
a 4.2-4.3 ppm.
The integral of these protons was compared
with the integral of the CH1-CH 2-group of the initiator moiety, which was not significantly shifted upon esterification.
The
inisurfs
(7)
were
isolated
by
flitration
of
the
formed
1,3-.!!i_çyclohexylQrea (DCU) (not soluble in toluene), the dimethylaminopyridine (DMAP), and the p-toluenesulfonic acid (pTSA) foliowed by evaporation of the solvent. The yields were determined from both the isolated amounts of inisurf (7) and 1,3-dicyclohexylurea (DCU), and are reported in Tabel 2.1.
The resultant
inisurfs (7) were kept at 277 K and were used without further purification.
Table 2.1.
lsolated Yields of the Formed lnisurfs (7).
Inisurf
Isolated Yield (%) Inisurf (7)
1,3-Dicyclohexylurea (DCU)
SE0-350
99
97
SE0-550
96
95
SC0-630
99
99
SC0-880
98
98
AE0-550
99
98
AC0-630
98
98
AC0-880
100
99
lnisurfs were mainly characterized by their 1H NMR spectrum as described above. The
13
C NMR spectra were rather complex due to the fact that the used
surfactants were commercial products, containing ethylene oxide units, with a varying chain length and, in the case of Antarox-based inisurfs, isomerie nonyl phenoxy units. Nevertheless, these spectra revealed the presents of the ester group
Synthesis
25
and of the nitrile group, and, in case of the asymmetrical inisurf, of the t-butyl group. Biemental analyses of the inisurfs were not perforrned since the used surfactants are mixtures containing poly(ethylene oxide) chains of different lengths.
2.4 Conclusions From the synthetic work presented in this chapter it can he concluded that the method for synthesizing an asymmetrical azo-initiator is suitable. It can also he concluded that the chosen esterification method is suitable for the syntheses of inisurfs in a very high yield.
2.5 Experimental Section Reagents 4,4' -Azobis( 4-.Çyano~entanoic Acid) (la) (ACPA, purum, Fluka
AG, Buchs, Switzerland) (see Fig. 2.1) was used as received.
4-t-Butylazo-4-
cyanopentanoic acid (lb) (BACA) (see Fig. 2.2) was obtained from commercially available levulinic acid (3) and t-butylhydrazine hydrochloride (4) as described below and was used without forther purification. Levulinic acid (3) (80%, Janssen Chimica, Geel, Belgium) was distilled once before use.
t-Butylhydrazine
hydrochloride (4) (>95%, Janssen Chimica, Geel, Belgium), sodium cyanide (pro analysi, Merk, Darmstadt, FRG), and chlorine (Hoekloos, Schiedam, The
Netherlands) were used as received. Monomethyl poly(ethylene oxide)s (2a) and (2b) (Aidrich Chemie, Bornem, Belgium), as well as nonylphenyl poly(ethylene oxide)s (2c) and (2d) (Antarox® Co, GAF, New York, USA) were dried by means of azeotropic
distillation
with
toluene.
Dicyclohexy lcarbodiimide
(DCC,
C6H 11 N=C=NC6H 11 , 99%, Janssen Chimica, Geel, Belgium) was recrystallized from
26
Chapter 2
diethyl ether. Belgium)
p-Toluenesulfonic acid (pTSA, 99%, lanssen Chimica, Geel,
was
dried
at
313
K
in
vacuum
shortly
before
use.
Dimethylaminopyridine (DMAP, 97%, lanssen Chimica, Geel, Belgium) was used as received.
Solvents Dichloromethane (pro analysi, Merk, Darmstadt, FRG) was used as received.
Toluene (pro analysi, Merk, Darmstadt, FRG) was dried over
molecular sieves (pore size: 3 Á).
lnstrumentation
The 1H and
13
C NMR spectra were recorded at room
temperature with a Bruker AM 400 spectrometer, using trichloromethane-d (CDC1 3) as the solvent and tetramethylsilane (TMS) as internal standard, unless stated otherwise.
Infrared spectra were obtained with a Mattson Polaris™ FT-IR
Spectrometer. Melting points were determined on a Büchi melting point apparatus and are uncorrected.
Synthesis of the Asymmetrical Azo-Initiator4 • To a salution of 31.2 g (0.25 mol) of t-butylhydrazine hydrachloride (4) and 12.3 g (0.25 mol) of sodium cyanide in 80 mL of water in a jacketed 0.5 dm 3 reactor equipped with an efficient mechanica! stirrer, thermometer, condenser and dropping funnel, was added dropwise 29.0 g (0.25 mol) of freshly distilled levulinic acid (3) over a period of 5 min. The reaction mixture heated up to 313-318 K during this addition period and was then heated for 2 h at 323 K.
The reaction mixture became more viscous
during this period. Subsequently, the reaction mixture was cooled to 273 K, and 50 mL of dichloromethane and 50 mL of water were added.
Chlorine was then
passed into the reaction mixture through a gas inlet tube reaching to the bottorn of the reactor, until approximately 28.4 g (0.40 mol) of chlorine had been absorbed. During this period the reaction temperature was maintained below 298 K. At the end of the chlorine addition, the dichloromethane layer was separated.
The
unreacted hydragen cyanide in the aqueous layer was treated with a 20 wt% salution of NaOH in water, foliowed by sodium hypochlorite, until bleaching of pH
Synthesis
27
paper persisted and was then discarded. The dichloromethane layer was stirred into 200 g of a 5 wt% solution of NaOH in water, and after 10 min the dichloromethane layer was separated and discarded.
The aqueous layer was
3
transferred into a 0.5 dm beaker and acidified to a pH of 1-2 with concentrated hydrochloric acid.
The precipitated 4-t-Butylazo-4-cyanopentanoic acid (lb) was
filtered off, washed with 200 mL of water and air dried. 4-t-Butylazo-4-cyanopentanoic acid (lb):
white powder; mp 353.5 K (lit4 :
353 K); IR (KBR) 3300-2500, 2241, 1742, 1722, 1229 cm· 1; 1H NMR (CDC1 3) ö 1.10-1.40 (br, 9 H, C-(CH 3) 3), 1.60 (s, 3 H, CH 3), 2.10-2.60 (m, 4 H, CH 2CH 2); 13
C NMR (DMSO-d6) ö 23.3 (C-5), 26.3 ((ÇH3) 3C), 28.8 (C-3), 32.7 (C-2), 68.1
(q-C, C-4), 71.0 (q-C, (CH 3)&}, 119.1 (q-C, CN), 172.9 (q-C, COOH). All esterifications have been carried out under the
Synthesis of lnisurf
same conditions. The recipe used in these syntheses can be found in Table 2.2. For each reaction, surfactant (2), dicyclohexylcarbodiimide, p-toluenesulfonic acid and dimethylarninopyridine, as well as 60 mL of toluene were charged into a dry 250 mL round-bottom flask, equiped with a stirring bar and a drying tube. The initiator (1) was crushed in a mortar and added in smal! portions to the solution over a period of 1 h. Another 60 mL of toluene were added to the solution. The flask was wrapped in aluminium foil and the contents was stirred at 277 K. After two days, the 1,3-dicyclohexylurea (DCU) (insoluble in toluene) was removed together with p-to1uenesulfonic acid and dimethylaminopyridine, by filtration using a fritted glass disc filter.
The toluene in the filtrate was evaporated on a rotary
evaporator at 303 K, to give the inisurfs (7), in yields as specified in Table 2.1. SE0-350 (7a): pale yellow Iiquid; 1H NMR (CDCI 3)
ö 1.68-1.80 (m, 6 H,
CH 3), 2.20-2.60 (m, 8 H, CH 2CH2), 3.4 (s, 6 H, CH,O), 3.50-3.90 (m, CH 2CHzO), 4.10-4.30 (m, 4 H, CH20CO). SE0-550 (7b): pale yellow wax; 1H NMR (CDCI 3)
ö 1.68-1.80 (m, 6 H,
CH 3), 2.20-2.60 (m, 8 H, CH 2CH 2), 3.4 (s, 6 H, CH,O), 3.50-3.90 (m, CH 2CH 20), 4.10-4.30 (m, 4 H, CH 20CO).
28
Chapter 2
Table 2.2.
Recipe for the Esterification of Initiator and Surfactant.
Amount (mmol)
Components Surfactant (2)
21.8", 10.9b
Dicyclohexylcarbodiimide
21.8", 10.9b
Initiator (1)
10.9
p-Toluenesulfonic acid
1.1
Dimethy1arninopyridine
1.1
120 mL
Toluene
a: Recipe for esterification with symmetrical initiator (la) b: Recipe for esterification with asymmetrical initiator (lb)
SC0-630 (7c): pale yellow wax; 1H NMR (CDC1 3)
o 0.50-0.90
(br 6 H,
CH 2C!! 3), 1.20-1.40 (m, 24 H) and 1.50-1.90 (m, 14 H) (C 9H19 and CH3), 2.20-2.70 (m, 8 H, CH 2CH 2), 3.55-3.75 (m, CH 2CH20), 3.85 (t, 4 H, J
15 Hz,
ArOCH2C!! 2), 4.11 (t, 4 H, J = 12 Hz, Ar0CH 2CH 2), 4.20-4.30 (m, 4 H, CH10CO), 6.90-7.10 (m, 8 H, Ar). SC0-880 (7d): pale yellow solid; 1H NMR (CDC13)
o 0.50-0.90
(m, 6 H,
CH2C!!J), 1.20-1.40 (m, 24 H) and 1.50-1.90 (m, 14 H) (C9H19 and CH 3), 2.20-2.70 (m, 8 H, CH 2CH 2), 3.55-3.75 (m, CH2CH 20), 3.85 (t, 4 H, J = 15 Hz, ArOCH2CH 2), 4.11 (t, 4 H, J
=
12 Hz, Ar0C!! 2-CH2), 4.20-4.30 (m, 4 H,
CH 20CO), 6.90-7.10 (m, 8 H, Ar). AE0-550 (7e): pa1e yellow wax; 1H NMR (CDC1 3)
o 1.05-1.40 (br,
9 H,
C(CH3h), 1.68-1.80 (br, 3 H, CH 3), 2.20-2.60 (m, 4 H, CH 2CH 2), 3.4 (s, 3 H, CHp), 3.50-3.90 (m, CH2CH20), 4.10-4.30 (m, 2 H, CHzOCO). AC0-630 (7f): pa1e yellow liquid; 1H NMR: (CDC1 3)
o0.60-0.95 (br, 3 H,
CH2 CH 3), 1.05-1.45 (m, 21 H) and 1.55-1.80 (m, 7 H) (C(CH 3h, C9 H 19 , and CH3), 2.20-2.65 (m, 4 H, CH2CH 2), 3.55-3.75 (m, CH2CH20), 3.85 (t, 2 H, J ArOCH 2C.tl 2), 4.11 (t, 2 H, J
= 12
C-CHzOCO), 6.75-6.90 (m, 4 H, Ar).
15 Hz,
Hz, ArOCH 2CH2 ), 4.20-4.30 (m, 2 H,
Synthesis
29
AC0-880 (7g): pale yellow wax; 1H NMR: (CDC1 3) ö 0.60-0.95 (br, 3 H, CH2CH 3), 1.05-1.45 (m, 21 H) and 1.55-1.80 (m, 7 H) (C(CH3h, C9H 19, and CH 3), 2.20-2.65 (m, 4 H, CH 2CH 2), 3.55-3.75 (m, CH 2CH 20), 3.85 (t, 2 H, J = 15 Hz, ArOCH2C!! 2), 4.11 (t, 2 H, J = 12 Hz, ArOCH 2CH2), 4.20-4.30 (m, 2 H, CHpCO), 6.75-6.90 (m, 4 H, Ar).
References: I. Sheppard, C.S.; Kamath, V. In Kirk-Othmer Encyclopedia of Chemica/
2. 3.
4. 5. 6. 7. 8. 9. 10. 11. 12. 13.
Technology; Mark, H.F., Othmer, D.F., Overberger, C.G., Seaborg, G.T., Eds.; Wiley: New York, 3. ed., 1981; Vol. 13; p 355. Overberger, C.G.; Hale, W.F.; Berenbaum, M.B.; Finestone, A.B. J. Am. Chem. Soc. 1954, 76, 6185. Sheppard, C.S. In Encyclopedia of Polymer Science and Engineering; Mark, H.F., Bikales, N.M., Overberger, C.G., Menges, G., Eds.; Wiley: New York, 2. ed., 1985; Vol. 2; p 143. MacLeay, R.E.; Sheppard, C.S. U.S. Pat. 3,931,143, 1976; Chem. Abstr. 1976, 84, 12258lv. Willes, J.M. lnd. Eng. Chem. 1949,41, 2272. Kotera, K.; Furusawa, K.; Takeda, Y. Kolloid Z. Z. Polym. 1970, 239, 677. Goodwin, J.W.; Heam, J.; Ho, C.C.; Ottewill, R.H. Brit. Polym. J. 1973, 5, 347. Griffin, W.C. J. Soc. Cosmet. Chem. 1949, 1, 311. Ottewill, R.H. Ann. Rep. A 1969, 66, 183. Napper, D.H. lnd. Eng. Chem. Prod. Res. Develop. 1970, 9, 467. Holmberg, K.; Hansen, B. Acta Chem. Scand. 1979, B 33, 410. Hassner, A.; Alexanian, V. Tetrahedron Lett. 1978, 4475. Shinkai, S.; Tsuji, H.; Hara, Y.; Manabe, 0. Bull. Chem. Soc. Jpn. 1981, 54, 631.
30
Chapter 3
Theory of Emulsion Polymerization
Summary: After an overview of free-radical polymerizalion, qualilalive and quantitative models developed for emulsion polymerizatlon will be discussed. The "zero-one" system, a special case in emulsion polymerization, will be discussed in greater detail. Some equations wil/ be derived to obtain entry (p) and exit (k) rate coefficients from raw conversion-time data at different stages of the emulsion polymerization reaction.
3.1 Introduetion
In the years following the appearance of the first papers on emulsion polymerization 1·2, the number of contributions to the onderstanding of the mechanism of emulsion polymerization have increased substantially. by Fryling and Harringtonl, Hohenstein and co-workers 4·5·6 ·7 , 8
Work done Kolthoff and
9
Dale , and Filette and Hohenstein , made a large contribution to the unravelling of the mechanism of emulsion polymerization. One of the most important models put forward in this period is the one by Harkins. In 1945 and 1946 short notes were published 10•11 • Descriptions with greater detail of this model appeared in 1947 12 and 1950 13 •
This model will be outlined briefly in Section 3.3, but first all
32
Chapter 3
relavant reactions occurring in free-radical polymerization will be discussed (Section 3.2). The presentation of the model of Harkins will be foliowed in Section 3.4 by a discussion of the effect of several parameters on the reaction rate, whereafter the model description of Smith and Ewart will be given.
In this description more
recent developments have been incorporated. In Section 3.6 the various nucleation models will be discussed. After this more general description of emulsion polymerization, attention will be drawn, in Section 3.7, to a more specific system: the "zero-one" system. Inisurfs will have an influence on those kinetic events, which involve the partiele surface, i.e., entry and exit. Hence, if the effect an inisurf has on the kineties is studied, this should be done in systems where entry and exit dominate, i.e. in "zero-one" systems. Within the boundaries of this system, Interval III kineties will be diseussed as well. This chapter will close with a discussion of the fate of the exited free radieals (Section 3.8).
