Formation of 2D spherulites in Langmuir films of amphiphilic T-shaped liquid crystals

Journal of Colloid and Interface Science 372 (2012) 192–201

Contents lists available at SciVerse ScienceDirect

Journal of Colloid and Interface Science www.elsevier.com/locate/jcis

Formation of 2D spherulites in Langmuir films of amphiphilic T-shaped liquid crystals Sascha Reuter a, Elkin Amado a, Karsten Busse a, Martin Kraska b, Bernd Stühn b, Carsten Tschierske a, Jörg Kressler a,⇑ a b

Department of Chemistry, Martin Luther University Halle-Wittenberg, D-06099 Halle (Saale), Germany Institute of Condensed Matter Physics, Technical University Darmstadt, D-64289 Darmstadt, Germany

a r t i c l e

i n f o

Article history: Received 18 October 2011 Accepted 20 January 2012 Available online 28 January 2012 Keywords: p-Terphenyl liquid crystal Langmuir film Brewster angle microscopy Langmuir–Blodgett film Multilayer formation Surface spherulites

a b s t r a c t Langmuir films of facial T-shaped amphiphilic liquid crystals were studied at the air–water interface. The liquid crystals were composed of three incompatible segments: a central rigid rodlike p-terphenyl (TP) group, two flexible hydrophobic n-alkyl terminal chains of identical length linked through ether bonds, and one hydrophilic lateral chain of three ethylene oxide units with a carboxyl end group. In order to determine the influence of the alkyl chain length on the characteristics of condensed films three TPs having n-alkyl chains with eight (TP8/3), ten (TP10/3) or 16 (TP16/3) carbon atoms were investigated. Surface pressure – mean molecular area isotherms revealed clear differences. TP8/3 and TP10/3 exhibit an extended plateau region where a phase transition from monolayer to multilayer takes place. On the other hand, the TP16/3 isotherm showed a distinct maximum (‘spike’) corresponding to a surprising surface crystallization process which is reported for the first time for a Langmuir film of a liquid crystal. Brewster angle microscopy clearly confirmed these differences: TP8/3 and TP10/3 formed circular domains with liquid crystalline order, while TP16/3 formed well-defined two-dimensional polycrystalline spherulites which are fractured after further compression. The film thickness determined by X-ray reflectivity measurements correlated with a multilayer formation for TP10/3. The morphology of Langmuir–Blodgett (LB) films transferred onto silicon wafers and studied by atomic force microscopy also confirmed the striking different behavior (multilayer formation vs. 2D crystallization) of the TPs under investigation. Ó 2012 Elsevier Inc. All rights reserved.

1. Introduction Amphiphilic molecules spread at the air–water interface are able to form insoluble Langmuir films [1,2]. The compression behavior of such films can be described through their Langmuir isotherms, recording the surface pressure p as a function of the mean molecular area mmA [3]. Brewster angle microscopy (BAM) enables to observe the morphology of the films directly on the water surface with a lateral resolution in the micrometer range [4,5]. Further techniques revealing detailed structural information of Langmuir films at molecular level are infrared reflection absorption spectroscopy (IRRAS) [6,7] and X-ray or neutron reflectivity measurements (XR or NR) [3,8]. After transfer of Langmuir films onto solid supports by the Langmuir–Blodgett (LB) technique, the morphology of the resulting LB films can be studied by atomic force microscopy (AFM) with resolutions in the nanometer range [9–14]. ⇑ Corresponding author. Address: Von-Danckelmann-Platz 4, D-06120 Halle (Saale), Germany. Fax: +49 345 5527017. E-mail address: [email protected] (J. Kressler). 0021-9797/$ - see front matter Ó 2012 Elsevier Inc. All rights reserved. doi:10.1016/j.jcis.2012.01.036

Anisometric molecules of amphiphilic nature form frequently liquid crystalline (LC) phases in the bulk [15] and are often able to arrange themselves into stable Langmuir films at the air–water interface [16–23]. A typical example are calamitic (i.e. rod-shaped) liquid crystals which find broad industrial application in liquid crystal displays for electronic devices [24]. A well-known calamitic molecule is 4-n-octyl-40 -cyanobiphenyl (8CB) consisting of a biphenyl group as rigid unit with a hydrophilic cyano group at one end and a hydrophobic n-octyl chain at the opposite end of the ring system. The Langmuir isotherm of 8CB shows a large plateau region of slightly increasing p values which is characteristic for a multilayer formation [25–28]. A possible collapse mechanism is the sliding of layers on top of each other typically resulting in a trilayer, as firstly proposed by Ries for the collapse of monolayers of 2-hydroxytetracosanoic acid [29]. Similarities to this behavior have been found for 8CB as demonstrated by BAM [26,27] and by AFM investigations [28]. A plateau region is also observed in Langmuir isotherms of other LC molecules and is linked to the formation of multilayers [30–34]. Another explanation for a plateau region is the reorientation of the molecules from a horizontal to a vertical alignment at the air–water interface [35]. Also studies on Langmuir films of the

