Eur. Phys. J. E 12, 497–505 (2003) DOI: 10.1140/epje/e2004-00021-8 T HE EUROPEAN P HYSICAL JOURNAL E Crystal growth rates of diblock copolymers in thin ﬁlms: Inﬂuence of ﬁlm thickness G. Reiter a and L. Vidal Institut de Chimie des Surfaces et Interfaces, CNRS, 15, rue Jean Starcky, B.P. 2488, 68057 Mulhouse Cedex, France Received 9 September 2003 Published online 20 January 2004 – c  EDP Sciences, Societ`a Italiana di Fisica, Springer-Verlag 2004 Abstract. Using optical and atomic force microscopy we determined the growth rate of polymer crystals in thin ﬁlms of low molecular weight poly(styrene-ethyleneoxid) block copolymers. We focused in particular on ﬁlms either thinner or slightly thicker than the thickness of a crystalline lamella (L). At a given temperature, three distinctly diﬀerent growth rates were observed for i) crystals grown from adsorbed monolayers thinner than L, ii) “primary” compact lamellar crystals in ﬁlms thicker than L and iii) “secondary” crystals formed from molecules remaining on top of ii). The growth rate of primary crystals did not indicate a dependence on ﬁlm thickness. Crystals from monolayers grew by a factor of 30 more slowly than primary crystals. Secondary crystals grew faster than crystals from monolayers but slower than primary crystals and the degree of deviation from the growth rate of primary crystals was found to depend on the value of L, which varied with temperature. The mechanisms responsible for the diﬀerences in growth rate are discussed in terms of the amount of available polymers and the controlling inﬂuence of diﬀusion towards the crystal front. PACS. 61.41.+e Polymers, elastomers, and plastics – 68.47.Mn Polymer surfaces 68.55.-a Thin ﬁlm struc- ture and morphology – 81.10.Aj Theory and models of crystal growth; physics of crystal growth, crystal morphology and orientation 1 Introduction Due to the connectivity of the individual monomers, crys- tallization of long chain-like polymers is a complex pro- cess [1–5]. Many diﬀerent morphologies and meta-stable states are formed. On the other hand, because of the pecu- liarities like connectivity and the consequential constraints controlling conformations, the multitude of possibilities of how polymers can be ordered and the resulting morpholo- gies of such chain-like molecules may also be less numerous than for crystallization of small molecules. Because of this chain-like shape of the molecules (they are much longer in one direction than in the other two) polymer crystals form lamellar structures of thickness L which grow es- sentially only in two directions. The observation of such growth for “ﬂat-on” lamellae, i.e. lamellae growing paral- lel to the surface, in thin ﬁlms of a thickness comparable to the thickness of crystalline lamellae is expected to provide information on the the modes of attachment of molecules to the crystal front. Such observations may allow to test theoretical predictions as the growth direction is always linked to the plane of the ﬁlm. Various studies on crystallization of homopolymers in thin ﬁlms have been already performed [6–15]. Interest- ingly, several authors [6,12,14,15] found a signiﬁcant inﬂu- a e-mail: g.reiter@univ-mulhouse.fr ence of ﬁlm thickness on the rate of crystallization. In par- ticular, the non-monotonic thickness dependence of crys- tal growth found in [15] is highly intriguing. At this point, no satisfying argument is yet presented which could ex- plain a thickness dependence for ﬁlms much thicker than L or the size of the polymer. Conﬁnement of crystallization to 2D lamellar struc- tures is even more pronounced if an incompatible amor- phous block is attached to the crystallisable polymer. As already stated by Lotz [16], an amorphous polystyrene block is not necessarily prohibiting the formation of lamel- lar polymer crystals and often even does not much af- fect the growth rate. However, such a block enhances the separation of individual crystalline lamellae and so allows for largely independent growth of these lamellae. The sys- tem poly(polystyrene-block-ethyleneoxide) has been par- ticularly well studied [16–24]. In preparing thin ﬁlms, the initial thickness h 0 of the molten ﬁlm is usually not identical with the thickness (h) of one (h = L) or a stack of n (h = nL) crystalline lamel- lae. Thus, the number of molecules contained in a molten ﬁlm may either be too small or too large to form com- pletely ﬁlled crystalline lamellae. For h 0 <L, one has to expect that the lamella can only cover part of the substrate. This is the case for crystallisation of adsorbed monolayers [7,9,10] or spincoated thin ﬁlms [11–15] where