3.2 Free-Radical Polymerization
As already outlined in Chapter 1, a free-radieal polymerization is a ehain reaction polymerization in which the aetive eentres are free radieals, and an emulsion polymerization is a partieular type of free-radieal polymerization. free-radical
reaelions
involved in
terminalion and ehain transfer
polymerization -
initiation,
All
propagation,
occur simultaneously. These reactions wiJl now be
explained in more detail.
Initiator Decomposition and Initiation. Chain polymerization eannot occur without an aetive eentre.
In free-radieal polymerization these aetive eentres are
free radieals. The possible sourees of free radicals have already been mentioned in Chapter 2. For ehemieal initiation, initiators undergo homolytic cleavage and form
Theory of Emu/sion Polymerization
33
two primary free radicals, according to:
2 R'
I
(3.1)
where I and R' denote the initiator and the formed primary free radical, respectively, and kd is the first-order rate constant for dissociation (s- 1). Two fates are possible for these free radicals. The fastest of these is the relatively facile recombination with the other primary free radical derived from the same initiator molecule (cage-effect) 14•15 , or reaction with other radical-bearing substances,
ultimately
producing
an
inert
species
incapable
of initiating
polymerization. The other fate is the actdition of a monomer unit to a primary free radical, called initiation:
R' + M
RM;
(3.2)
where M denotes monomer and RM; denotes a radkal composed of a primary free radical moiety and one single monomerunit now carrying the radical activity. The second-order rate constant for the addition of the first monomer unit to a primary free radical is denoted as kP 1 (dm3 mor 1 s- 1). Propagation. The addition of a monomer unit to a growing chain is called propagation.
A polymer chain with n+ I monomer units is formed from a chain
with n monomer units. The general equation for the propagation reaction is:
M~
where
M~
+ M
(3.3)
and M:+J represent growing polymer chains with n and n+ 1 monomer
units, respectively, and kP is the second-order rate constant for propagation (dm3 mor 1 s- 1), which is presumed to be virtually independent of chain length: the starting free radical influences the rate of propagation only for a very small number of consecutive monomer unit additions.
34
Chapter 3
Bimolecular Termination. Two growing polymer chains can react with each
other and form "non-living" polyrner material.
This reaction, called termination,
can occur in two distinct ways. The first one, called combination, is the reaction of the free radical ends of two growing chains, resulting in one saturated "non-living" polymer chain. The second one takes place by hydrogen abstraction. A hydrogen atom is transferred from one free radical to another, which results in the formation of two "non-living" polymer chains, one saturated, the other with a terminal double bond.
This process is called disproportionation.
A general
equation for terminalion is:
M~
+
M~ ---~
Non-living polymer
(3.4)
where k, is the second-order rate constant for termiaation (dm3 mor 1 s- 1). It was recognized by Benson and North 16, and has been shown conclusively by Adams et al. 11 , and by Russen et a/. 18•19 that terrnination is a ditfusion controlled process and depends on the weight fraction of polymer in the polymer particles and the length of the growing polymer chains.
Chain Transfer. This is the reaction by which an atom is transferred to (or
abstracted from) a growing polymer chain. The growth of the original polymer chain is stopped. The molecule T, from which the atom is abstracted (or to which the atom is transferred), is turned into a radical, which can reinitiate a new polymer chain (if sufficiently reactive):
M~
+ T
M. + T
(3.5)
where k" is the second-order rate constant for transfer (dm3 mol" 1 s- 1). In principle, polymerie free radicals can undergo chain transfer with just about any other species present in a polymerizing system. However, in emulsion polymerization, without any added chain transfer agent and otherwise usual conditions, it is chain transfer to monomer that is by far the most important and prevalent form of chain transfer
Theory of Emulsion Polymerization
(T=M).
35
This produces a monomerk free radkal and a terminally unsaturated
polymer chain.
3.3 Model of Harkins As already mentioned, a basic recipe for emulsion polymerization contains monomer, water, surfactant and initiator.
At the beginning of the reaction the
monoroer is dispersed in dropiets in the continuous aqueous phase by agitation. Depending on the amount of added surfactant, micelles of this surfactant will be present and some of the monoroer may be present in the hydrophobic core of these micelles. Usually, the initiator will be dissolved in the aqueous phase. The whole of the emulsion polymerization process in a batch reactor can be divided into three intervals.
Interval 1: Partiele Formation Only the case where the overall concentration of the added surfactant is higher than the Çritical Micelle Çoncentration (CMC) wiJl be considered, i.e, the case where micelles are present at the start of the reaction. Primary free radicals, formed by dissociation of the initiator, will enter the monoroer swollen micelles and will initiate the polymerization. Subsequently, the micelles will be converted into polymer particles. The entry of a radkal into a monoroer droplet is kinetically negligible due to the small total surface area of all the monomer droplets, as compared to the total surface area of all the monoroer swollen micelles.
The
formed particles will be stabilized by adsorption of extra surfactant from not as yet initiated micelles.
This model for partiele formation is called the micellar
nucleation model (the various nucleation models will be discussed more extensively
in Section 3.6). This partiele nucleation is thought to be the primary reason for the increase in reaction rate, which is one of the characteristics of this interval.
36
Chapter 3
When the surfactant concentration eventually drops below the CMC, due to adsorption of emulsifier molecules onto the monomer-polymer particles, partiele nucleation can only occur in the aqueous phase 11 and in the monomer droplets 10 • Since the monomer concentration in the aqueous phase is usually very low, this can not be an effective souree for polymer particles. As already mentioned, due to the low ratio of total surface area of the monomer dropiets to the total surface area of the particles, monomer dropiets do not play a significant role in partiele nucleation either. Thus, after the disappearance of the micelles, the nucleation rate will drop dramatically.
Interval 11: Constant Reaction Rate
In this interval all three possible phases are present: there will be an aqueous phase, polymer particles saturated with monomer, and monomer droplets. lt is usually characterized by a constant number of polymer particles and a constant
reaction rate.
It is presumed that the radical production does not change
significantly during the reaction, as is in the case of long-living initiators. formed radicals are absorbed by the growing polymer particles.
The
These polymer
particles are the main loci of polymerization and they grow in size. The monomer concentration inside the polymer particles is kept constant during this interval by diffusion of monomer from the monomer dropiets through the aqueous phase into the polymer particles.
In order to keep the monomer concentration inside the
polymer particles constant, not only the monomer which has recently been converted into polymer, needs to be replaced, but also the monomer which is needed for the swelling of the newly formed polymer, has to be incorporated. The monomer dropiets function as monomer reservoirs only.
At a certain conversion
these reservoirs are exhausted and disappear, which marks the end of Interval Il.
Interval III: Decreasing Reaction Rate
At the beginning of this final stage of the reaction only the polymer
Theory of Emulsion Polymerization
particles and the aqueous phase are present.
37
The monomer is present in both
phases at the saturation concentrations. This interval is characterized by a constant number of particles and by a decreasing rate of polymerization. The latter results from the decrease of the monomer concentration in the polymer particles, as polymerization proceeds. This depletion of monomer in the polymer phase leads to an increase in the intrapartiele viscosity, which eventually results in a slower terminalion rate. Consequently, the radical concentration wil! increase and so will the overall polymerization rate.
This is called the Trommsdorff, or gel, effect.
The end product is a latex: a dispersion of polymer particles in an aqueous medium, stabilized by surfactant. In this model, as it was first described by Harkins 12, it is presumed that the monomer is essentially insoluble in water and the polymer is miscible with the monomer in all proportions.
This type of system is referred to as an "ideal"
emulsion polymerization system. Fig. 3.1 shows idealised conversion-time and rate-time curves of an emulsion polymerization. The three different intervals are indicated in this figure. The Trommsdorff effect can be observed as an increase in the rate of polymerization in Interval lil.
In virtually any particular system some of the features,
shown in this idealised picture, may be absent, or other may be so dominant that they mask them.
Figure 3.1
ldealised Plot of Fractional Conversion (x) and Rate (dxldt) against Time (t) in an Emulsion Polymerization System.
38
Chapter 3
3.4 The Rate of Polymerization
The rate of polymerization
Rpo~
is in general given by the following
equation: (3.6)
Rf>Ol
where [M] and [R'] are the monomer and free radical concentrations, respectively. The constituentpartsof Eq. 3.6 will be discussed briefly. As long as propagation is chernically controlled, it can be presumed that kP
is independent of w/n, the weight fraction polyrner in the solution of polymer in monomer, in which the propagation is taking place. Purthermore, kP is independent of chain length, as already mentioned in Section 3.2. Thus, the so-caUed "long chain" value for kP can be used, except for very high wP. In emulsion polymerization, as described by Harkins (see Seetion 3.3), the principle loci of polymerization are the polymer particles. For that reason it bas become customary to use CM, inslead of [M], to denote the concentration of monomer in
the
polymer particles.
The value
of CM is deterrnined
thermodynarnically by the striving of the system to rninirnize its free energy: an equilibrium is reached between the opposing effeets of lowering the surface tension and of reducing the free energy of mixing.
It is generally accepted21 •22 •23 •24
that in Intervals I and 11 CM depends on the diameter of the partiele for small particles up to about 30 nm and that CM is constant for particles above this size. In Interval lil CM will deerease with încreasing conversion and can be trivially calculated by means of a mass balance·25 • It is assumed that CM is the same in each polymer particle. The radical concentration [R'] is usually unknown and is difficult to predict.
This straightforward calculation is only applicable with sparingly warer-soluble monomers, where the arnount of monomer dissolved in the aqueous phase, is negligible. Since in these studies maînly styrene, which has a maximum water solubility of 4.3 10~ 3 mol dm~' at 323 K, is used as monomer, the foregoing assumption holds.
Theory of Emu/sion Polymerization
39
According to the Harkins Model, the rate of polymerization can be described as the rate of reaction within a single particle, times the number of particles per unit volume of aqueous phase (NJ. The reaction rate per partiele is the average number of radicals per partiele (n) multiplied by (kp CM)INAv• where NAv is Avogadro's constant. The overall polymerization rate in mole per unit of time, per unit volume of aqueous phase, is then expressed as: kp cMnNc
(3.7)
NAv
The parameters now most difficult to predict are
n
and Ne.
lt is most
convenient to express the rate of polymerization as the increase in fractional conversion of monomer into polymer per unit of time.
This is achieved by
dividing Eq. 3.7 by the initia! number of moles of monomer present, per unit volume of aqueous phase, n~: dx
(3.8)
dt
where dx/dt represents the change in percentage conversion per unit of time. Since n~ is known from the recipe, this equation can be used if the values of Ne, kP and
CM are known, and Eq. 3.8 can then be simplified to:
dx
dt where the value of A
An
(3.9)
kP CM Ne I (n~ NA.), this, so-called conversion factor, is
independent of conversion in Interval Il.
3
40
3.5 The Smith and Ewart Theory
Besides the qualitative theories about emulsion polymerization, such as the Harkins model, quantitative theories have also been developed. With these theories it should be possible to predict fi and Ne. One of the earliest theories was publisbed by Corrin 26 in 1947.
This
theory is concerned only with that part of the process, that occurs after free micellar soap has disappeared from the aqueous phase. A more general theory was developed by Smith and Ewart27 •
This theory is also based on the model of
Harkins.
3.5.1 The Recurrence Relationships
In their theory, Smith and Ewart formulated population balance equations for the relative number of polymer particles containing i free radicals, N;.
The
average number of radicals per partiele is then defined as:
n
(3.10)
For convenience N; is normalized, so that: (3.11)
The population balance equations are derived by taking the following possible processes into account:*
More recent additions to the theory of Smith and Ewart have been incorporated in the present description of the various kinetic events.
Theory of Emulsion Polymerization
Entry.
41
In emulsion polymerization the initiator used is in general water
soluble and the polymer particles are the main loci of polymerization.
This
suggests that free radicals are created in the aqueous phase by dissociation of the initiator, and that they ultimately migrate into the particles. As has been pointed out by several authors24 '28 •29 •30•31 •32, the newly formed free radicals do not show a tendency to enter a polymer partiele or micelle. Free radicals have to add a critica!, monomer-depending number of monomer units (z), before they become surface-active and will adsorb onto the surface of a polymer partiele or micelle, foliowed by true entry.
Entry will increase the number of radicais in a given
partiele by unity and will be an important determinant of the rate of polymerization.
The rate of entry is quantified by a pseudo-first-order rate
coefficient p and represents the average number of free radicals entering a latex partiele per unit of time (s. 1). The quantity p depends on the initiator type, the initiator concentration, the partiele number, and the type of monomer.
After a
radical enters a particle, it will propagate. Although propagation does not influence i directly, it will affect events which do influence i. For instance, the propagation
of a free radkal and thus the growth of a polymer chain decreases the likelibood of termination and exit.
Given the definition of p, the following equation for the
change in the population of loci with i radicals can be derived: (3.12) enlry
Termination.
The presence of two or more radicals in a single polymer
partiele will lead to termination of two of these radicals and to the production of "non-living" polymer materiaL
As already mentioned before, terminatien is a
diffusion-controlled process, and, therefore, the second-order rate constant for termination (k,) is a function of conversion.
It can also be inferred that
macroscopie terminatien rate constants are a function of chain length. From this it is clear that there is not just one k, for a given system.
Nevertheless, it is
presumed for simplicity that termination can be characterized by one single
42
Chapter 3
termination rate constant k,. It is customary 33 to use a pseudo-first-order rate constant, designated c, to
quantify termination in emulsion polymerization: c is the frequency of bimolecular termination per free radical per partiele (s- 1) (c = k,l NA.Vs• where V' is the swollen volume of a polymer particle). According to the above definition the contribution of terminatien to dNJ dt must be:
(i+2)(i+l)cN;.2
-
i(i-l)cN;
(3.13)
termlootion
Exit. Apart from termination, a growing chain can he stopped by transfer of
the active centre toanother molecule. This other molecule may be monomer, but it may also be any added substance with the ability of chain transfer and even polymer.
Similarly to propagation, transfer does not affect i, but it produces a
"non-living" polymer chain and a reactive site (free radical). In the situation where no added chain transfer agent is present and transfer to polymer is neglected, transfer wi11 only be to monomer. This produces a monomeric free radical. Just as free radicals can enter a polymer particle, it must also be possible that free radicals exit from a polymer partîcle. Based on the nature of the various free-radical species, it is, in the case of styrene, argued that only monomeric free radicals can exit from a polymer particle. Exit is then presumed to follow a threestep mechanism34 : 1) a monomerk free radical must be generated by transfer, 2) this spedes must diffuse through the interior of the latex partiele to the partiele surface, and 3) thereupon it must undergo desorption. Exit of a free radical leads to a decrease of the radical concentration in the polymer partiele and thus to a decrease in the rate of polymerization. Normally, the rate of exit is described by an overall first-order rate coefficient, usua11y denoted as k: this is the freqilency with which desorption occurs from a single polymer particle, per free radical (s- 1). above that the maximum value of k is k,,CM.
lt wil! he clear from the
Here k" is the second-order rate
constant for transfer. The exit-related contri bution to dNJ dt is:
Theory of Emulsion Polymerization
dN; dt
-
'
43
(i+1)kN I ,.
=
2R. /
a
Aqueousph
hetemtennin
withentrant free radtcal (l= -I
Figure 3.2
Exil
V
Aqueous phase bomotennination with ar10ther desorbed free radical (l:O
or
Re-entry M~
~·
~
ll>w"--/1Propagation . Terminalion
Kinetic Events in Emulsion Polymerization.