S. Reuter et al. / Journal of Colloid and Interface Science 372 (2012) 192–201

193

calamitic 4-cyano-400 -n-pentyl-p-terphenyl (abbreviated as T15 or 5CT) [36–39] have been reported. Compared to 8CB, 5CT has an additional phenyl group in the rigid unit (i.e. the rigid unit consists of a p-terphenyl group) to which a shorter hydrophobic n-pentyl chain is attached. The Langmuir isotherm of 5CT differs clearly from the Langmuir isotherm of 8CB, since the monolayer collapse occurs at higher surface pressure and lower mmA value [38]. This means, 5CT forms a more stable monolayer than 8CB, since due to the presence of an additional benzene ring its molecules are densely packed and in a nearly perpendicular orientation to the water surface [40]. The formation of multilayers does not take place which is indicated by the absence of a plateau region in the Langmuir isotherm of 5CT. Similar molecular arrangements as for 5CT are reported also for terthiophene derivatives [41]. A novel class of calamitic LC molecules with amphiphilic character is that of T-shaped facial p-terphenyl derivatives (TPs) [42–44]. In this class of compounds, two hydrophobic n-alkyl chains are attached in terminal positions to the rigid central pterphenyl group, and a hydrophilic chain in lateral position. Some new types of thermotropic LC phases, such as the honeycomb-like cylindrical phase, have been reported for TPs recently [45,46]. A characteristic feature of Langmuir isotherms for TPs is the presence of a plateau region that has been assigned either to a multilayer formation (similar to 8CB) or to a reorientation of the monolayer molecules from a horizontal to a vertical alignment with respect to the water surface (similar to 5CT). The preferred organization depends on type, position, and length of the hydrophilic group of the TPs [42–44]. In this study, we investigate Langmuir and LB films of T-shaped p-terphenyl derivatives in which the hydrophilic lateral chain is composed of three ethylene oxide units with a carboxyl end group. The focus lies on the behavior of the TPs at the air–water interface depending on the length of the hydrophobic n-alkyl chains. Three TPs differing in the length of their n-alkyl chains are investigated by Langmuir trough measurements and BAM experiments. Layer thickness and surface roughness are measured in situ by XR. The surface morphologies of the resulting LB films are studied by AFM after transferring the Langmuir films onto solid supports. The formation of two-dimensional polycrystalline spherulites in monolayers at the air–water interface is a seldom phenomenon that has been reported before almost exclusively for polymerized monolayers of diacetylene [47]. To the best of our knowledge, this is the first report on the formation of two-dimensional polycrystalline spherulites in monolayers of liquid crystals.

2.2. Surface pressure measurements

2. Experimental section

LB films of TPs were transferred at different p values. The deposition was performed onto silicon wafers cut into pieces of around 25  15 mm and cleaned with double distilled water to gain a hydrophilic SiO2 surface. The compression was performed with a compression rate of 15 cm2 min1 until the desired transfer surface pressure was achieved. Then, the films were allowed to equilibrate for 20 min. Afterwards, the transfer of the films was carried out at constant surface pressure by vertical upstroke of the submersed silicon wafers through the Langmuir film with a constant rate of 5 mm min1. Finally, the LB films were stored for drying in a desiccator at room temperature for 24 h.

2.1. Materials The facial p-terphenyl derivatives (TPs) were synthesized and characterized as reported elsewhere [48–50]. A general chemical structure of the T-shaped TPs is shown in Fig. 1a. Fig. 1b is a calotte model with the corresponding molecular dimensions for TP10/3. The length of the p-terphenyl group including the terminal oxygen atoms of the ether bonds is 15.3 Å and its width is 4.6 Å. In the calotte model, a and b indicate the lengths of the terminal n-alkyl chains and of the lateral chain for the different TPs, respectively. A height of 4–5 Å for the p-terphenyl group is assumed from the height of a flat lying biphenyl group [51]. Length data for a and b are given in Table 1. The notation for the different TPs is as follows: the first number after ‘TP’ represents the number m of carbon atoms in the terminal n-alkyl chains and the second number is the number n of ethylene oxide units in the lateral chain.

Langmuir isotherms at the air–water interface were obtained at 20 °C using a Langmuir trough system (KSV Instruments Ltd., Finland) placed in a sealed box. A microroughened Wilhelmy plate of platinum was used for measuring the surface pressure p (calculated from c0–c where c0 is the surface tension of pure water and c is the reduced surface tension due to the presence of surface-active molecules at the interface) as a function of the mean molecular area mmA. It should be mentioned that the mmA values are average values which do not take into account the possible formation of multilayers or the submersion of parts of the molecules into the subphase. Double distilled water was purified by a Purelab option system (ELGA Ltd., Germany) equipped with an organic removal cartridge (conductance < 0.06 lS cm1). Spreading solutions were prepared by dissolving the TPs in chloroform of HPLC grade (1–5 g L1). Predetermined amounts of these solutions (10–50 ll) were spread evenly and dropwise at the air–water interface using a digital Hamilton microsyringe. After waiting for 20 min to ensure the total evaporation of chloroform and the uniform distribution of the molecules at the air–water interface, the measurements were performed using a compression rate of 15 cm2 min1 (total trough area of 768 cm2). 2.3. Brewster angle microscopy (BAM) BAM was carried out for direct observation of Langmuir films of the TPs at the air–water interface using a MiniBAM instrument delivering images from a surface of around 7  5 mm with a lateral resolution of approximately 20 lm (Nanofilm Technologie GmbH, Germany). 2.4. In situ X-ray reflection (XR) For XR, a modified D8 Advance instrument (Bruker, USA) was used. The X-rays passed a Goebel mirror yielding Cu Ka radiation with k = 1.54 Å. The beam was collimated by two 0.1 mm slits. The reflectometer was equipped with a Vantec-1 line detector. Reflectivity was directly measured at room temperature on a MicroTrough system (Kibron, Finland) placed in a box. While collecting data, the mmA value was held constant. Typical measurement times were between 30 and 45 min. Finally, the data were corrected for background scattering. 2.5. Deposition of Langmuir–Blodgett (LB) films

2.6. Atomic force microscopy (AFM) The morphologies of LB films were studied using an atomic force microscope NanoWizard (JPK Instruments AG, Germany) working in tapping-mode. Silicon cantilevers of type Arrow (NanoWorld AG, Switzerland) with a resonance frequency of about 285 kHz and a force constant of about 42 N m1 were used.