44
Chapter 3
3.5.2 Solutions to the Recurrence Relationships
Smith and Ewart themselves did not present a general solution to their recurrence relationships. Instead, they considered three limiting cases.
Case 1: The number of free radicals per polymer partiele is small compared with unity. This situation occurs when exit of free radicals out of the loci is much faster than entry of free radicals into such loci, i.e., p « k. Termination can be neglected due to the Iow value of i. The recurrence relationships then reduce to:
(3.16)
In the steady-state conditions: dN0 I dt fl
=Ni I N
0
= pIk.
= 0.
Since Ni «
N0 , it follows that:
The overall rate of polymerization Rpo1 per unit volume of
aqueous phase can then be written as:
(3.17)
Case 2: The number of free radicals per polymer partiele is approximately equal to 0.5. This is the best known solution in the Smith-Ewart theory.
The following
conditions have to be satisfied simultaneously: (I) bimolecular termination is instantaneous when a second radkal enters a partiele containing one radical, and (2) radical exit is negligible, i.e., k « p « c.
Trivially, this leads to N0 = N1•
Since it follows from the above mentioned conditions that only particles with zero
Theory of Emu/sion Polymerization
or one radical will be present, N0
45
= N =0.5 and thus fi =0.5. 1
The overall rate of
polymerization per unit volume of aqueous phase can then be expressed as:
(3.18)
Case 3: The number of Jree radicals per polymer partiele is large compared with unity. This situation occurs when entry is much faster than bimolecular termination,
i.e., c « p.
In the treatment of this case Smith and Ewart neglected exit.
By
assuming that fi is large, the system can be approximated as one, in which all particles contain the same number of free radicals and for which the steady-state condition is: p = 2cfi 2 •
The overall rate of polymerization per unit volume of
aqueous phase can then be expressed as:
R
1 po
)0.5
_ ( p Ne -kCM-P
2c
N
(3.19)
Av
In later days new solutions to the Smith-Ewart equations were derived by, for example, Stockmayer15 , O'Toole36 , and Ugelstad et al. 31 .
Even though
these solutions were more genera!, all of them suffered from the shortcoming of being applicable to steady states only.
The first time-dependent solution of the
Smith-Ewart equations was obtained by Gilbert and Napper38 •
Due to the
assumptions used in their derivation, the solution obtained was limited to SmithEwart Case 1 kinetics.
A short time later, a general time-dependent solution for
Smith-Ewart Cases 1 and 2 was found by the same group39·40 .
46
Chapter 3
3.5.3 Partiele Number
Smith and Ewart also developed an expression for the partiele number under the lirnitations of their Case 2.
Since their theory is based on the model of
Harkins, they presumed that particles are formed by micellar nucleation only. Other nucleation models will be discussed in the next section.
In order to develop an expression for the partiele number they also assumed that: I) surfactant is only present in micelles or adsorbed at the surface of polymer particles.
The amount adsorbed at the surface of the monomer dropiets and
dissolved in the water phase is neglected. 2) The surface-coverage by one surfactant molecule is independent of the kind of surface (rnicelle-water, polymer-water),
i.e., the total surface of micelles and polymer particles is constant in Interval L 3) The reaction rate per partiele is constant in the nucleation stage of the reaction, since it is assumed that the monomer/polymer ratio is constant as wel!.
Smith and Ewart took two limiting cases into account. The first case, too
many particles, prediets more and the second case, too few particles, prediets less particles than in the actual situation.
Too many particles. As long as micelles are present, it is assumed that they capture all available radicals.
This leads to a calculation of too many particles
since free radicals can also be captured by already existing particles.
Too few particles. A given interfacial area is assumed to capture the same number of free radicals regardless of the size (and therefore curvature) of the particle.
It is known from classica! diffusion theory that the flux of diffusing
matter across a unit area of interface is inversely proportional to the radius of curvature at the interface.
This means that more particles will be formed than
predicted, since bigger particles are less effective in capturing free radicals.
Both cases lead to a formula of the same form for calculating the number of particles formed during the nucleation stage (Interval I) in an emulsion
47
Theory of Emulsion Polymerization
polymerization system:
x
[:·r("
(3.20)
s)""
where p, is the total entry-rate coefficient per unit of volume of aqueous phase (s· 1 m· 3), J1 is the volume growth rate of a partiele (m3 s· 1), 2
occupied by unit mass of surfactant (m g-
1 ),
ti1
is the interfacial area
and S is the total mass of surfactant
contained in the system per unit volume of aqueous phase (g m- 3). Further, X is a constant of which the true value lies in between 0.37 (too few particles) and 0.53
(too many particles).
3.6 Nucleation Models Since the pioneering work of Harkins 12 , and Smith and Ewart27 on the partiele formation in emulsion polymerization, some other nucleation models have been developed.
In this section a qualitative description of these models will be
given.
Micellar Nucleation.
As already explained, radicals will be absorbed into
monomer-swollen surfactant micelles, which are then transformed into polymer particles.
As long as micelles are present in the system, new particles will be
formed and nucleation ceases with the disappearance of the micelles. This model is able to predict the partiele number for systems with sparingly water-soluble monomers (e.g., styrene) and with surfactauts with a Jow CMC41 •
The
shortcomings of this theory are that I) in some cases particles are formed even when no micelles are present, 2) partiele numbers estimated with Eq. 3.20 deviate about a factor of two from what is found experimentally, even for styrene, and 3) more water-soluble monomers do not fit the theory at all.
48
Chapter 3
Homogeneaus Nucleation. A radical will react with dissolved monomer to
gîve a growîng polymer chaîn dissolved in the aqueous phase. radical will continue to grow.
Thîs oligomeric
Upon reaching a critica! chaîn Iength it will
precipitate and form a polymer partiele stabilized against flocculatîon by adsorbed emulsifier and swollen with absorbed monomer.
The critica! chain Iength is
directly related to the type of monomer used. The micelles act only as surfactant reservoirs, and the number of particles is only determined by the amount of surfactant and its intrinsic efficiency. This model, first described by Priest42 , and later quantified by Roe43 , by Hansen and Ugelstad44 , and by Fitch and Tsai 45 has become known as the "HUFT'' (Hansen-Ugelstad-Fitch-Tsai) theory46 •
Homogeneous!Coagulative Nucleation.
This is an extension of the HUFT
model, introduced by Gilbert and Napper and co-workers 47 •48 •49 •
As in the case
of the homogeneaus nucleation mechanism, radicals will react with monomer and will precipitate above a critica! chain length.
The so formed particles are not
"true" latex particles and are called "precursor" particles.
These particles differ
from mature latex particles in at least two important respects: firstly, they are colloidally unstable, undergoing coagulatîon with other precursor particles or a mature latex particle; secondly, they polymerize very slowly.
This slow
polymerization may arise from the reduced swelling of the particles by monomer 1.47 and/or from the rapid exit of any free radical in the precursor particles due to their small size23 • The precursor particles are presumed to grow mainly by coagulation, although some growth must occur by polymerization.
Mature latex
particles are formed by coagulation of precursor particles, until the coagulated entities are stabie and clearly segregated hydrophilic and lipophilic regions have been formed.
These mature latex particles grow rapidly.
Coagulation of a
precursor partiele with a mature latex partiele wil! contribute to the process of free-radical en try. When the surfactant is exhausted, its surface coverage per area on the mature latex particles decreases more rapidly than on the precursor particles, due to the higher growth rate of the mature particles.
This leads to a situation
where precursor partiele-mature partiele coagulation is favoured over coagulation of
Theory of Emulsion Polymerization
49
two precursor particles and the partiele nucleation rate falls rapidly.
Droplet Nucleation.
Usually, monomee dropiets are not believed to play
any other role in emulsion polymerization than being a reservoir of monomer. In a series of papers5051.s2.s3.s4.s5, it has been shown, however, that in cases with very smal! monomer droplets, these may become important, or even the sole, loci for partiele nucleation.
The system may then be regarded as a microsuspension
polymerization with a water-soluble initiator.
Up till now, experimental data can not refute one of the nucleation models, and it has been shown by Morrison et al. 56 that none of the above mentioned mechanisms is solely responsible for the partiele formation above the CMC. These authors condurled that the nucleation mechanism above the CMC involves both homogeneouslcoagulative nucleation and growth of captured oligomeric radicals in micelles with the incorporation of the extended entry model, as developed by Maxwell et a/. 32 • The capture of oligomeric radicals in micelles competes with the capture of these radicals in (pre-)existing particles (entry). For systems with low or zero surfactant concentration, nucleation is dominated by homogeneous nucleation and by coagulation of precursor particles.
3.7 Zero-One System33
As described in Section 3.5, the value of fi is determined by values of the rate parameters p, k and c. lt is very difficult to determine the values of any of these parameters independently. Thus, it is very difficult to extract the values of specific rate parameters from the experimental values of ii. This problem can be circumvented by not consirlering emulsion polymerizations in genera!, but rather by restricting oneself to specific types of kinetic systems.
One of these specific
systems is the so-called "zero-one" system, in which a latex partiele contains at
50
Chapter 3
most one free radical at any time. Moreover, it has been shown 19 recently that the Smith-Ewart equations, as they have been described in Section 3.5, are invalid for systems where fi is larger than 0.5. Proper chain-length dependent termination has to he incorporated for these systems.
Therefore, only zero-one systems are of
consideration in this investigation. In the "zero-one" system it is assumed that re-entry can he neglected as a first approach.
Re-entry is the entry of an exited free radical into a partiele
irrespective of this being the "parent" partiele or another particle. The process of re-entry will he discussed in greater detail in Section 3.8. The parameter p is then denoted as the rate of entry of free radicals per particle, regardless of their history, and k is denoted as the rate of exit of free radicals per particle, regardless of their ultimate destiny. The assumption that a partiele contains, at all times, either one single free radical or none, is accomplished when the entry of a second radical leads to instantaneous termination. Thus, terminalion is so rapid that it is not rate determining (p, k « c), and only p and k need to he considered in the kinetic analysis of a "zero-one" system.
Hence, this system is very well suited for
studying the characteristics of inisurfs.
3.7.1 Population Balance Equations
According to the above given definition, the populations of the two species of particles being present, are denoted as N0 and N 1, and the total population is normalized for reasons of convenience:
fl =
N 1•
The different species can he
related to each other as follows. A partiele with zero free radicals is formed by exit of a radical out of a partiele with one free radical or by entry of a free radical into a partiele with one free radical, foliowed by instantaneous termînation.
A
partiele with zero free radicals is converted into a partiele with one free radical by entry of a free radical. From this, the following population balance equations can he derived:
Theory of Emu/sion Polymerization
51
(3.21)
Assuming that p and k are constant, Eq. 3.21 can be solved trivially. lt is then found that: p (2p
+
k)
P
(2p + k)
) exp{ -(2p + k)t}
(3.22)
where ii0 is the initial (t =0) value of ii. The steady state value of ii (fl5 ,,) is found by taldng the long time limit of Eq. 3.22:
The assumption that p and k are constant, is valid in genera!.
In the
experimental situation of a seeded system (see Section 4.2.3), there is usually an induction period at the beginning of a polymerization, which is associated with the residual dissolved oxygen acting as an inhibitor. Thus, p increases from zero to a constant value, but this happens on a time scale much smaller than that of the kinetic events that govern iï. Since CM is constant and the surface of a partiele does not change much during Interval 11 of a polymerization, k will be constant during Interval 11 (Section 3.7.2). Interval 111 will be discussed in Section 3.7.3.
3.7.2 Slope and Intercept Method 23
A theoretica! expression for the variation of conversion with time in a simple "zero-one" system can be obtained by substituting the population Eq. 3.22 into the conversion Eq. 3.9. During Interval 11 kP, CM, and Ne are constant and so
52
Chapter 3
is the conversion factor from Eq. 3.9.
Given this, substitution of Eq. 3.22 into
Eq. 3.9 and integration yields:
x(t) -x0
=
A (2p + k)
[pt (n - (2p p k) J(l +
0
+
oxp{-(2p •
k)t})l (3.24)
where Xo is the fractional conversion at t = 0.
For long reaction times the
exponential part of Eq. 3.24 becomes negligible and this equation reduces to:
A (2p + k)
(pt
+
no
(3.25)
This theoretica! expression shows that at long times the conversion-time curve can be described by a straight line. It is indeed known from experiment that the rate of polymerization in this type of systems climbs from zero to a steady state value, whereafter x(t) is linear: x(t) - x0 = a + bt.
Fig 3.3 shows a typical
conversion-time curve for a seeded system, in which intercept a and slope b are
Slope b
Figure 3.3
Typical Plot of Fractional Conversion (x) vs. Time (t) for a Seeded Emulsion Polymerization System.
indicated. The intercept a and the slope b can be obtained from experimental data. From Eq. 3.25 it follows that:
Theory of Emulsion Polymerization
53
(3.26)
This metbod has the advantage that by rnaicing use of the intercept, the information contained in the approach to steady state is effectively incorporated as well.
In terms of accuracy this metbod is also preferabie to, for example, a
non-linear least-square fit method.
The intercept is not subject to as much
uncertainty as any single early-time value of conversion, and thus the problem of noise in the approach to the steady state region of the experimental data is, to a large extent, overcome. Another souree of uncertainty is the error in the value of the conversion factor A. lt can be easily shown that an uncertainty of 10% in the value of A can lead to an ultimate uncertainty of more than 50% in the value of p. Fortunately, for "zero-one" systems A can be determined with a high accuracy from kinetic data alone, and thus this difficulty can be overcome.
3.7.3 Interval 111 Kinetics
Up till now, it bas been presumed that CM is constant. But during Interval III CM decreases with conversion, which means that the above derived expressions do not hold for Interval lil data. However, most of the kinetic experiments have been carried out in Interval lil for reasons, which will be explained in Chapter 4. Thus, a closer look at Interval lil is necessary.
As might be clear from all of the above, Ne and kP are constant, at least in the early stage of this polymerization phase. Presuming that in Interval lil all the monomer is contained in the particle, i.e., the monomer is not water soluble, and that the volume contraction due to polymerization can be neglected, it can be shown that:
54
Chapter 3
(3.27)
(1-x) c.~
where C~ is the value of CM at the beginning of the experiment.
Substituting
Eq. 3.27 into Eq. 3.8 leads to: -dln(I-x) dt
(3.28)
This can be simplified to:
-dln(l-x)
dt
kr_n __ VsNAv
A'n
(3.29)
where V, is strictly speaking the swollen partiele volume at the beginning of the experiment, but since the volume contraction has been assumed to be negligible, Vs is just the swollen partiele volume and is constant.
The conversion factor for
Interval III is denoted as A'. As can be seen, Eq. 3.28 takes a form simHar to Eq. 3.9, indicating that data from Interval III experiments should be analyzed by using values of -ln(l-x) instead of the conversion x directly. Since Eqs. 3.9 and 3.28 are of the same form, the method given above for the calculation of p and k can be altered trivially to take variation of CM with conversion into account. Although it seems that p and k can be calculated from Interval lil data, a few problems arise when applying this altered method. Firstly, it is presumed that p and k are independent of conversion. Because of the decrease of CM and, hence, the increase of wl' during Interval lil, this assumption may no Jonger be valid for this interval. mainly determined by aqueous-phase events
32
,
In the case of p, which is
it might intuitively be anticipated
that it is relatively insensitive to variations in CM, except for low CM. In the case of k, it is expected that this rate parameter changes throughout Interval lil. Only when w" changes very little during the course of the experiment can k be presumed
Theory of Emu/sion Polymerization
constant.
55
Only so-called y-relaxation experiments (see Chapter 7) meet this
criterion, since relaxation occurs over a smal! conversion range, and these are therefore recommended for determining Interval lil values of k. Secondly, with a decrease in CM, the viscosity increases and this leads to a decrease in the rate of terrnination.
If c becomes sufficiently smal!, the
instantaneous terrnination assumption may no Jonger be tenable, and the "zero-one" approximation may no Jonger be valid. In those cases the method for deterrnining p and k loses its applicability.
3.8 The Fate Parameter
Up till now, no attention has been paid to the fate of the free radicals which have exited from a partiele into the aqueous phase. For clarity it has been assumed that exit is a first-order loss process.
However, this is only the case when an
exited free radical homo-terminates (with another exited free radical). actually two other fates possible for the exited free radicals: latex particle.
There are
1) re-entry into a
Most like1y, this wil! be a latex partiele other than the one, from
which this free radical exited.
Exit is then not a true radical loss process.
2) Termination in the aqueous phase with an initiator-derived free radical.
This
hetero-termination does not only make exit a loss process, but it also reduces the effective rate of entry. Ugelstad et al. 31 were the first to consicter the effect of reentry (read exit) on the overall rate of entry. expanded by Ugelstad and Hansen
44
•
This work has been reviewed and
However, Whang et al. 51 simplified the
model and proposed that p be correctly expressed as: (3.30) where PA is defined as being the value of p in the absence of exit and a is known as the fate parameter, which has a value between -1 and 1. Eq. 3.30 expresses all the possible fates of the desorbed free radicals. They may either re-enter a latex
56
partiele (a.
= 1),
undergo aqueous-phase homo-termination (a. = 0), or experience
aqueous-phase hetero-termination (a.
= -I).
The advantage of applying a fate
parameter is that the complex model of emulsion polymerization can be simplified. A disadvantage is that a value of a. does not describe one unique physical situation.
Nevertheless, the fate parameter will be applied because of its simplicity. Substitution of Eq. 3.30 into Eq. 3.21 gives:
(3.31)
Solving this set of equations and substitution into the conversion equation, results in an equation, which again describes a straight line at long times, comparable with Eq. 3.24: k
A~ (2a.a)
(3.32)
with: G
2a.b 2 +A( 1 - a.)b A(A - 2b)
(3.33) F
0.5
+
2G + (l - a.) + 4a.ÏÏ0
---=-------..,:-----------::-:: 1
+4G(a. + 1)
where a and b are the intercept and slope of the long-time experimental conversion-time curve, respectively, and tï0 is the initia! value of tï. Eq. 3.32 is not applicable in the case of a.
=0 and Eq. 3.26 bas to be applied.
With the same ease
as already explained in Section 3.7.3, Interval III data can be used by plotting
Theory of Emulsion Polymerization
57
-ln(l-x) versus time and by determining intercept and slope of the long time straight line.
References 1. Dinsmore, R.P. U.S. Pat. 1,732,795, 1929; Chem. Abstr. 1930, 24, 266. 2. Luther, M.; Heuck, C. U.S. Pat. 1,860,681, 1932; Chem. Abstr. 1932, 26, 3804. 3. Fryling, C.F.; Harrington, E.W. lnd. Eng. Chem. 1944, 36, 114. 4. Hohenstein, W .P.; Vingiello, F.; Mark, H. India Rubber Wld. 1944, 1JO, 291. 5. Hohenstein, W.P.; Siggia, S.; Mark, H.1ndia Rubber Wld. 1944, 111, 173. 6. Siggia, S.; Hohenstein, W.P.; Mark, H.1ndia Rubber Wld. 1945, 111, 436. 7. Hohenstein, W.P.; Mark, H. J. Polym. Sci. 1946, 1, 549. 8. Kolthoff, I.M.; Dale, W.J. J. Am. Chem. Soc. 1945, 67, 1672. 9. Frilette, V.J.; Hohenstein, W.P. J. Polym. Sci. 1948, 3, 22. 10. Harkins, W.D. J. Chem. Phys. 1945, 13, 381. 11. Harkins, W.D. J. Chem. Phys. 1946, 14, 47. 12. Harkins, W.D. J. Am. Chem. Soc. 1947, 69, 1428. 13. Harkins, W.D. J. Polym. Sci. 1950, 5, 217. 14. Franck, J.; Rabinowitsch, E. Trans. Faraday Soc. 1934, 30, 120. 15. Rabinowitsch, E.; Wood, W.C. Trans. Faraday Soc. 1936, 32, 1381. 16. Benson, W.S.; North, A.M. J. Am. Chem. Soc. 1962, 84, 935. 17. Adams, M.E.; Russell, G.T.; Casey, B.S.; Gilbert, R.G.; Napper, D.H.; Sangster, D.F. Macromolecules 1990, 23, 4624. 18. Russell, G.T.; Gilbert, R.G.; Napper, D.H. Macromolecules 1992, 25, 2459. 19. Russen, G.T.; Gilbert, R.G.; Napper, D.H. Macromolecules 1993, 26, 3538. 20. Ballard, M.J.; Napper, D.II.; Gilbert, R.G.; Sangster, D.F. J. Polym. Sci., Polym. Chem. Ed. 1986, 24, 1027. 21. Morton, M.; Kaizermann, S.; Altier, M.W. J. Colloid Sci. 1954, 9, 300. 22. Gardon, J.L. J. Polym. Sci., Polym. Chem. Ed., Part A-11968, 6, 2859. 23. Hawkett, B.S.; Napper, D.H; Gilbert, R.G. J. Chem. Soc., Faraday Trans. 1 1980, 76, 1323. 24. Penboss, LA.; Gilbert, R.G.; Napper, D.H. J. Chem. Soc., Faraday Trans. l 1986, 82, 2247. 25. Lane, W.H. 1nd. Eng. Chem. 1946, 18, 295. 26. Corrin, M.L. J. Polym. Sci. 1947, 2, 257. 27. Smith, W.V.; Ewart, R.H. J. Chem. Phys. 1948, 16, 592. 28. Penboss, LA.; Gilbert, R.G.; Nappper, D.H. J. Chem. Soc., Faraday, Trans. 1 1983, 79, 1257. 29. Alexander, A.E.; Napper, D.H. Prog. Polym. Sci. 1971, 3, 145. 30. Nomura, M.; Harada, M.; Eguchi, W.; Nagata, S. Polym. Prepr.-Am. Chem. Soc., Div. Polym. Chem. 1975, 16, 217.
58
Chapter 3
31. Barrett, K.E.J. Dispersion Polymerization in Organic Media; Wiley, New York, 1975. 32. Maxwell, I.A.; Morrison, B.R.; Napper, D.H.; Gilbert, R.G. Macromolecules 1991, 24, 1629. 33. Gilbert, R.G.; Napper, D.H. J. MacromoL Sci., Rev. Macromol. Chem. Phys. 1983, C23, 127. 34. Nomura, M.; Harada, M. J. Appl. Polym. Sci. 1981, 26, 17. 35. Stockmayer, W.H. J. Polym. Sci. 1957, 24, 314. 36. O'Toole, J.T. J. Appl. Polym. Sci. 1965, 9, 1291. 37. Ugelstad, J.; M~rk, P.C.; Aasen, J.O. J. Polym. Sci., Polym. Chem. Ed., Part A-l 1967, 5, 2281. 38. Gilbert, R.G.; Napper, D.H. J. Chem. Soc., Faraday Trans. /1974, 70, 391. 39. Hawkett, B.S.; Napper, D.H.; Gilbert, R.G. J. Chem. Soc., Faraday Trans. I 1975, 71' 2288. 40. Hawkett, B.S.; Napper, D.H.; Gilbert, R.G. J. Chem. Soc., Faraday Trans. 1 1977, 73, 690. 41. Gerens, H. Ber. Bunsenges. Phys. Chem. 1963, 67, 741. 42. Priest, W.J. J. Phys. Chem. 1952, 56, 1077. 43. Roe, C.P. lnd. Eng. Chem. 1968, 60, 20. 44. Ugelstad, J.; Hansen, F.K. Rubber Chem. Techno[. 1976, 49, 536. 45. Fitch, R.M.; Tsai, C.H. In Polymer Colloids; Fitch, R.M., Ed.; Plenum Press: New York, 1971; p 73. 46. Hansen, F.K., Ugelstad, J. In Emulsion Polymerization; Piirma, 1., Ed.; Academie Press: New York, 1982; p 45. 47. Lichti, G.; Gilbert, R.G.; Napper, D.H. J. Polym. Sci., Polym. Chem. Ed., Part A-l 1983, 21, 269. 48. Feeney, P.J.; Napper, D.H.; Gilbert, R.G. Macromolecules 1976, 49, 536. 49. Richards, J.R.; Congalidis, J.P.; Gilbert, R.G J. Appl. Polym. Sci. 1989, 37, 2727. 50. Sujkov, A.V.; Grizkova, LA.; Medvedev, S.S. Kolloid Z. 1972, 34, 203; Eng. Trans!.: Colloid J. USSR 1972, 34, 154. 51. Ugelstad, J.; El-Aasser, M.S.; Vanderhoff, J.W. J. Polym. Sci., Polym. Lett. Ed. 1973, 11, 503. 52. Ugelstad, J.; Hansen, F.K.; Lange, S. Makromol. Chem. 1974, 175, 507. 53. Hansen, F.K.; Ofstad, E.B.; Ugelstad, J. In Theory and Practice of Emulsion Technology; Smith, A.L., Ed.; Acadamic Press, New York, 1976; p 13. 54. Azad, A.R.M.; Ugelstad, J; Fitch, R.M.; Hansen, F.K. ACS Symp. Ser. 1976, 24, 1. 55. Hansen, F.K.; Ugelstad, J. J. Polym. Sci., Polym. Chem. Ed., Part A-1 1979, 17, 3069. 56. Morrison, B.R.; Maxwell, LA.; Gilbert, R.G.; Napper, D.H. ACS Symp. Ser. 1992, 492, 28. 57. Whang, B.C.Y.; Napper, D.H.; Ballard, M.J.; Gilbert, R.G.; Lichti, G. J. Chem. Soc. Faraday Trans. 1 1982, 78, 1117.
Chapter 4
Experimental Procedures
Summary: In this chapter the various experimental and analytica! techniques used in monitoring the emulsion polymerizations wilt be described. The chapter will close with a description of the various methods used in determining important characteristics, e.g., surface tension of an aqueous salution and dissociation rate of the synthesized inisurfs.
4.1 Introduetion
In order to enhance the clarity of the subsequent chapters, the experimental techniques used in this investigation will be described and explained in this chapter·.
The various reactor designs used and the different polymerization
procedures will be described (Section 4.2), along with the preparation and eleaning of a seed latex (Section 4.3), and the various ways of monitoring an emulsion polymerization (Section 4.4).
The various ways of measuring partiele size and
partiele size distribution wiJl be explained in Section 4.5.
Surface tension
measurements of aqueous solutions of inisurf wiJl be discussed in Section 4.6.
•
The techniques used for the syntheses of the various inisurfs have been described in Chapter 2.
60
Chapter4
Finally a description of the measurements used for determining the dissociation rate of the synthesized initiator and inisurfs will be given in Section 4.7.
4.2 Emulsion Polymerization Reaelions
In the course of this investigation two types of emulsion polymerization reaelions have been carried out in different set-ups.
Besides ab initia reactions,
seeded reaelions have been carried out. Befare going into more detail about these reactions, a description of the various types of reactors will be given.
4.2.1 Reactor Design
Three different reactors have been used, depending on the reaelions investigated. The different set-ups wil! now be discussed.
4.2.1.1 Batch reactor
For the ab initia reaelions (see Section 4.2.2 ), a glass-jacketed 1 dm3 stainless steel reactor, equipped with a twelve-blade flat-blade turbine impeller, thermocouple and sample withdrawing tube, was used. This reactor (see Fig. 4.1) could also be flushed with an inert gas, and it could be evacuated by a pump. The reactions were monitored by taking samples to delermine conversion and/or partiele size.
4.2.1.2 Dilalometer
A dilalometer is one of the two reactor designs used for monitoring seeded reaelions (see Section 4.2.3). Originally, dilalometers were used for measuring the
Experimental Procedures
61
1.
2. 3. 4. 5. 6. 7. 8.
9. 10. 11. 12.
13. 14. 15.
Figure 4.1
Stainless Steel Reactor Glass Jacket Electric Motor Speed lndicating Controller Twelve-Blade Flat-Blade Turbine Raffles Chromel-Alumel Thermocouple Temperature Transmitter to Thermostalie Bath Filling Tube with Funnel Sample Tube Bottom Tube Gas Tube Evacuation Tube Pressure Indicator Safety Valve
Batch Reactor
thermal expansion or contraction of liquids or solids. When it was found that a contraction in volume occurred during polymerization, the dilatometer was adapted to study these reactions 1• A dilatometer for studying polymerization rates consists of a reaction vessel, a volume-sensing device, associated equipment for filling, deoxygenation and stirring, and a temperature-control apparatus. Dilatometers for emulsion polymerization require provisions for agitation to maintain homogencity of the oil in water emulsion. The use of dilatometers in emulsion polymerization was first reported by Fryling 2 •
He studied successfully the rate of emulsion copolymerization of
butadiene and styrene in end-over-end rotating tubular dilatometers. Figure 4.2 shows a dilatometer, which was used in our studies. Since most experiments were carried out in Interval lil, agitation with a magnetic stirrer was sufficient. This type of dilatometer is equipped with an automatic tracker to follow
62
Chapter 4
the meniscus of the liquid in the capillary automatically in order to determine the fractional conversion of the polymerization as a function of time. The accuracy of this metbod is determined by 1) the precision of the temperature control of the bath in which the dilatometer was immersed, so that a constant temperature of the dilatometer could be maintained, and 2) the size and uniformity of the measuring capillary. The disadvantage of this highly accurate apparatus is that during the reaction samples can only be taken or substances can only be added with serious difficulties and at the cost of losing accuracy.
4.2.1.3 Densimeter
The second reactor design used to monitor seeded reactions, was the densimeter. Since the density of polymer is higher than the density of monomer, the density of an emulsion will
increase
during
polymerization.
Thus,
emulsion
polymerization can be monitored by the change in density of the mixture.
Poeblein et al. 3 were the first to report on the
application of densimetry: a y-ray densimeter was used to measure conversion for control purposes in a continuous emulsion polymerization experiment. In our studies density was measured on-line by an Anton Paar DMA digital densimeter (see Fig. 4.3).
The measuring
principle of this densimeter is based on the change of the natura! frequency of a hollow oscillator when filled with different liquids or gases. The oscillator consists of a hollow elastic UFigure 4.2 One Piece Glass Vilatometer
shaped glass tube. This is electronically excited, perpendicular to the plane of the U-shaped glass tube, in an undamped harmonie oscillation. significant figures.
The density can be measured in six
Experimental Procedures
63
In our set-up, the latex was pumped from a reactor (similar to the one described in Section 4.2.1.1, but with a volume of 0.5 dm3) to the densimeter and via a peristaltic pump back into the reactor, as shown in Figure 4.3. The tubing used in the peristaltic pump, had to be mechanically stabie and inert. 1t must also be resistant to the dispersion pumped through, i.e., it must not be swollen by monomer and it must be free from any substances liable to be washed out by the emulsion.
Viton tubing (Watson Marlow Ltd., Falmouth, UK) was found to be
suitable in this investigation.