194

S. Reuter et al. / Journal of Colloid and Interface Science 372 (2012) 192–201

a

H2 m +1 C m O

a

15.3 Å

OCm H 2m +1 O

O

4.6 Å

n

(a)

b

COOH

(b)

Fig. 1. (a) General chemical structure of TPs. (b) Calotte model for the TPs using the example of TP10/3. The symbols a and b indicate the lengths of terminal n-alkyl chains and lateral chain respectively. Values of a and b for the TPs are provided in Table 1.

to the mmA requirement of the molecule. Upon further compression, the surface pressure increases up to a kink at mmA = 66 Å2 (mmAbp, value at the beginning of the plateau region) and p = 22 mN m1 (pbp). During this increase, the n-alkyl chains are assumed to lift off from the water surface. Thus, a monolayer of densely packed, flat lying p-terphenyl groups remains at the air– water interface. This seems to be reasonable, since the mmAbp value of 66 Å2 coincides well with the theoretical mmA requirement of 70 Å2 for one flat lying p-terphenyl group including the two ether bonds linking the n-alkyl chains. Theoretical mmA requirements are based upon calculations from the dimensions provided in Fig. 1 and Table 1. After the kink in the Langmuir isotherm, a plateau region follows in which the surface pressure is only slightly increasing up to 23 mN m1 with further compression (pep, value at the end of the plateau region). This plateau indicates a phase transition in which multilayer domains are formed out of the monolayer. The exact mechanism is still not clear, but it is probable that a sliding of layers on top of each other takes place as it has been reported for other LC molecules [25–28,30–34,42–44,52– 54]. Since the end of the plateau is observed at mmA = 18 Å2 (mmAep) which is around one-third of the mmAbp value of the monolayer (66 Å2), the formation of three monolayers packed into a trilayer is a reasonable assumption [25–28,30,42–44,53,54]. Surface potential measurements within the plateau of similar TPs

Table 1 Molar masses and lengths of the terminal n-alkyl chains (a) and the lateral chain (b) for the different TPs. m and n represent the number of carbon atoms in the terminal nalkyl chains and the number of ethylene oxide units in the lateral chain. TPs a

TP8/3 TP10/3 TP16/3

m

n

Molar mass (g mol1)

a (Å)

b (Å)

8 10 16

3 3 3

693 749 917

8.8 11.3 18.9

15 15 15

a Investigations on TP8/3 reveal an identical behavior at the air–water interface as observed for TP10/3. Thus, the single results are neglected in this report.

3. Results and discussion 3.1. Langmuir isotherms Fig. 2a shows the Langmuir isotherm of TP10/3 (results for TP8/ 3 are nearly identical and are therefore given only as Supporting Information). Characteristic features are two steep increases of the surface pressure upon compression which are separated by an extended plateau region. The first significant deviation of p from zero occurs at mmA = 105 Å2 (mmAfi). At this point, the hydrophilic lateral chain is still submerged into the subphase and the n-alkyl chains are mostly located on the water surface contributing

50

50 C10H21O O

O

30

OO

O O

40

3

bp

ep

10

O

0

nd

2 maximum

30

20

O

fi

60

90 2

mmA [Å ]

120

150

0

30

3

COOH

st

1 maximum

O

O fi

ep

(b)

0 30

O

fc

10

(a)

0

OC16 H33

O

COOH

-1

fc

20

C16H33O

π [mN m ]

-1

π [mN m ]

40

OC10 H21

60

90

OO O

120

150

2

mmA [Å ]

Fig. 2. Langmuir isotherms of (a) TP10/3 and (b) TP16/3 measured at 20 °C. The abbreviations and lines denote the determination of characteristic p and mmA values (see text). BAM images of the TPs were recorded at mmA values indicated by (O), XR measurements were performed at positions shown as (h) and LB films were transferred at surface pressures marked by (D).

195

S. Reuter et al. / Journal of Colloid and Interface Science 372 (2012) 192–201

revealed that a reorientation of the molecules to a perpendicular orientation with respect to the water surface does not occur [42]. Upon further compression, a second increase of the surface pressure is observed resulting in the final collapse of the trilayer at mmA = 16 Å2 (mmAfc, value at the final collapse) and p = 33 mN m1 (pfc). In Table 2, an overview of all characteristic mmA and p values is given. Fig. 2b displays the Langmuir isotherm of TP16/3 which differs clearly from that of TP10/3 and presents two distinct maxima. Only in the low surface pressures region are both Langmuir isotherms similar suggesting the same initial state of the molecules at the air–water interface. The mmAfi value at 106 Å2 in the Langmuir isotherm of TP16/3 thus indicates that the lateral chain is submerged into the subphase at this point and that the n-alkyl chains are located on the water surface contributing to the mmA requirement of the molecule. Considering the fact that the mmAfi values of TP16/3 and TP10/3 are similar, it is concluded that the n-alkyl chains of TP16/3 should be oriented more upright to the water surface than the n-alkyl chains of TP10/3, since they are longer and thus more hydrophobic. A similar behavior has been reported for fatty acids and acyl-triazoles [3,55]. Upon further compression, the Langmuir isotherm of TP16/3 shows in analogy to TP10/3 an increase of the surface pressure due to the further lift-off of the n-alkyl chains. In contrast to TP10/3, this increase abruptly ends up in a maximum at mmA = 79 Å2 and p = 14 mN m1 followed by a steep decrease in surface pressure forming a peak. Similar ‘spikes’ in compression isotherms have been reported previously, although their origin remains elusive. In the case of fatty acid monolayers of palmitic (C16), stearic (C18) and arachidic acid (C20) the formation of a crystalline trilayer has been postulated as explanation. [56–58]. There are also some isolated reports on ‘spikes’ in isotherms of liquid crystals. For a family of phenyl benzoates linear calamitic liquid crystals [52] they were assumed to be related to some undetermined crystallization phenomena. For 4-n-alkyl-40 cyanobiphenyls (nCB) with C11 and C12 alkyl chains (11CB, 12CB) and the related compounds obtained by replacing one benzene ring of bisphenyl by cyclohexane trans-4-n-alkyl(40 -cyanophenyl)cyclohexanes (nPCH) with C9 and C10 alkyl chains (9PCH, 10PCH), the presence of a ‘spike’ was mentioned as indication of instability of the film during compression [59] but was not further analyzed. The same indication was given for 4-n-alkyl(40 -cyanophenyl)benzoates (nCPB) and 4-n-alkoxy-40 -cyanobiphenyls (nOCB) with C8 and C9 alkyl chains (8CPB, 9CPB, 8OCB, 9OCB) [60]. Also in the case of the terphenyl derivative 5CT, more closely related to the TPs subject of this investigation, a ‘spike’ is shown in its Langmuir isotherm at room temperature [38,61], but it has not been investigated. Based on this background, a crystallization process in the case of TP16/3 triggered by the packing of its n-hexadecyl chains and leading to a smaller mmA requirement is postulated as the origin of the discussed peak in its isotherm. After the first maximum a small plateau with slightly increasing p values (12–13 mN m1) follows in which the crystallization further proceeds. Finally, a second maximum at mmA = 53 Å2 and p = 21 mN m1 is interpreted as the completion of the crystallization process followed by the breaking off of the crystalline structures formed. The surface crystallization of TP16/3 is confirmed by reversibility studies of Langmuir isotherms and BAM investigations of Langmuir films as discussed in the next sections.