Densimeter
I I I I I I I I
@
---------
t IHo~ngboili Figure 4.3
Reactor
Densimeter Set-up
Since density strongly depends on temperature, the temperature of the latex had to be kept constant.
This was achieved by keeping the length of the tubing
between the reactor and the densimeter as short as possîble, by insulating the tubing from the reactor to the densimeter and back, and by thermostaling the densimeter.
Moreover, the densimeter had a built-in heat exchanger to maintain
the temperature of the latex at its reaction temperature. through the heat exchanger and then through the oscillator.
The latex flowed first
64
Chapter 4
4.2.2 Ab Initio Reaelions
During an ab initia experiment all three intervals, as described in Section 3.3, will occur. In this study, all the ab initia reactions were carried out in a batch reactor. Befare filling the reactor, it was flushed with argon and evacuated several times in order to remave oxygen and any volatile materials. First the inisurf was dissolved in doubly distilled de-ionized and argon-flushed water.
Details of the
recipe will be given in the relevant sections of the following chapters.
This
salution was then charged into the reactor along with the balance of the water needed.
Befare charging the reactor with the monomer (styrene, for synthesis,
Merck, Darmstadt, FRG, distilled under nitrogen at reduced pressure, and subsequently stored at 277 K under nitrogen until use), it was again flushed with argon and evacuated to remave the last traces of oxygen.
After charging the
monomer into the reactor, the reactor content was stirred at a speed of 250 rpm and heated to the reaction temperature (323 K or 343 K). During the reaction, samples were taken at regular intervals to determine the conversion by gravimetry (see Section 4.4.1) or for partiele size analyses (see Section 4.5).
4.2.3 Seeded Reaelions
In a "seeded" study, the reaction is deliberately started in either Interval II or lil.
In those studies a preformed cleaned monodisperse seed latex (for
preparation and cleaning see Section 4.3) is further polymerized by actdition of more monomer, surfactant and initiator. Since the nucleation stage of the emulsion polymerization, which is difficult to control, is avoided, kinetic data obtained from those reaelions give more reliable information about the various kinetic events and the mechanism of emulsion polymerization4·5.
One of the first groups to use
seeded polymerization reactions was Vanderhoff et al. 6·7·8•
They examined the
competitive growth of seeded polymerization of styrene under Case 3 conditions. Theoretica! and experimental studies of seeded polymerization have been published
Experimental Procedures
65
by for example Gilbert and Napper9•10•11 • The monomer added to the seed latex will preferentially migrate to the latex particle. Depending on the amount of monomer added, the monomer concentration in the particles will reach its equilibrium concentration,
Ct,;t.
lf suffïcient monomer
is added, monomer dropiets will be formed, and Interval II polymerization will (initially) take place.
If, however, the amount of the added monomer is just
sufficient or insufficient to fully saturate the particles with monomer, the system will commence under conditions corresponding to Interval lil.
In this case, the
concentration of monomer at the beginning of the polymerization, C~. can be calculated as follows, assuming molar-volume additivity: 0
ms mw5
0
c~
ns
Vs
0
ms
(4.1) 0
mps + --
Ps
PPs
where n~ is the initia! number of moles of styrene per particle, V, is the swollen volume of a particle, m~ and m~5 are the initial amounts of styrene and polystyrene respectively, mw5 is the molecular weight of styrene, and Ps and PPs are the density of styrene and polystyrene, respectively.
Seed latices were prepared in such a way as to obtain a latex with an essentially monodisperse partiele size. This avoided complications when analyzing the kinetic data, since rate coefficients may very well be partiele size dependene 2 ·13 • Another advantage of seeded polymerization is that the approach to steady state can be monitored.
In ab initio polymerizations Interval II begins with a
steady state concentration of free radicals.
This is in contrast to seeded
polymerizations, where the experiment begins in a non-steady state with regard to the free radical concentration.
As has been seen in Chapter 3, this approach to
steady state provides crudal information in rate and mechanistic analyses.
66
Chapter 4
Seeded experiments were carried out in two different ways.
Most of the
experiments were carried out with a chemica) initiator, inisurf in our case.
In a
number of experiments the initiation was by means of a y-ray souree (see Chapter 7).
Since the recipes for the latter polymerizations were different, i.e.,
depended on the monitoring technique that was applied, the various recipes and techniques of carrying out these polymerizations will be discussed in the relevant sections of Chapter 7. For all these seeded polymerizations it was important that the preselected number of polymer particles per unit volume of aqueous phase, Ne, did not change during the polymerization. This meant that no new partiele nucleation ("secondary nucleation") should occur and that no coagulation of seed particles should take place. This was checked by comparing the partiele size distribution as obtained via electron microscopy (see Section 4.5.3) before and after the polymerization.
4.3 Seed Preparation As has been mentioned in the previous section a seed latex must be monodisperse, but it should also be stabie against secondary nucleation and coagulation. Various recipes for different monomers have been reportedl2,14,l5,16,17.tB,t9. One condition to favour monodispersity is to carry out the polymerization at higher temperatures (353 K or 363 K). This leads to a higher dissociation rate of the initiator, and, thus, to a higher concentration of free radicals and a shorter nucleation period. Therefore, the final latex, in which most polymer formation has occurred in the absence of new partiele formation, will be monodisperse. Besides that the seed applied in these studies had to be monodispers it had to be easily cleaned from all surfactant, oligomers and residues of initiator; and should still be stabie enough to be swollen with monomer and not show secondary nucleation or coagulation during polymerization.
The latex had to be totally
Experimental Procedures
67
cleaned from any physically adsorbed material at the surface of the particles, since these materials can interfere with the adsorption equilibrium of the inisurfs used. Colloidal stability, monodispersity and cleanability were achieved by carrying out a surfactant-free emulsion polymerization20•21 •22 using a small amount of an ionic co-monomer (sodium p-styrenesulphonate) 23•24 and an ionic initiator (sodium peroxodisulphatei5•26•
The surfactant-like ionic components,
formed during the polymerization, are either chemically bonded to the polymer or consist of oligomers physically adsorbed onto the latex particles. these components add to the electrostalie stabilization.
In both cases
Applying an ionic co-
monomer increased the electric charge at the surface and, thus, contributed to the electrostatic stabilization. Table 4.1 shows the recipe used for the production of the seed latex. These have been carried out either in a bottie polymerizer or in a stainless steel tank reactor with a volume of 5 dm3 (the recipe was proportionally increased).
Table 4.1. Basic Recipe fortheSeed Latex. Amount (gram) Styrene
29.30
Water
150.00
Sodium p-Styrenesulphonate
1.42
Sodium Peroxodisulphate
0.39
Stirring rate: Temperature: Reaction time:
40 rpm for bottie polymerizer (end-over-end) and 300 rpm fortank reactor. 358 K. 24 hours.
The cleaning of the latex can be done by dialysis27 , ion exchange28.z9 , ultracentrifugation30•31 andlor serum replacemene2•33 • but faits to remove oligomers. cleaning a latex.
Dialysis removes ions,
Nevertheless, it is a very useful first step in
Although similar in scope to dialysis, ion exchange bas an
additional disadvantage: the resin granules require extensive precleaning to ensure that all extraetabie contaminants are removed, which makes this a very tedious
method.
The centrifugation process, as well as the serum replacement process
remove both oligomers and ionic species. In our studies, dialysis was used as a first step in cleaning the seed latex. As a second step either ion exchange or ultracentrifugation was used.
Ion
2
exchange was applied following the procedure of Vanderhoff et al. s, in which a mixed bed of anionic (OH· form) and cationic (W form) ion-exchange resins removed the unreacted materials and reaction side-products from the latex and replaced the sodium ions with protons quantitatively.
Ultracentrifugation was a
faster method compared to ion exchange, and used less material as welL
The
ultracentrifugation was carried out on a Kontron Instruments, eentrikon T-2000, at a speed of 45,000 rpm for 3 h. The precipitate was redispersed by ultrasonification in doubly distilled de-ionized water. The seed latex was centrifuged twice before use.
4.4 Conversion Measurements
For the determination of the conversion of the polymerization as a function of time various methods have been applied.
The three methods used in these
studies will now be discussed. Besides the experimental procedure of each method, the method of calculating the conversion of the polymerization from the raw data will be included.
4.4.1 Gravimetry
The conversion was determined by means of the solid weight of a sample and calculated according to Eq. 4.2:
69
Experimental Procedures
x(t)
DS(t) - DSmin DSmax
(4.2)
DSmin
where x(t) is the fractional conversion at time t, DS(t) is the fraction dry solid content in the sample at time t, and
DSm~n
and DS,_ are the dry solid contents at
the start of the reaction (0% conversion) and at the end of the reaction (100% conversion), respectively. DSmin and DS,_ were calculated from the recipe. The sample was drawn from the reactor directly into a dry, clean aluminium cup, where the reaction was shortstopped with added hydroquinone (jor synthesis, Merck, Darmstadt, FRG). The sample was weighed and dried on a steambath and in a vacuum oven at 313 K, until constant weight was obtained. With this method conversion could be determined within 1% accuracy.
4.4.2 Dilatometry
With dilatometry the conversion was determined by means of the change in volume during the polymerization. The following equation for the calculation of the fractional conversion, x(t), was used:
1tr21!Jt x(t)
(4.3)
0
gm
I dm
dp
where r is the bore radius, Ah is the incremental change in the meniscus level, dm and ~ are the densities of monoroerand polymer, respectively, and g~ is the initial mass of monomer. The movements of the meniscus were followed, measured and recorded by
70
Chapter 4
an automated tracker·: the presence of the meniscus is determined by an optical detector, which was moved by a stepping motor to follow the height of the liquid in the capillary. The number of steps was monitored by a computer, and, tagether with the previously determined step height, db could be calculated and substituted in Eq. 4.3. The resolution obtained with this method depended on the step-heights of the motor and was in the order of microns. The accuracy in the conversion was 0.01%. Filling Procedure
The required amount of seed tagether with the monomer were added into the dilatometer and were stirred slowly overnight, to swell the polymer particles of the seed with the monomer.
A salution of inisurf was made in doubly distilled
de-ionized and degassed water·· and the required arnount was added\ into the dilatometer. The dilatometer was further filled with doubly distilled de-ionized and degassed water up to the required level. The whole dilatometer was placed into an ultrasonic bath for emulsification of the reaction mixture.
The dilatometer was
immersed in a water bath and the tracker was aligned. The emulsion level in the capillary of the dilatometer was sealed with hexane and the tracker was adjusted. Every fifteen to thirty seconds a reading was taken of the level of the meniscus, and the fractional conversion was calculated applying Eq. 4.3.
4.4.3 Densimetry
Densimetry was- used in a relative way for the determination of the conversion of the polymerization.
The polymerization was foliowed not only
densimetrically, but also gravimetrically (small number of samples).
The
Apparatus designed by Mr. D.F. Sangster and Mr. G. Baxter of CSIRO, Di vision of Chemieals and Polymers, Lucas Heîghts Research Laboratories, Menai, NSW 2234, Australia. Degassed water was needed sincc a gas bubble formed in the dilalometer could change thc level in the capîllary and thus innuencc the accuracy of the measurements.
71
Experimental Procedures
gravimetrie data were used for calibration of the densimetric data. Conversion was calculated from the gravimetrie data and plotted against time-corresponding density. The resulting curve enabled the translation of density into conversion.
Fig. 4.4
shows a typical density versus conversion plot of a seeded styrene emulsion polymerization, which was used in converting density into conversion. Filling Procedure
Similar to dilatometry, the required amounts of seed latex and monomer were first charged into the reactor and stirred slowly, for swelling. A solution of the required amount of inisurf in doubly distilled de-ionized and argon-flusbed water was then added, foliowed by the remainder of the water.
0.985
Figure 4.4
0.987
0.989
0.991
0.993
0.995
A Typical Plot of Density versus Conversion of a Seeded Styrene Emulsion Polymerization.
4.5 Partiele Size Measurements
For the determination of the various kinetic parameters it is necessary to know the partiele diameter in order to calculate the number of particles per unit volume of aqueous phase, Ne, and the polymerization rate per partiele (or the average number of radicals per particle). Different methods have been used to measure the partiele diameter and the partiele size distribution. These methods will now be discussed below.
72
Chapter4
4.5.1 Dynamic Light Scattering
Dynamic light scattering is a relatively rapid metbod for determining partiele sizes. The dynamic light scattering technique is based on the scattering of a beam of coherent laser light by a number of particles present in a diluted and filtered latex sample. The intensity of the scattered light beam is measured at a certain angle to the primary beam as a function of time at room temperature. The fluctuation of the intensity with time is directly related to the (Brownian) motion of the particles in the dispersion. This diffusion is determined by the size and sh:.:pe of the particles. For spherical particles the Stokes-Einstein relation can be applied for the relation between the diffusion coefficient and the radius of the particles. In this study a Malvem Autosizer Ilc was used, with a 5 mW He-Ne laser, which produced a coherent light of 633 nm wave length. A pboton multiplicating detector is placed at an angle of 90°. 2,000 nm can be analysed.
Particles with a size ranging from 20 to
A so called z-average diameter is measured, from
which a weight-average diameter is calculated.
4.5.2 Transmission Electron Microscopy
Another technique used to obtain an average partiele size of the sample, was Iransmission Qlectron Microscopy (TEM).
With this technique it was also
possible to obtain a total partiele size distribution.
Diluted samples (0.05 wt.%
solids) were dried on 400 Mesh Formvar covered grids, and afterwards covered with a carbon film to enhance stability and conductivity in the electron microscope (Jeol 2000 FX). In the case of a polystyrene latex the polymer was stabie enough in the electron beam and provided enough contrast for taking micrographs. Typically 750- 1000 particles were counted with a Zeiss TGA-10 partiele analyzer. It was found that counting extra particles did not change the partiele size
distribution any more.
Experimental Procedures
73
4.5.3 Disk Centrifuge Photosedimentometry
Another metbod to obtain partiele size distributions is Disk Çentrifuge fhotosedimentometry (DCP). This rapid metbod is used for measuring partiele size distributions of colloidal systems, e.g., pigments and synthetic latices, of partiele sizes ranging from 50 nm to 2 J..Ull. DCP is based on the fact that particles witb different density and diameter migrate at a different speed througb a fluid in a force field, as described by Stokes law 34 • The field applied in tbis technique is a centrifugal force field within a bollow rotating disk. Presuming tbat tbe particles to be analyzed all have tbe same density, Stokes law can be written as:
18 Tl ln(R4 /Rm)
r.
Ap
ar
(4.4)
where dw is the partiele diameter (nm), Tl is tbe viscosity of the spinfluid (kg m· 1 s- 1), R4 is the radius at wbich tbe detector is positioned (m), Rm is tbe radius at whicb the meniscus of the spinfluid (suspension medium through wbich the particles migrate) is and at wbicb the sample starts to migrate (m), t, is the time needed for sedimentation of a partiele from Rm to R4 (s), Ap is the difference in partiele density and spinfluid density (kg m·3), and ro is tbe rotational speed of tbe disk (s- 1).
The tbeory of sedimentation and detection are well described in
literature35 •36•37 •38 •
DCP eliminales problems encountered witb microscopy of
deformable partieles, the need for stains or gold coating on the particles, and the requirement for image analysis of a large number of particles for accurate distribution averages. It bas been found that tbe partiele size distributions obtained witb DCP are in good agreement with those obtained with TEM39•40•41 • Two different procedures are used for obtaining tbe partiele size distribution: 1) the Jine-start metbod (LIST)42.43, wbere a small sample is injected into tbe spinning disk, whicb already contains tbe spinfluid; 2) the homogeneous-start metbod (HOST)44 '45.46, where the spinfluid initially contains a uniform concentration of the colloid, wbose partiele size distribution is to be determined.