3.2. Reversibility studies of Langmuir isotherms Reversibility experiments are carried out by measuring compression and expansion cycles of the TPs at the air–water interface. Fig. 3a reveals a certain reversibility for TP10/3, although a slight hysteresis to lower mmA values is observed between the compression and expansion runs. The hysteresis is explained by the fact that not all molecules of TP10/3 ordered in multilayers during the first compression (black line) disaggregate back fast enough to form a monolayer during the first expansion (blue line). Fig. 3b shows reversibility studies for TP16/3. The large hysteresis is caused by the molecules of TP16/3 forming more stable structures at the air–water interface during compression compared to TP10/3. TP16/3 molecules crystallized during the first compression (black line), as found by BAM and discussed below, but only a small amount of the crystals disaggregated back to single molecules during the subsequent expansion (blue line) and formed only an incomplete monolayer. 3.3. BAM of Langmuir films Fig. 4a–c shows BAM images of TP10/3 recorded at different mmA values during and after compression. As shown in Fig. 4a, broad bands having a contrast different to the original homogeneous monolayer are observed upon compression near to the end of the plateau region of the Langmuir isotherm at mmA = 30 Å2 (see Fig. 2a). These bands posses an internal texture which lies below the resolution of BAM and is discussed for LB films (see below). They are interpreted as indication of the formation of microscopic multilayer domains as reported for Langmuir monolayers of liquid crystals exhibiting a similar plateau region in their Langmuir isotherms [62,63]. Multilayer formation is accompanied by an increase in layer thickness as observed by XR (see next section). After compression is stopped at mmA = 25 Å2 (and the surface pressure at this mmA is kept constant for 3 min), domains of variable size grow further and are already visible by BAM (Fig. 4b). At longer waiting times, the domains with circular or elliptical shape reach diameters up to 500 lm (Fig. 4c). The circular domains posses an internal liquid crystalline order in a probably smectic phase [64]. The liquid crystalline order is due to the pi-stacking of the pterphenyl groups of the molecules. In literature, circular domains were observed by BAM during compression within the plateau region in Langmuir isotherms of similar TPs [42,44] and 8CB [26– 28,30,38] and were discussed as bilayers lying on top of a monolayer. Fig. 4d–f present BAM images of TP10/3 during expansion. It is observed that the circular domains disappear again with expansion and finally form a homogeneous monolayer. This confirms the partially reversible behavior of TP10/3 during the compression and expansion cycles already observed during the reversibility measurements (see Fig. 3a). Fig. 5a–c present BAM images of TP16/3 recorded during compression at different mmA values. Characteristic features are crystalline structures resembling spherulites as shown in Fig. 5a that occur when the first maximum in the Langmuir isotherm (see Fig. 2b) has been passed and the plateau region has been reached. Thus, they confirm that the first maximum is due to surface crystallization of TP16/3. Spherulites are polycrystalline aggregates composed of radially oriented microcrystals arranged within a

Table 2 Characteristic data for the Langmuir isotherms of the TPs. The meaning of the subscripts of the mmA and p values is given in the text.

a

TPs

mmAfi (Å2)

mmAbp/pbp (Å2)/(mN m1)

mmAep/pep (Å2)/(mN m1)

mmAfc/pfc (Å2)/(mN m1)

TP10/3 TP16/3

105 106

66/22 79/14a

18/23 60/13

16/33 10/33

Data indicate the values at the first maximum in the Langmuir isotherm of TP16/3.