74
Chopter 4
The LIST-metbod is to be preferred, since it is able to obtain results with a higher precision and a better resolution. The main advantage of the HOST-metbod is its applicability to colloids with partiele densities less than that of the spinfluid47 • To obtain even better results, a modified LIST-metbod has been applied in this investigation. In this so-called buffered line-start method48 •49 •50, 1 mL of methanol is added to 15 mL of doubly distilled de-ionized water already present in the cavity of the spinning disk.
The density gradient formed in this way at the
inner boundary eliminates to a great extent the density discontinuity and interfacial tension between the suspension medium and spinfluid. Hence, stabie sedimentation and, thus, satisfactory separation was obtained. The latex sample is prepared by actding 10-15 drops of polystyrene latex to 15 mL of doubly distilled de-ionized water foliowed by actdition of 5 mL of methanol. The solid weight content in the samples is estimated to be 0.025-0.05%. A small amount of this sample (0.25 mL) is injected into the cavity of the spinning disk, where the water and methanol is already present, as described above. The used instrument was a Brookhaven BI DCP Partiele Sizer.
4.5.4 Partiele Number Concentration
After the partiele size bas been determined, the partiele number per unit volume of aqueous phase, N,, can be calculated as follows:
6xmsoPw 1td3
(4.5)
mw p PS
where x is the fractional conversion, d is the partiele diameter, mw is the amount of water, and Pw is the density of water.
Experimental Procedures
75
4.6 Surface Tension Measurements To determine the Çritical Micelle ,Ç_oncentration (CMC) of the synthesized inisurfs the surface tension has been measured.
This bas been done by two
methods, and both will be described bere. Surface tension measurements have also been used to determine the adsorption isotherm of inisurf on polystyrene latex particles.
4.6.1 Du Nouy Ring-Method51 This metbod is based on the measurement of the force required to detach a frame, usually in the form of a ring, from the surface of a liquid or solution. The commercially available equipment consists of a torsion balance and a platinum ring.
This metbod is very quick and simple, does not require large volumes of
liquid or solution, and is very suitable for comparative measurements. For absolute measurements of surface tension, it is essential to apply correction factors, as described by Harkins et al. 52 , and by Fox et a/. 53 • 1t is essential to use a clean platinum ring (cleaned by flaming) and to
ensure that the surface of the liquid or solution is clean as well. The ring should be free of kinks and as flat as possible.
4.6.2 Maximum Pressure Bubble Tensiometry The determination of surface tension by measuring the pressure required to liberate bubbles from a vertical capillary tube immersed in a liquid was first suggested by Simon54 • This is a quick method, requiring simple apparatus, and it is very accurate. Since the surface is renewed quite quickly (almost every second), surface contamination is minimized. Contact-angle problems as with the capillaryheight method 55 , do not usually arise, although due care must be taken to ensure
76
Chapter 4
that the bubble forms on either the inside or outside diameter of the capillary. 1t must also be ensured that the diameter of the capillary does not change during the measurement, due to contamination on the inside of the capillary. Some modifications for an even simpler application have been described by Sugden56•57 •
The application of two capillaries with different radii has been
suggested, with both tubes vertically immersed to the same depth in the liquid or solution. The surface tension, y, can then be calculated from the difference of the pressures, required to liberate bubbles from each capillary, using the following simple empirica! equation: y
=
(4.6)
l flp
where flp is the difference in the pressures needed for releasing bubbles from the narrow and wide capillary, and l is an apparatus constant, found through calibration. The radius of the larger capillary is usually in between 1 and 4 mm, and that of the smaller one below 0.5 mm.
The rate of bubble release is
approximately one per second for the small capillary and slightly lower for the large capillary. The method gives an accuracy of ca. 0.5% and can also be used for liquid-liquid interfaces58 , on remote control, and in continuous flow cells to obtain a surface tension reading as a function of time.
Accuracy can even be
increased by re-orienting the capillaries so that the eentres of the formed bubbles, rather than the tips of the capillaries, are level.
A SensaDyne 6000 Surface
Tensiometer was used in our investigations.
4.7 Initiator Decomposition Measurements
The dissociation rates of the formed initiator and inisurfs were measure different temperatures by means of UV-analysis applying a Hewlett Packard Diode Array Spectrophotometer.
The disappearance of the clearly-m
absorption-band of the azo-group (À = 350 nm) was monitored as a fu•
r
d.
able .m of
Experimental Procedures
time and temperature.
77
Since the rate of dissociation of azo-compounds is not
influenced by concentration and other substances dissolved and hardly changes with solvent59'60 , the dissociation rate of the initiator and the inisurfs was only determined in water at one concentration. The solutions of the initiator or inisurfs were kept in a thermostated bath, and samples were withdrawn at constant time intervals. The UV spectrum of each sample was obtained and the decrea'ie in the absorption of the azo-group, at 350 nm, was followed.
Keferences 1. 2. 3. 4. 5.
6. 7. 8. 9. 10. ll. 12. 13. 14. 15. 16. 17.
Starkweather, H.W.; Taylor, G.B. J. Am. Chem. Soc. 1930, 52, 4708. Fryling, C.F. Ind. Eng. Chem., Anal. Ed. 1944, 16, 1. Pochlein, G.W.; Dougherty, DJ. Rubber Chem. Techno/. 1977, 50, 601. Pitch, R.M.; Shih, L.-B. Progr. Colloid Polym. Sci. 1975, 56, 1. Gutta, G.; Benetta, G.; Talamini, G.P.; Vianello, G. Adv. Chem. Ser. 1969, 91, 158. Vanderhoff, J.W.; Vitkuske, J.F.; Bradford, E.B.; Alfrey, T., Jr. J. Colloid Sci. 1956, 11' 135. Vanderhoff, J.W.; Bradford, E.B.; Tarkowski, H.L.; Wilkinson, B.W. J. Polym. Sci. 1961, 25, 265. Poehlein, G.W.; Vanderhoff, J.W. J. Polym. Sci., Polym. Chem. Ed., Part A-1 1973, 11, 447. Hawkett, B.S.; Napper, D.H.; Gilbert, R.G J. Chem. Soc., Faraday Trans. I 1975, 71' 2288. Gilbert, R.G.; Napper, D.H. J. Chem. Soc., Faraday Trans. I 1974, 70, 391. Gilbert, R.G.; Napper, D.H. J. Macromol. Sci., Rev. Macromol. Chem. Phys. 1983, C23, 127. Hawkett, B.S.; Napper, D.H.; Gilbert, R.G. J. Chem. Soc., Faraday Trans. I, 1980, 76, 1323. Morrison, B.R.; Maxwell, I.A.; Gilbert, R.G.; Napper, D.H. ACS Symp. Ser. 1992, 492, 28. Smith, W.V. J. Am. Chem. Soc. 1948, 70, 3695. Verdurmen, E.M.; Albers, J.G.; Dohmen, E.H.; Zirkzee, H.F.; Maxwell, LA.; German, A.L. to be published. Hjertberg, T.; Sörvik, E.-M. J. Polym. Sci., Polym. Chem. Ed., Part A-1 1978, 16, 645. Lee, C.K.; Forsyth, T.H. ACS Symp. Ser. 1981, 165, 567.
78
Chapter 4
18. Nak:amuro, Y.; Tabata, H.; Suzuki, H.; Iko, K.; Okubo, M.; Matsumoto, T. J. Appl. Polym. Sci. 1986, 32, 4865. 19. Lange, D.M.; Poehlein, G.W.; Hayashi, S.; Komatsu, A.; Hirai, T. J. Polym. Sci., Polym. Chem. Ed., Part A-I 1991, 29, 785. 20. Willes, J.M. Ind. Eng. Chem. 1949, 4I, 2272. 21. Kotera, K.; Furusawa, K.; Tak:eda, Y. Kolloid Z. Z. Polym. 1970, 239, 677. 22. Goodwin, J.W.; Hearn, J.; Ho, C.C.; Ottewill, R.H. Brit. Polym. J. 1973, 5, 347. 23. Liu, L.; Krieger, I.M. In Emulsion, Latices and Dispersions; Becher, P., Ed.; Marcel Dekker Inc.: New York, 1978; p 26. 24. Juang, M.S.; Krieger, I.M. J. Polym. Sci., Polym. Chem. Ed., Part A-I 1976, I4, 2089. 25. Hearn, J.; Wilkinson, M.C.; Goodall, AR.; Chainey, M. J. Polym. Sci., Poly:n. Chem. Ed., Part A-I 1985, 23, 1869. 26. Fitch, R.M.; Tsai, C.H. In Polymer Colloids; Fitch, R.M., Ed.; Plenum Press: New York, 1971; p 103. 27. Ottewill, R.H.; Shaw, J.N. Kolloid Z. Z. Polym. 1970, 2I8, 34. 28. Van den Hul, H.J.; Vanderhoff, J.W. J. Colloid Interface Sci. 1968, 28, 336. 29. Vanderhoff, J.W.; Van den Huil, H.J.; Tausk, R.J.; Overbeek, J.T. In Clean Surfaces: Their Preparation and Characterization for lnterfacial Studies; Goldfinger, G., Ed.; Marcel Dekker: New York, 1970; p 15. 30. Voegtli, L.P.; Zukoski C.F. J. Colloid Interface Sci. 1991, I4I, 92. 31. Chonde, Y.; Krieger, I.M. J. Colloid Interface Sci. 1980, 77, 138. 32. Ahmed, S.M. Ph.D. Thesis, Lehigh University, Bethlehem, PE, USA 1979. 33. Ahmed, S.M.; El-Aasser, M.S.; Pauli, G.H.; Poehlein, G.W.; Vanderhoff J.W. J. Colloid Interface Sci. 1980, 73, 388. 34. Koglin, B.; Leschonski, K.; Alex, W. Chem. Ing. Techn. 1974, 46, 563. 35. Langer, G. Colloid Polym. Sci. 1979, 257, 522. 36. Oppenheimer, L.E. J. Colloid Interface Sci. 1983, 92, 350. 37. Coll, H.; Haseler, S.C. J. Colloid Interface Sci. 1984, 99, 591. 38. Coll, H.; Searles, C.G. J. Colloid Interface Sci. 1987, 115, 121. 39. Zirkzee, H.F.; Maxwell, I.A.; German, AL. to be published. 40. Verdurmen, E.M.; Albers, J.G.; German, A.L. Colloid Polym. Sci. accepted for publication. 41. Koehler, M.E.; Provder, T. ACS Symp. Ser. 1987, 332, 231. 42. Coll, H.; Oppenheimer, L.E.; Searles, C.G. J. Colloid Interface Sci. 1985, 104, 193. 43. Coll, H.; Searles, C.G. J. Colloid Interface Sci. 1986, I JO, 65. 44. Lloyd, P.J.; Scarlet, B.; Sinclair, I. In Partiele Size Analysis I970; Groves, M.l., Wyatt, J.L., Eds.; Wiley: New York, 1970; p 60. 45. Groves, M.J.; Yalabik, H.S. Powder Techno!. 1977, 17, 213. 46. Svarovsky, L.; Svarovska, J. J. Phys., Appl. Phys., Sec. D 1975, 8, 181. 47. Groves, M.J.; Yalabik, H.S. Powder Techno!. 1975, I2, 233. 48. Jones, M.H. Proc. Soc. Anal. Chem. 1966, 3, 116. 49. Beresford, J. Oil Col. Chem. Assoc. 1967, 50, 594.
Experimental Procedures
79
50. Faser, 1.; Patton, T.C. In Pigment Handbook; Patton, T.C., Ed.; Wiley: New York, 1973; Vol. 3, p 327. 51. DuNouy 1. Gen. Physiol. 1918, 1, 521. 52. Harkins, W.D.; Jordan, H.F. 1. Am. Chem. Soc. 1930, 52, 1751. 53. Fox, H.W.; Chrisman, C.H. 1. Phys. Chem. 1952, 56, 284. 54. Sirnon P. Ann. Chim. Phys. 1851, 32, 5. 55. Adamson, A.W. Physical Chemistry of Surfaces; Wiley: New York, 1982; p 4. 56. Sugden, S. 1. Chem. Soc. 1922, 121, 858. 57. Sugden, S. 1. Chem. Soc. 1924, 124, 27. 58. Hutchinson, E. Trans. Faraday Soc. 1943, 39, 229. 59. Overberger, C.G.; Hale, W.F.; Berenbaum, M.B.; Finestone, A.B. 1. Am. Chem. Soc. 1954, 76, 6185. 60. Sheppard, C.S. In Encyclopedia of Polymer Science and Engineering; Mark, H.F., Bikales, N.M., Overberger, C.G., Menges, G., Eds.; Wiley: New York, 2. Ed., 1985; Vol. 2, p 143.
80
Chapter 5
The Application of Inisurfs A Survey of the Performance of the Various Inisurfs in the ab Initio Emulsion Polymerization of Styrene
Summary: The results obtained with different inisurfs in ab initia emulsion polymerization experiments of styrene will be presented. ft will be shown that with an increase in the length of the poly( ethylene oxide) part in the surfactant moiety, the colloidal stability during emulsion polymerization increases. Furthermore, systems with an anionic co-surfactant and a nonionic inisurf show similar behaviour as systems with anionic and nonionic surfactant mixtures.
5.1 Introduetion
In this chapter the first results of the application of inisurfs in the emulsion polymerization of styrene wil! be discussed. A criterion for a "good" inisurf is that it should be able to perfarm as an emulsifier and as an initiator in an emulsion polymerization in the absence of any other emulsifying and/or initiating substance. Only ab initia experiments (see Section 4.2.2) will be of concern in this part of the investigation. This chapter has been divided into two major sections. The results obtained with symmetrical inisurfs (7a) - (7d) are reported in Section 5.2 and those obtained
82
with asymmetrical inisurfs (7e) - (7g) are reported in Section 5.3. For the ab initio reactions a standard recipe with styrene as monomer has been used which is given in Table 5.1. The concentration of inisurf was based on the volume of the aqueous phase. The monomer-water ratio was 0.1 weight by weight.
The reactions were carried out in a stainless steel batch reactor with a
volume of l dm3 (Section 4.2.l.l) at 323 K, with an impeller speed of 300 rpm and monitored by gravimetry, unless stated otherwise.
This temperature was
chosen because a considerable amount of data is available for the emulsion polymerization of styrene at 323 K, which can be used for comparison.
The
applied impeller speed was similar to the one used for systems where surfactant is present. The type and amount of co-surfactant wil! be mentioned in the relevant sec ti ons.
Table 5.1 Standard Recipe for ab lnitio Reactions.
Component
Amount
Water
900 g
Monomer
90 g
6.3 1o-3 mol dm-3
Inisurf Co-surfactant
variabie
As was already mentioned in Section 3.2, primary free radicals originating from the same initiator molecule can combine, before they diffuse away from the vicinity of forrnation, and form an unreactive compound. this so-called cage-effect
12 •
As can be envisioned,
wil! be more pronounced in the case of large free
radicals. In the case of the inisurfs it can be anticipated that with an increase in the length of the surfactant moiety, the probability of geminate recombination increases as well. Moreover, it is expected that symmetrical inisurfs show more geminate recombination than asymmetrical ones, since a large surface-active free radkal diffuses more slowly than the much smaller t-butyl free radicals.
This
The Application of lnisurf
83
effect will be discussed to some extent in Section 5.4.
5.2 Symmetrical Inisurfs
The results of ab initia reactions with the symmetrical inisurfs (7a)-(7d) will be discussed in order to elucidate the effect of increasing length of the poly(ethylene oxide) chain in the surfactant moieties of these inisurfs.
5.2.1 Ab lnito Reaelions with Inisurfs SE0-350 (7a) and SE0-550 (7b)
With SE0-350 and with SE0-550 it was not possible to start an emulsion polymerization with inisurf as the only stahilizing agent in the system.
We
reasoned that, since these inisurfs probably were hardly surface-active (no welldefined hydrophobic and hydrophilic moieties in these molecules), the formed particles could not be colloidally stabilized. The envisioned in situ formation of the surfactant was probably too slow to form sufficient surface-active material for stabilization and subsequent normal emulsion polymerization. Therefore with these inisurfs, Antarox C0-630 was used as a co-surfactant in order to have only steric stabilization in these systems. The ab initio reactions of SE0-350 and SE0-550 were carried out with 5.56 g and 7.82 g of inisurf, respectively, and with 16.63 g of Antarox C0-630 (dried as described in Chapter 2). The conversion-time curves of these experiments are shown in Fig. 5.1.
As can be seen from these curves, the reaelions with SE0-350 and SE0-550 proceeded at a very slow rate (approximately 1 10·5 and 7
w-6
mol dm- 3 s- 1,
respectively, between 20% and 40% conversion) and stopped altogether at about 70% conversion due to massive coagulation. No region with a constant reaction rate was observed.
84
Chapter5
From these experiments it could be concluded that inisurfs SE0-350 and SE0-550 were not suited for ab initio emulsion polymerizations.
100,_----------------------------------~
80
-
t (min)
Figure 5.1
Conversion-Time Curves for ab Initio Reactions with Symmetrical Inisuif SEO-#, with e: SE0-350; ~: SE0-550.
5.2.2 Ab lnitio Reaelions with luisurf SC0-630 (7c)
Applying the standard recipe, an ab initio reaction with SC0-630 was carried out using 8.53 g of this inisurf. Although a co-surfactant was not necessary to start the emulsion polymerization, the reaction already stopped at a conversion of approximately 40% (see Fig. 5.2) due to coagulation. Stabilization problems, originating from the binding of the surfactant moiety to the initiator moiety, could have been a reason for the cessation of the polymerization.
Alternatively, poor initiator performance resulting from the
altachment of the initiator moiety to the surfactant moiety could also have been a reason.
The Application of lnisurf
85
80
~160
40
20
60
120
180
240
300
360
420
480
t (min)
Figure 5.2
Conversion-Time Curve for Symmetrical lnisurf SC0-630.
ab
lnitio
Reaction
with
As a check on the stahilizing ability of the inisurf SC0-630, several experiments were performed in the presence of an anionic co-surfactant (§,odium f!.odecyl§.ulphate, SDS, for synthesis, Merk, Darmstadt, FRG). The influence of the SDS concentration on polymerization with SC0-630 has been presented in Table 5.2. The conversion-time curves are plotted in Fig. 5.3.
Table 5.2
Inftuence of the Co-Surfactant Concentration on the Polymerization with SC0-630. R!_l
.
[Co-surfactant] (10' 3 mol dm- 3)
Ne (10 17 dm- 3)
(10' 4 mol dm· 3 s· 1)
(nm)
1.88
1.36
0.78
107.6
7.51
2.85
1.65
84.7
30.0
4.12
2.36
75.2
pol
dw
* Determined by light scattering as described in Section 4.5.3.
As would be expected, the increase in surfactant concentration resulted in an increase in partiele number, which was seen as an increase in reaction rate in Interval II (see Figure 5.3), and a drop in partiele diameter (see Table 5.2). From these results it can be concluded that the stahilizing ability of the surface-active
86
Chapter 5
moiety in the inisurf SC0-630 alone is insufficient to stabilize the formed polymer partides.
..-·
~·,________,_··_,.·~ ~~ i ......!.....
60
Figure 5.3
120
180
240
300
360
Conversion-Time Curves for ah Initio Reactions with SC0630 and Different Co-Suifactant Concentrations, with •: 1.88 mol dm-1,- A: 7.51 mol dm'1 ; • : 30.0 mol dm·1.
As a check on the initiating moiety an emulsion polymerization was carried out with the initiator and surfactant used in the synthesis of SC0-630.
Thus, a
"traditional" emulsion polymerization was carried out, where initiation and stabilization were performed by two different substances. ACPA (la) was used as initiator and Antarox C0-630 (2c) as surfactant. An otherwise identical recipe and identical conditions as reported in Table 5.1 were used (Table 5.3). Figure 5.4 shows the conversion-time curve of this "control" reaction as well as the conversion-time curve of the reaction with inisurf and co-surfactant. As can be seen, the polymerization rates in Interval 11 per unit volume of aqueous phase were approximately the same. The partiele diameters were determined at the end of the reaction using dynamic light scattering.
These partiele diameters, as
well as the calculated partiele numbers per unit volume of aqueous phase, the rates per partiele Table 5.4.
(R~)
and the average number of radicals per partiele (fl) are given in
The Application of lnisurf
Table 5.3
87
Recipe for Reaction with Initiator and Surfactant. Amount (Concentration)
Component Water
900 g
Monomer
90 g
Initiator (ACPA)
1.59 g (6.3 10"3 mol dm"3)
Surfactant (Antarox C0-630)
6.99 g (12.6 10·3 mol dm-3)
Co-Surfactant (SDS)
7.88 g (30.0 10"3 mol dm"3)
loo r------···········~······.··-__--:w:~ ... ~ ... ~..~ ...-=,7'==-=---=.,:.:..::_..;: __ :::.:: .. ~.;:::=~•.-l ,
..~..······
60
40 20
60
0
120
180
240
1 (min)
Figure 5.4
Conversion-Time Curves for ah lnitio Reactions with lnisurf SC0-630 ( "') and with Initiator and Surfactants (e).
Since the polymerization rates were comparable and the initiation of polymerization with SC0-630 was entirely due to the inisurf it can be concluded from the above results that the performance of the initiating moiety in the inisurf was not significantly influenced by the link of the initiator moiety to the surfactant moiety.
Only the surfactant moiety of the inisurf should be changed in order to
improve the performance.
88
Chapter 5
Table 5.4
The Effect of Linking Initiator and Suifactant on the Emu/sion Polymerization of Styrene in the Presence of a Co-Suifactant.
Ne (10 17 dm-3)
RI/pol (104 mol dm·3 s· 1)
(nm)
(10"22 mol s" 1)
SC0-630
4.12
2.36
75.2
5.73
0.23
ACPA
6.41
2.81
65.1
6.41
0.26
Initiator
* n is calculated kP
by appling Eq. 3.7, were CM
= 258 dm' mol·' s·'
dw
R~ol
= 5.8 mol dm·' and
5.2.3 Ab Initio Reaction with SC0-880 (7d)
For the ab initio reaction with SC0-880 the standard recipe was applied with 18.89 g of this inisurf. Figure 5.5 shows the resultant conversion-time curve. As was the case with SC0-630 (7c), no co-surfactant was necessary to start an emulsion polymerization, but in this case coagulation did not occur in the course of the reaction.
lt can be concluded that this inisurf, SC0-880, satisfied the
requirements, as set out in Chapter 2.
Emulsion polymerizations with SC0-880
will be discussed in more detail in Chapter 6.
5.3 Asymmetrical Inisurfs
In this section the results obtained with the asymmetrical inisurfs (7e) - (7g) will be discussed. As was the case with the symmetrical inisurfs, the results will be presented in the sequence of the increasing length of the poly(ethylene oxide) chain in the surfactant moiety of the inisurfs.
The Application of lnisurf
89
100 80
~1
60 40 20 0 0
Figure 5.5
120
240
360
480
600
Conversion-Time Curve for the ab Initio Reaction with Symmetrical Inisurf SC0-880.
5.3.1 Ab lnitio Reactions with Inisurf AE0-550 (7e)
The reaction with AE0-550 was carried out according to the recipe in Table 5.1 with 4.22 g of this inisurf. As with the symmetrical inisurf SE0-550, a cosurfactant (Antarox C0-630,
16.63 g) was needed to start an emulsion
polymerization. The resultant conversion-time curve is shown in Fig. 5.6. As was the case with the symmetrical inisurf SE0-550, an emulsion polymerization commenced, but after 20% conversion the reaction stopped, probably due to probieros with the colloidal stability of the particles.
5.3.2 Ab lnitio Reactions with Inisurf AC0-630 (70
The batch reaction with AC0-630 was carried out using the standard recipe with 4.68 g of this inisurf.
The resultant conversion-time curve is shown in
Fig. 5.7. The emulsion polymerization started normally, but after 20% conversion massive coagulation occurred and the reaction stopped.
90
Chapter 5
100,----------------------------------,
80
60
tf
~
1
40
20
..................•.~
0
200
400
600
800
1000
1200
....
1400
-
... 1600
t(min)
Figure 5.6
Conversion-Time Curve for the ab Initia Reaction with Asymmetrical Inisuif AE0-550.
80
tf
~
60
1
40
20 o~~~=d~~~~~~~~~~~~~~
0
250
500
750
1000
1250
1500
1750
t (min)
Figure 5.7
Conversion-Time Curve for the ab Initia Reaction with Asymmetrical Inisuif AC0-630.
Table 5.5
Influence of the Co-Suifactant Concentration on the Polymerization with AC0-630.
R~ol
[Co-surfactant]
Ne
( 10" 3 mol dm" 3)
(10 16 dm. 3)
(10" 5 mol dm" 3 s" 1)
(nm)
1.88
0.91
0.67
262.6
7.51
3.07
2.40
175.1
30.0
3.82
3.00
162.7
dw
The Application of Inisurf
91
With this inisurf reactions have also been carried out in the presence of an anionic co-surfactant (SDS). The influence of different co-surfactant concentrations on the polymerization is shown in Table 5.5. The resultant conversion-time curves are shown in Fig. 5.8. As was observed m the case of the symmetrical inisurf SC0-630, an increase in the co-surfactant concentration resulted in an increase in the partiele number.
Hence, this led to an increase of the overall reaction rate and to a
decrease in the mean partiele diameter, as shown in Table 5.5.
100~--------------------------------------,
80
i?: ';'
60
1
40
20
0
100
200
300
400
500
600
700
800
t (min)
Figure 5.8
Conversion-Time Curves for ab Initia Reactions with AC0630 and Different Co-Surfactant Concentrations, with e: 1.88 mol dm- 3; + : 7.51 mol dm- 3 ; .11: 30.0 mol dm- 3•
It can be coneluded that upon addition of co-surfactant normal emulsion
polymerization behaviour was observed.
Thus, the steric stahilizing ability of
inisurf AC0-630 was insufficient and, if an emulsion polymerization has to be carried out where initiation and stabilization stem solely from inisurf, the stahilizing ability of the surfactant moiety of the inisurf has to be improved. It can also be concluded that the initiating ability has not been impaired significantly by attaching the surfactant moiety to the initiator moiety.
92
Chapter 5
5.3.3 Ab Initio Reaction with luisurf AC0-880 (7g)
For the ab initio reaction with AC0-880 the standard recipe was applied with 9.86 g of this inisurf. Fig. 5.9 shows the resultant conversion-time curve. As was the case with SC0-880 (7d), a co-surfactant was not necessary to perform an emulsion polymerization and coagulation did not take place during the reaction. It can be concluded that this inisurf satisfied the requirements, as set in Chapter 2. Further results of emulsion polymerizations with AC0-880 will be presented in Chapter 6.
100 80
~~
60 40 20
o' 0
200
400
600
,.
800
1000
t (min)
Figure 5.9
Conversion-Time Curve for the ab lnitio Reaction with Asymmetrieal/nisurf AC0-880.
5.4 Discussion
In the series, where the co-surfactant concentration was varied, the results obtained with the symmetrical and the asymmetrical inisurf show no significant differant behaviour.
Fig. 5.10 shows the size of the particles at the end of the
reaction versus the ratio [anionic co-surfactant]/[nonionic inisurf], defined as v. The partiele sizes obtained with asymmetrical inisurf are larger than those obtained with symmetrical inisurf.
Since it is presumed that partiele formation
occurred mainly via the micellar nucleation mechanism, the larger partiele nurnber
The Application of Inisurf
93
for symmetrical inisurfs can be explained as follows. The symmetrical inisurf has two surfactant moieties whereas the asymmetrical inisurf has only one; hence, the concentration of nonionic surfactant for the symmetrical inisurf was twice as large. Thus in the case of symmetrical inisurfs, more micelles and consequently, more particles were formed. 300
,--r----------------
········ ....•.. . .................. .
V
Figure 5.10
Partiele Size (dw) versus the Ratio (v) of Anionic CoSurfactant to Nonionic lnisurf, with +: SC0-630; ,.,,. AC0630.
Sirnilar to the results obtained by Woods et a/. 3 , by Chu et al.\ and by Kato et a/. 5 , it was found that the size of the polystyrene particles decreased with an increasing amount of the anionic component (co-surfactant) in the surfactant mixture. It is known that an increasing amount of anionic surfactant in a mixture of surfactants leads to a decreasing size of the mixed micelles6• The decreasing size of the polymer particles with increasing SDS concentration was not only the result of the amount of surfactant itself, but also of the smaller micelle size3•7 •
In the emulsion polymerization of styrene, according to the Smith-Ewart theory, a logarithmic plot of the number of particles formed and, therefore, the rate of polymerization,
Rpot•
versus the total surfactant concentration, [S], 01 , should
generate a straight line (substitution of Eq. 3.20 into Eq. 3.18).
However,
experimental results of the present study did not give a straight line, as is
94
Chapter 5
illustrated in Fig. S.ll.
Similar results have been obtained by Kamath8 when
applying mixed emulsifier systems. Other factors were possibly affecting the rate of polymerization. It was assumed7·8 that the size of the micelle was a contributing factor. Additionally, it was shown by Piirma et
ae, and this can also be deduced
9
from results presenled by Medvedev , that the probability of the micelle to become a latex partiele depends on the size of the micelle. This problem lies beyond the scope of the present investigation. However, it bas been shown that systems where nonionic inisurfs are used in combination with anionic co-surfactant, behave similarly as systems with a traditional initiator-emulsifier combination.
10
18 -2
10
10
1~
17
A_::--3
JO
'"' o:-e
."
1016
~v
-4
JO
! r:.:.l
w's -5
JO 1014
5 10"
3
Jo·'
10-2
[SJ
Figure 5.11
rm
(mol
Number of Polymer Particles (N) and Rate of Polymerization (R~01 ) versus the Total Surfactant Concentration ([S],0 ,), with .1,o: SC0-630; -',•: AC0-630.
At last, the results obtained with inisurfs synthesized with Antarox Co (SC0-630, SC0-880, AC0-630, and AC0-880) are discussed in more detail. Fig. 5.12 shows the conversion-time curves for these inisurfs. In the case of the asymmetrical inisurfs an increase in the chain length of
the ethylene oxide chain (increasing HLB) resulted in a higher reaction rate 10 and in a more stabie latex 11 , as was expected. For the symmetrical inisurf a lower initia! reaction rate bas been observed
The Application of lnisurf
95
for SC0-880 as compared with SC0-630. As can be anticipated intuitively, with an increase in length of the inisurf molecule the influence of a cage-effect would be more pronounced.