196

S. Reuter et al. / Journal of Colloid and Interface Science 372 (2012) 192–201

50

50 C10 H21 O

OC10H21 O

-1

3

O

O

COOH

-1

O

30

OC16H33

40

π [mN m ]

40

π [mN m ]

C16 H33 O

20

30

3

COOH

20

10

10

(a)

0 0

(b)

0 30

60

90

120

150

0

30

60

90

120

150

2

2

mmA [Å ]

mmA [Å ]

Fig. 3. Reversibility studies of Langmuir isotherms of (a) TP10/3 and (b) TP16/3 measured at 20 °C with a compression/expansion rate of 15 cm2 min1. The different cycles show first compression (black line), first expansion 2 min after reaching the collapse pressure (blue line) and second compression (green line). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

(a)

(b)

500 µm

(c)

500 µm

(e)

(d)

500 µm

500 µm

(f)

500 µm

500 µm

Fig. 4. BAM images of TP10/3 recorded (a) during compression at mmA = 30 Å2, (b) 3 min after compression up to mmA = 25 Å2, (c) 15 min after compression up to mmA = 25 Å2, and during expansion at (d) mmA = 45 Å2, (e) mmA = 70 Å2, (f) mmA = 85 Å2. The measurements were carried out at 20 °C with a compression rate of 15 cm2 min1 and an expansion rate of 132 cm2 min1. The corresponding Langmuir isotherm is given in Fig. 2a.

spherical envelope which are formed during bulk crystallization far from equilibrium conditions (quenching) [65] of a broad variety of substances. They are the typical crystallization mode for polymers such as poly(propylene) and poly(e-caprolactone). However, spherulites have been observed in the bulk for a broad variety of substances including the melt of small organic molecules such as phtalic acid and sorbitol, elemental selenium, graphite and sulfur, metals such as iron and some metal alloys [65,66]. Spherulites are also grown from aqueous solutions of some low-soluble salts such as calcite and apatite, and protein solutions of insuline and lysozyme [66]. The shape of spherulites is due to the noncrystallographic branching that takes places during crystal growth, and spherulitic growth appears to be a universal crystallization mechanism for all crystalline substances, under some specific conditions

[66]. Surprisingly, the growth of two-dimensional spherulites in monolayers at the air–water interface is practically unknown and has been reported up-to-date only for polymerized monolayers of diacetylene removed from the water surface and observed by polarized optical microscopy [47]. This might be due to the physical constraints imposed by the extreme thinness of a monolayer to investigations by polarized microscopy: in order to observe the birefringence of an extremely thin film a highly anisotropic polarizability is required [47]. Thus the high birefringence of the terphenyl group originated from its elongated p-electron conjugated system [67] is probably the reason that makes the in situ observation of 2D spherulites at monolayers of TP16/3 possible. At this point a probable source of confusion must be mentioned: the spherulitic textures frequently seen for droplets of thermotropic

197

S. Reuter et al. / Journal of Colloid and Interface Science 372 (2012) 192–201

(a)

(b)

500 µm

(d)

(c)

500 µm

(e)

500 µm

500 µm

(f)

500 µm

500 µm

Fig. 5. BAM images of TP16/3 recorded during compression at (a) mmA = 70 Å2, (b) mmA = 40 Å2, (c) mmA = 25 Å2, and during expansion at (d) mmA = 105 Å2, (e) mmA = 110 Å2, (f) mmA = 120 Å2. The measurements were carried out at 20 °C with a compression rate of 15 cm2 min1 and an expansion rate of 132 cm2 min1. The corresponding Langmuir isotherm is given in Fig. 2b.

liquid crystals between crossed polarizers do not contain any solid crystal, since they build an homogeneous phase down to the molecular level [66]. Thus, the visualization of spherulitic textures for a thin sample of a liquid crystal has no direct implications regarding the possibility of having polycrystalline spherulites at monolayers of the substance as those reported in this study. In fact, both TP8/3 and TP10/3 exhibit spherulitic textures corresponding to columnar mesophases [48] when observed between crossed polarizers inside the temperature range of their liquid crystalline behavior, but none of them shows 2D spherulites. On the contrary, TP16/3 exhibits only a smectic A phase [48], and therefore no spherulitic texture between crossed polarizers is seen. However, it is the only sample exhibiting 2D spherulites. Coming back to the BAM images of TP16/3 in Fig. 5b, after the spherulites have fully developed they impinge on each other causing an increase in surface pressure up to a breaking point (second maximum) is reached Then, fracture lines appear inside the spherulites as seen in Fig. 5c and the surface pressure drops. During expansion (Fig. 5d–f), the crystalline structures of TP16/3 seem to retain their morphology irreversibly even at large mmA values,

i.e. the crystals formed during compression are very stable and they are not able to disaggregate during expansion in agreement with the reversibility studies. 3.4. X-ray reflection of Langmuir films Reflectivity curves have been analyzed in terms of Fresnel reflectivity [68] in the case of pure water and with the Abeles matrix method [69] for monolayer reflectivity, using MOTOFIT [70] and IGOR Pro [71] and applying a genetic fit algorithm. For all monolayer fits, only a single layer with a uniform density has been assumed. Fig. 6 shows the X-ray reflectivity patterns for TP10/3 and TP16/ 3 corrected for Fresnel reflectivity. The lowest curves in both diagrams represent the reflectivity of the water subphase without any monolayer. A typical surface roughness of 3.1 ± 0.1 Å was found. The surface pressure dependent raw data of the TPs show systematic changes in the reflection behavior. Starting at high mmA values, the interference increases with decreasing mmA. In the case of TP10/3 (Fig. 6a), a significant Kiessig fringe appears.

Fig. 6. X-ray reflectivity patterns of (a) TP10/3 and (b) TP16/3 at different mmA values. The curves are shifted for clarity. The respective positions for the scattering data on the Langmuir isotherms are indicated by (h) in Fig. 2.