Thus, a longer ethylene oxide chain in the symmetrical
inisurf leads to a deercase in the efficiency of initiation. A lower initiation rate results in less particles and a slower initial polymerization rate, and thus masks the effect of the increased stabilization by the longer ethylene oxide chain.
The
constant increase in the reaction rate, as observed in the case with SC0-880, will be discussed in more detail in Chapter 6. This effect of a longer ethylene oxide
chain is apparently very minor or even absent in the case of asymmetrical inisurfs.
80
20
250
500
750
1000
1250
1500
1750
t (min)
Figure 5.12
Conversion-Time Curves for the Varies lnisurfs Synthesized with Ll: SC0-630; o: SC0-880; "-: AC0-630; •: AC0-880.
Keferences
1. 2. 3. 4. 5.
Franck, J.; Rabinowitsch, E. Trans. Faraday Soc. 1934, 30, 120. Rabinowitsch, E.; Wood, W.C. Trans. Faraday Soc. 1936, 32, 1381. Woods, M.; Dodge, J.; Krieger, J.M.; Pierce, R. J. Paint. Tech. 1968, 40, 541. Chu, H.-H.; Piirma, I. Polym. Bull. 1989, 21, 301. Kato, K.; Kondo, H.; Esumi, K.; Meguro, K. Bull. Chem. Soc. Jpn. 1986, 59, 3741.
96
6. 7. 8. 9.
Chapter 5
Kuriyama, K.; Inoue, H.; Nakagawa, T. Kolloid Z Z Polym. 1962, 183, 68. Piirma, 1.; Wang, P.-C. ACS Symp. Ser. 1976, 24, 34. Kamath, V. Ph.D. Thesis, University of Akron, Akron, Ohio, USA, 1973. Medvedev, S.S.; Gritskova, LA.; Zuikov, A.V.; Sedak:ova, L.I.; Berejnoi, G.D. J. Macromol. Sci., Chem. 1973, Al, 715. 10. Chao, T.-C.; Piirma, I. Polym. Prepr., Am. Chem. Soc., Div. Polym. Chem. 1983, 24, 69. 11. Greth, G.G.; Wilson, J.E. 1. Appl. Polym. Sci. 1961, 5, 135.
Chapter 6
Inisurfs with Antarox C0-880 as Surfactant Moiety
Summary: Two inisurfs, one symmetrical and one asymmetrical, both with Antarox Co-880 as surfactani moiety, wil/ be discussed in detail. The dissociation rates and the Critica/ Micelle Concentrations will be given and discussed. The results of ab initio reactions, i.e., reaction rate, partiele size distribution and molecular weight, will be compared with the results obtained for conventional systems with initiator and surfactant similar to the initiator moiety and the surfactant moiety of the inisurfs.
6.1 Introduetion
In this Chapter two of the inisurf systems, described in Chapter 5 and which were found to be suitable in emulsion polymerization, will be discussed in more detail. The symmetrical inisurf SC0-880 and the asymmetrical inisurf AC0-880 have been investigated. First, some physical properties, i.e., the dissociation rate (Section 6.2), and the CMC and surface coverage (Section 6.3), will be discussed. This is foliowed in Section 6.4 by a eloser look at the rate of polymerization, partiele size and partiele size distribution. This chapter ends with a discussion of
98
Chapter 6
the obtained molecular weights (Section 6.5).
6.2 Dissociation Rate of the Inisurfs
The rate coefficient of dissociation of the inisurfs was determined in water at a concentration above the CMC of those inisurfs.
The dissociation was
monitored by following the deercase of the UV absorption at À. = 350 nm, which is caused by the disappearance of the azo-group. As has been reported by numerous researchers, the dissociation of azo compounds is a unimolecular, first-order rate process 1.2.3.4. independent of concentration medium
1 2 3 7 8 9 10 ••• ••• .
3 6
.5·
The rate constant is
and is virtually unaffected by the reaction
However, the dissociation rate is influenced by steric effects7
and by branching of the molecule4• It can be anticipated that the dissociation rate of the inisurfs will differ according to their solubilization state, i.e., whether they are adsorbed at the surface of a partiele or agglomerated in a rnicelle or whether they are molecularly dissolved in the aqueous phase. Since the behaviour of inisurf at the surface of polymer particles and in micelles is of interest, we will only consider
the
dissociation
rate
above
the
CMC.
The
contribution
of
monomolecularly dissolved inisurf to the dissociation rate is assumed to be negligible due to the low value of the CMC. It is known from literature that the absorption originating from to the azo-
group obeys Beer's law 5•11 • concentration.
Thus, absorbtion can be translated directly to
As shown in Figs. 6.1 and 6.2, excellent linear first-order
correlations were obtained by plotting -In (Abs, I Abs 0) versus time (t), where Abs 0 and Abs, denote the absorbtion at zero and at t seconds, respectively. absorbtion at
t
=
The
= is zero (water is used as a reference). The kd values for the
SC0-880 and AC0-880 were determined from the slopes of the plots and are summarized in Table 6.1. Fig 6.3 shows the plot of In kd versus T'.
lnisurfs from Antarox C0-880
99
1.20
'""'
[f ~
0.80
0.40
..!? 6
12
18
t (!0
Figure 6.1
24
3
30
36
s)
Logarithm of the UV Absorbtion at 350 nm versus Time for SC0-880 at Various Temperatures: + = 323 K, o = 343 K, ~ = 353 K.
l.20 .-----,~--rl- - - - - - ,
[
I
'""'c
~-
.
I I I
0.80
~ 0.40
c:
:
l
I
/ I // i'
,'
; I
I00
-
200
300
3
t (10
Figure 6.2
500
s)
Logarithm of the UV Absorbtion at 350 nm versus Time for AC0-880 at Various Temperatures: ~ = 323 K, o 334 K, + = 441 K.
Table 6.1
Dissociation Rate for SC0-880 and AC0-880
Temperature
Eact
400
kd
(K)
SC0-880 (10-6 s· 1)
AC0-880 ( 10-6 s· 1)
323
6.3 ± 0.6
0.93 ± 0.09
334
5.0 ± 0.5
341
11 ± 1
343
12 ± 1
353
30 ± 3
(kJ mof
k0 (s' 1)
1 )
47 ± 5 2.3
IOZ ± 0.2 102
1.3 102 ± 0.1 102 7.6 1014 ± 0.8 10 14
100
Chapter 6
The activation energies (Eac,) and the pre-exponential factor (k0 ) were determined and these are reported in Table 6.1. The values for the dissociation constant at 323 K are of the sarne order of magnitude as those reported for other azo-type initiators 1•2•5•7•8•10• In the case of the asymmetrical inisurf an activation energy and a pre-exponential factor were found simHar to the ones, obtained for other azo-type compounds. For the symmetrical inisurf a substantially lower activation energy and pre-exponential factor were found. The decreases in the activation energy and in the pre-exponential factor were attributed to steric and resonance effects.
For a
symmetrical inisurf at the surface of a particle, this implies that there is an increase in strain within the molecule, when compared to an asymmetrical one.
0.00310
Figure 6.3
The Dissociation Rate Constant versus Temperafure for SC0-880 (---) and AC0-880 r---1.
The values for the dissociation rate constants will be used in Chapter 7 to determine the initiator efficiency of the inisurfs.
6.3 Critical Micelle Concentration
The critical micelle concentration of the two inisurfs under investigation have been measured by means of maximum pressure bubble tensiometry (Section
lnisurfs from Antarox C0-880
4.6.2).
101
For comparison the CMC of the symmetrical inisurf has also been
measured by means of the Du Nouy-ring metbod (Section 4.6.1 ). Not withstanding the fact that the CMC is temperature dependent 12 , the measurements have been carried out at room temperature due to the thermolability of the inisurf. Fig. 6.4 and 6.5 show the surface tension versus concentration.
The CMC can be easily
determined by extrapolation of the two linear regions in these plots.
Table 6.2
gives the measured CMC of the two insurfs with the 'different methods.
Good
agreement was found between the two methods.
15.----------------------,
so.----------------------.
30L-~~~~~~~~~~~
0.10
0.20
0.30
0.40
0.50
0.00
0.20
0.40
0.60
0.80
1.00
1 molr )
Figure 6.4
Surface Tension versus Concentration of SC0-880.
Table 6.2
Surface Tension versus Concentration of AC0-880.
Criticle Micelle Concentration (mol dm'3). Du Nouy Metbod
Maximum Pressure Bubble Metbod 2.0
SC0-880 AC0-880
Figure 6.5
6.4 10·4
w-s
6.3 10'4
Since the behaviour of the inisurf on the surface of polymer particles is of interest, it is also important to know the surface coverage as a function of the concentration in the continuous phase, i.e., the adsorption isotherm must be known. The adsorption isotherm and, thus, the surface coverage as reported by Kronberg et a/. 13 for nonylphenol poly(ethylene oxide) are assumed to be applicable for the
102
Chapter 6
inisurfs 14 •
As a cheque the adsorption isotherm of the asymmetrical inisurf on
polystyrene polymer particles has been measured. The measurements were carried out by means of maximum pressure bubble tensiometry (see Section 4.6.2) on a cleaned seed latex (see Section 4.3). The surface tension was measured as function of the overall inisurf concentration. By using the curve, obtained by measuring the surface lension as a function of inisurf concentratien in pure water (Fig. 6.5), and by applying a mass-balance, it was possible to calculate the amount of inisurf adsorbed at the surface as a function of the inisurf concentratien in the aqueous phase (Fig 6.6). The adsorption (r) has been expressed as the number of mo!es per square meter of surface and the concentratien of inisurf (X) as the number of J.UllOles per mole salution (inisurf + water).
2,50
n 2.00 1.50
1.00
(....
0.50
0.00 0.0
2.0
4.0
6.0
8.0
10.0
12.0
X (f.llllol mol-I)
Figure 6.6
Adsorption Isothenn for AC0-880 at 293 K.
This adsorption isotherm seems to obey approximately a Langmuir-type expression:
r where
rM
r
KX Ml+KX
(6.1)
is the maximum amount of inisurf that can be adsorbed per unit surface
area and K is a constant, which can be seen as an equilibrium constant governing the partilioning of inisurf between the surface layer and the bulk phase 15 • The Langmuir equation applied to solutions is derived under the assumption that the following conditions are met: ( 1) the absorbent is homogeneous, (2) the
lnisurfs from Antarox C0-880
103
adsorption takes place in only one molecular layer, (3) the solvent and the solute have equal molecular cross-sectional surface areas, and (4) there is no net solutesolvent interaction at the surface or in the bulk phase, i.e., the denvation is based on an ideal mixing model. For nonionic surfactants on latex particles 16, and also for inisurfs based on these surfactants, the two first assumptions are reasonable, but the latter two conditions are clearly not met.
Nevertheless, an equation of the
Langmuir-type generally gives a good fit of the adsorption isotherms of surfactants. By applying the Flory-Huggins theory for polymer solutions, it has been shown17 that deviations from assumptions (3) and (4) cause oppositely directed deviations from
a Langmuir-type
adsorption,
i.e.,
the
Langmuir-type
expression
is
approximately obeyed due to a compensation of two counteracting effects. If llr is plotted versus 1/X (see Fig. 6.7), a straight line is obtained within
experimental error. This straight line is in accordance with the Langmuir equation. Rewriting Eq. 6.1 gives:
1 = rM1 KrM 1 (1) x
-
_+ __ _
r
(6.2)
The intercept of the line in Fig 6.7 gives llrM• which is equal to the cross-sectional surface area at complete coverage, aad, of the inisurf on polystyrene latex particles. K, the equilibrium constant, can be related to a standard molar free energy of
adsorption, AJ.t: AJ.t
= -RT ln(K)
(6.3)
The various values determined are summarized in Table 6.3. For comparison a set of literature data is given as well 13 . The values obtained are of the same order of magnitude as those obtained for the adsorption of nonionic surfactants on hydrophobic surfaces 13 •16 • Keeping in mind that the structure of the asymmetrical inisurf is only slightly different from that of the nonionic surfactants, can explain the deviation in the values obtained. The adsorption isotherm data were used for determining the fractional surface coverage (see Chapter 7).
104
Chapter 6
Table 6.3
rM
Inisurf AC0-880 NF-E20
Adsorption Data of AC0-880.
.
11~0
K
aad
(mol m·2)
(Hf m2 mor 1)
w-6 1.56 w-6
4.43
3.7 l(f
-31.2
6.41
6
-37.0
2.26
(kJ mol- 1)
3.0 10
• NF-E,o is a nonyl phenol poly(ethylene oxide) surfactani with 20 ethylene oxide units
sr-----------------------~
0.00
Figure 6.7
0.50
1.00
1.50
2.00
2.50
3.00
1 Ir versus 1 /X for AC0-880 on a Polystyrene Seed.
6.4 Ab Initio Reactions with SC0-880 and AC0-880
The results of ab initia emulsion polymerizations with the inisurfs containing the nonionic surfactant Antarox Co-880 in the absence of any co-surfactant at 323 K will be discussed in detail. Fig. 6.8 shows the conversion-time curve obtained for a system with SC0-880. For comparison the results, obtained for a system with ACPA (la) and Antarox Co-880 (2d), have been plotted as wel!. As can be seen the reaction rate of the "control" system was much higher than that of the inisurf system. It was assumed that due to the cage-effect and subsequent geminate recombination, the initiation rate was much smaller for the inisurf system. This led to Jess particles early in the reaction and, thus, to a smaller initia! rate.
105
lnisurfs from Antarox C0-880
As can be seen in Fig. 6.8, the rate of the inisurf system increased dramatically as the reaction proceeded and approached the rate of the "control" system closer.
This indicated that partiele nucleation occured up to higher
conversions than for the ACPN Antarox recipe.
The partiele size distri bution
(PSD) at the end of the reaction was obtained by means of transmission electron microscopy (see Section 4.5.2). Fig 6.9 shows the PSD, obtained with the inisurf system. As can be seen, this distribution had a long tail of very smal! amplitude in the higher partiele size range, thereby establishing, that at the start of the reaction a small number of particles was formed and that partiele nucleation continued in the course of the reaction. This continuous nucleation rate could be due to the cageeffect: the initiating ability disappeared but the substance formed by geminate recombination still was a surfactant. This surfactant could stabilize particles, but it could also form micelles, that functioned as precursers for new particles.
This
process would continue until the monomer dropplets disappeared. Fig. 6.10 shows the PSD of the control reaction. A rather narrow PSD with an average diameter, which is slightly smaller than the one obtained for the symmetrical inisurf system, was found. The large tailing in the higher range of the PSD did not occur, which indicates that a cage effect, if present, was not important in this system.
80
480
600
t (min)
Figure 6.8
Conversion-Time Curves for the ab lnitio Polymerization with Symmetrical Initiator SC0-880 (•) and with ACPA (la) & Antarox C0-880 (2d) ( t:.). ([SC0-880] = {ACPA] = 0.5 [Antarox C0-880])
106
Chapter 6
o.Js,-------;;:::;;:'Flr"""'""""'==----..,~.oo
Partiele Diameter (nm)
Figure 6.9
Partiele Size Distribution at the End of the ab Initio Reaction with SC0-880 (7d).
0.35 ~
1.00
0.30
~
~
0.80
c:::
0 ·.:::: 0.25
i\ ·c: ~
i5 ~
0.60
0.20
c:::
.g :::> .0
·.s i5"'
0.15
0.40
0 .."
~
Ii:
0.20
5
55
105
155
205
255
295
,
.~
0.06
•