198

S. Reuter et al. / Journal of Colloid and Interface Science 372 (2012) 192–201

Table 3 mmA dependent layer thickness and surface roughness of Langmuir films of TP10/3 and TP16/3. TPs

mmA (Å2)

Layer thickness d (Å)

Surface roughness r (Å)

TP10/3

105 85 70 22

9.2 ± 0.2 9.3 ± 0.2 10.0 ± 0.2 21.8 ± 0.3

3.8 ± 0.3 4.6 ± 0.3 5.0 ± 0.1 4.0 ± 0.2

TP16/3

100 85 67 40

10.3 ± 0.2 11.0 ± 0.1 12.2 ± 0.1 17.3 ± 0.2

6.2 ± 0.2 7.2 ± 0.1 8.5 ± 0.2 4.3 ± 0.1

The simple monolayer model provides an excellent fit of nearly all data, and deviations between measurements and the fit appear only at the highest compression. However, the main feature of the data is still captured by the fit. Remaining differences may be due to internal structures within the Langmuir film not accounted for by the single layer model. The resulting fit parameters (layer thickness d and surface roughness r) for the Langmuir films of TP10/3 and TP16/3 are summarized in Table 3. It should be mentioned that the lateral chains of the ethylene oxide units do not contribute to the film thickness because of the similarity of their electron density with that of water. It is clear that the layer thickness for TP10/3 increases with compression of the Langmuir film. The layer thickness data at mmA of 85 and 70 Å2 can be explained with the assumption that the nalkyl chains in the monolayer are lifted off from the water surface

and contribute to the monolayer thickness. Near to the end of the plateau at mmA of 22 Å2, a considerably increased layer thickness of 21.8 ± 0.3 Å is measured which is significantly larger than the theoretical height of a monolayer. This confirms the formation of a certain multilayer structure in the plateau region of TP10/3. The surface roughness is almost independent on mmA, but decreases slightly when the plateau region of the Langmuir isotherm is passed. The layer thickness of the Langmuir film of TP16/3 (Fig. 6b) follows initially the same pattern as for TP10/3. Compression of the film up to the first maximum in the Langmuir isotherm (see Fig. 2b) results in slightly increased layer thickness and surface roughness. Crystallization, as clearly seen by BAM (see Fig. 5), causes a sudden decrease in surface roughness (see Table 3 at mmA values before and after crystallization, e.g. at 85 Å2 and 40 Å2). Concerning the coherence length of the X-ray beam (Ic  1 lm), the method is sensitive to surface roughness of lengths smaller than Ic. Taking into account the BAM images, the reflections of the crystalline surface dominate the signal when the second maximum in the Langmuir isotherm at mmA = 53 Å2 is passed resulting in smaller surface roughness. In contrast, the signal at larger mmA is an average reflection by crystalline and noncrystalline areas leading to the higher surface roughness observed. 3.5. Morphology of LB films Fig. 7a presents the AFM images for the LB film of TP 10/3 transferred within the plateau region of the Langmuir isotherm at

Fig. 7. (a) AFM height image of the LB film of TP10/3 transferred at p = 22 mN m1, mmA = 55 Å2 (indicated by D in Fig. 2a). (b) AFM height image of the LB film of TP10/3 transferred at p = 22.5 mN m1, mmA = 34 Å2 (indicated by D in Fig. 2a). (c) Height profile taken along the line in (b).

199

S. Reuter et al. / Journal of Colloid and Interface Science 372 (2012) 192–201

Fig. 8. (a) AFM height image of the LB film of TP16/3 transferred at p = 12.6 mN m1, mmA = 72 Å2 (indicated by D in Fig. 2b). (b) AFM height image of the LB film of TP16/3 transferred at p = 15 mN m1, mmA = 60 Å2 (indicated by D in Fig. 2b). (c) Height profile taken along the line in (b).

(a)

(b)

Fig. 9. Schematic representation for the behavior of TPs at the air–water interface upon compression for (a) TP10/3 (TP8/3) and (b) TP16/3.

p = 22 mN m1 and mmA = 55 Å2 (indicated by D in Fig. 2a). Characteristic features are circular, fluid like domains of various lateral dimensions between 0.2 and 5 lm. The height image (Fig. 7b) for a second LB film transferred at a higher p value of 22.5 mN m1 and a smaller mmA of 34 Å2 (towards the end of the plateau region and indicated by D in Fig. 2a) reveals that the circular, fluid like

domains coalesce to laterally larger domains upon further compression. However, these domains are still too small to be resolved by BAM and appear only as a region of different contrast as in Fig. 4a. In the corresponding height profile (Fig. 7c) a uniform height of the domains in the range of 3.5–3.8 nm is found which is significantly larger than the theoretical height of a monolayer of

200

S. Reuter et al. / Journal of Colloid and Interface Science 372 (2012) 192–201

TP10/3. This confirms that the circular domains are indeed multilayers. The difference between the height of the circular domains measured by AFM and the layer thickness obtained from XR measurements is caused by the fact that the XR signal is an average over a measuring area of several square centimeters. Compared to TP10/3 rather different morphologies are observed for LB films of TP16/3. For a LB film transferred within the plateau region of the Langmuir isotherm at p = 12.6 mN m1 and mmA = 72 Å2 (indicated by D in Fig. 2b), needle-shaped structures of variable heights are found (Fig. 8a). These structures are crystalline fragments of the same 2D-spherulites, visualized for TP16/3 before the transfer of the LB film onto silicon wafers had been carried out (see BAM images in Fig. 5). Only fragments of the spherulites are observed in the LB film, since they break up during transfer. This indicates the extrem fragility of surface spherulites cause by mechanical stressed built in during their growth at conditions far from equilibrium, in analogy with the mechanical instability observed for some bulk spherulites [72]. LB film transferred at a higher p value of 15 mN m1 and mmA of 60 Å2 (Fig. 8b and c, indicated by D in Fig. 2b) show fragments coming from fully developed spherulites for which extensive impingement has already occurred resulting in larger structures. Both thick bundles of needles up to 150 nm width, and far thinner single needleshaped crystals are observed. The different behaviors of the TPs during compression at the air–water interface as concluded from the experimental results are schematically summarized in Fig. 9. The striking differences between TP16/3 and TP10/3 are caused by the longer n-hexadecyl chains of TP16/3 being able to trigger crystallization (Fig. 9b). In contrast, the n-alkyl chains of TP10/3 are too short to crystallize and thus, the liquid crystalline nature of the rod-shaped molecules, with their tendency to pi-stacking of the p-terphenyl groups (Fig. 9a) dominates the self-assembly in TP10/3 monolayers. 4. Conclusions A totally different behavior of facial T-shaped p-terphenyl derivatives (TPs) during compression of Langmuir films at the air–water interface has been found depending on the length of their two n-alkyl terminal chains. TPs with shorter n-octyl (TP8/3) and n-decyl chains (TP10/3) go through a phase transition from monolayer to multilayers organized in circular fluid domains with diameters around 0.2–18 lm. In contrast, TP16/3 having longer n-hexadecyl chains crystallizes as two-dimensional polycrystalline spherulites, with sizes in the hundredths of micrometers range. The formation of surface spherulites has been observed before only for polymerized monolayers of diacetylene [47]. The shape of the surface pressure/area (p – mmA) isotherm of TP16/3 exhibits a characteristic ‘spike’ that has been reported previously in the literature, but without any link to spherulitic crystallization, for some other substances: fatty acids such as palmitic (C16), stearic (C18) and arachidic acid (C20) [56–58]; linear calamitic liquid crystals such as phenyl benzoates [52], 4-n-alkyl-40 -cyanobiphenyls (nCB) with C11 and C12 alkyl chains (11CB, 12CB) [59]; trans-4-n-alkyl(40 -cyanophenyl)cyclohexanes (nPCH) with C9 and C10 alkyl chains (9PCH, 10PCH) [59]; 4-n-alkyl(40 -cyanophenyl)benzoates (nCPB) and 4-n-alkoxy-40 -cyanobiphenyls (nOCB) with C8 and C9 alkyl chains (8CPB, 9CPB, 8OCB, 9OCB) [60]; and also for the terphenyl derivative 5CT [38,61]. The occurrence of such ‘spikes’ in isotherms of the amphiphilic substances mentioned having relatively long n-alkyl chains suggests that the formation of 2D spherulites might also occur for those substances and be the actual origin for some reports on monolayer instability to compression and the observation of undefined ‘‘irregular strips’’ and ‘‘hair-like objects’’ in the BAM images for some of them [59,60].

On the other hand, there have been isolated reports on bulk spherulitic crystallization from the melt in thin films (some tenths of micrometers thick) between two glass plates for palmitic (C16) and stearic (C18) acid [73], and for the liquid crystals 4-n-alkoxy40 -cyanobiphenyls with a C8 alkyl chain (8OCB) [74] and a C10 chain (10OCB) [72]. Interestingly, for those substances the characteristic bulk crystallization is not spherulitic and the growth of spherulites is therefore related to the crystallization of the amphiphiles under confinement. The main effect of confinement could be to slow down diffusion, which is considered to be a key requirement for spherulitic growth, besides supercooling [66]. Analogously, the occurrence of 2D spherulitic crystallization at Langmuir monolayers might be a general phenomenon, at least for amphiphilic molecules bearing long alkyl chains, arising from the added effects of molecular confinement at the air–water interface, plus the anchoring of the molecules to the water subphase through their hydrophilic groups, which effectively slows down their diffusion at the interface. Acknowledgments S.R. thanks Agrochemisches Institut Piesteritz e.V. for financial support, J.K. and C.T. thank Deutsche Forschungsgemeinschaft (FOR 1145). Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.jcis.2012.01.036. References [1] G.L. Gaines Jr., Insoluble Monolayers at Liquid–Gas Interfaces, WileyInterscience Publishers, New York, 1966. [2] A. Ulman, Introduction to Ultrathin Organic Films: From Langmuir–Blodgett to Self-Assembly, Academic Press, San Diego, 1991. [3] P. Dynarowicz-Latka, A. Dhanabalan, O.N. Oliveira Jr., Adv. Colloid Interface Sci. 91 (2001) 221. [4] S. Hénon, S. Meunier, J. Rev, Sci. Instrum. 62 (1991) 936. [5] D. Hönig, D. Möbius, J. Phys. Chem. 95 (1991) 4590. [6] E. Amado, A. Kerth, A. Blume, J. Kressler, Langmuir 24 (2008) 10041. [7] H. Hussain, A. Kerth, A. Blume, J. Kressler, J. Phys. Chem. B 108 (2004) 9962. [8] V.M. Kaganer, H. Möhwald, P. Dutta, Rev. Mod. Phys. 71 (1999) 779. [9] M.-P. Krafft, J.G. Riess, Chem. Rev. 109 (2009) 1714. [10] G. Roberts, Langmuir–Blodgett Films, Springer, Berlin, 1990. [11] M.C. Petty, Langmuir–Blodgett films, Cambridge University Press, Cambridge, 1996. [12] K.S. Birdi, Self-Assembly Monolayer Structures of Lipids and Macromolecular Interfaces, Springer, Berlin, 1999. [13] A.T. Hubbard, Encyclopedia of Surface and Colloid Science, Dekker, New York, 2002. [14] D.K. Schwartz, J. Garnaes, R. Viswanathan, S. Chirovolu, J.A.N. Zasadzinski, Phys. Rev. E 47 (1992) 452. [15] C. Tschierske, Prog. Polym. Sci. 21 (1996) 775. [16] C. Jego, B. Agricole, M.-H. Li, E. Dupart, H.T. Nguyen, C. Mingotaud, Langmuir 14 (1998) 1516. [17] O. Albrecht, W. Cumming, W. Kreuder, A. Laschewsky, H. Ringsdorf, Colloid Polym. Sci. 264 (1986) 659. [18] G. Decher, H. Ringsdorf, Liq. Cryst. 13 (1993) 57. [19] M.D. Everaars, A.T.M. Marcelis, E.J.R. Sudhölter, Thin Solid Films 242 (1994) 78. [20] D. Joachimi, C. Tschierske, A. Öhlmann, W. Rettig, J. Mater. Chem. 4 (1994) 1021. [21] D. Joachimi, A. Öhlmann, W. Rettig, C. Tschierske, J. Chem. Soc. Perkin Trans. 2 (1994) 2011. [22] M. Woolley, R.H. Tredgold, P. Hodge, Langmuir 11 (1995) 683. [23] D. Janietz, J. Mater. Chem. 8 (1998) 265. [24] D. Pauluth, K. Tarumi, J. Mater. Chem. 14 (2004) 1219. [25] J. Xue, C.S. Jung, M.W. Kim, Phys. Rev. Lett. 69 (1992) 474. [26] M.C. Friedenberg, G.G. Fuller, C.W. Frank, C.R. Robertson, Langmuir 10 (1994) 1251. [27] M.N.G. de Mul, J.A. Mann Jr., Langmuir 10 (1994) 2311. [28] J. Fang, C.M. Knobler, H. Yokoyama, Physica A 244 (1997) 91. [29] H.E. Ries Jr., Nature 281 (1979) 287. [30] M.N.G. de Mul, J.A. Mann Jr., Langmuir 11 (1995) 3292. [31] A. El Abed, J. Daillant, P. Peretti, Langmuir 9 (1993) 3111. [32] N.C. Maliszewskyj, P.A. Heiney, Langmuir 11 (1995) 1666.

S. Reuter et al. / Journal of Colloid and Interface Science 372 (2012) 192–201 [33] M. Ibn-Elhaj, H. Möhwald, M.Z. Cherkaoui, R. Zniber, Langmuir 14 (1998) 504. [34] M. Ibn-Elhaj, M.Z. Cherkaoui, R. Zniber, H. Möhwald, J. Phys. Chem. B 102 (1998) 5274. [35] A.K. Dutta, J. Phys. Chem. B 101 (1997) 569. [36] M.F. Daniel, O.C. Lettington, S.M. Small, Thin Solid Films 99 (1983) 61. [37] T. Sakuhara, H. Nakahara, K. Fukuda, Thin Solid Films 159 (1988) 345. [38] T. Martynski, R. Hertmanowski, D. Bauman, Liq. Cryst. 28 (2001) 437. [39] R. Hertmanowski, L. Chudzinski, T. Martynski, P. Stempniewicz, E. Wolarz, D. Bauman, Liq. Cryst. 31 (2004) 791. [40] P. Jedlovsky, L.B. Pártay, J. Mol. Liq. 136 (2007) 249. [41] H. Nakahara, J. Nakayama, M. Hoshino, K. Fukuda, Thin Solid Films 160 (1988) 87. [42] J.A. Schröter, R. Plehnert, C. Tschierske, S. Katholy, D. Janietz, F. Penacorada, L. Brehmer, Langmuir 13 (1997) 796. [43] R. Plehnert, J.A. Schröter, C. Tschierske, Langmuir 14 (1998) 5245. [44] R. Plehnert, J.A. Schröter, C. Tschierske, Langmuir 15 (1999) 3773. [45] F. Liu, B. Chen, U. Baumeister, X. Zeng, G. Ungar, C. Tschierske, J. Am. Chem. Soc. 129 (2007) 9578. [46] C. Tschierske, Chem. Soc. Rev. 36 (2007) 1930. [47] D. Day, J.B. Lando, Macromolecules 13 (1980) 1478. [48] B. Chen, Ph.D. Thesis, Martin Luther University Halle-Wittenberg, 2004. [49] B. Chen, X. Zeng, U. Baumeister, G. Ungar, C. Tschierske, Science 307 (2005) 96. [50] B. Chen, U. Baumeister, G. Pelzl, M.K. Das, X. Zeng, G. Ungar, C. Tschierske, J. Am. Chem. Soc. 127 (2005) 16578. [51] L. Xu, M. Salmeron, S. Bardon, Phys. Rev. Lett. 84 (2000) 1519. [52] H. Diep-Quang, K. Ueberreiter, Colloid Polym. Sci. 258 (1980) 1055. [53] I. Giner, I. Gascón, R. Giménez, P. Cea, M.C. López, C. Lafuente, J. Phys. Chem. C 113 (2009) 18827.

201

[54] B. Rapp, H. Gruler, Phys. Rev. A 42 (1990) 2215. [55] S. Reuter, K. Busse, U. Radics, H.-J. Niclas, J. Kressler, J. Colloid Interface Sci. 340 (2009) 276. [56] C. McFate, D. Ward, J. Olmsted, Langmuir 9 (1993) 1036. [57] D. Vaknin, W. Bu, S.K. Satija, A. Travesset, Langmuir 23 (2007) 1888. [58] S. Kundu, A. Datta, S. Hazra, Langmuir 21 (2005) 5894. [59] K. Inglot, T. Martynski, D. Bauman, Liq. Cryst. 33 (2006) 855. [60] A. Modlinska, K. Inglot, T. Martynski, R. Dabrowski, J. Jadzyn, D. Bauman, Liq. Cryst. 36 (2009) 197. [61] T. Martynski, Colloids Surf., A 198–200 (2002) 179. [62] J.-L. Gallani, S. Mery, Y. Galerne, D. Guillon, J. Phys. Chem. B 108 (2004) 11627. [63] Y. Tabe, T. Yamamoto, I. Nishiyama, K.M. Aoki, M. Yoneya, H. Yokoyama, J. Phys. Chem. B 106 (2002) 12089. [64] Y. Tabe, H. Yokoyama, J. Chem. Phys. 115 (2001) 1041. [65] L. Gránásy, T. Pusztai, G. Tegze, J.A. Warren, J.F. Douglas, Phys. Rev. E 72 (2005) 011605. [66] A.G. Shtukenberg, Y.O. Punin, E. Gunn, B. Kahr, Chem. Rev. (2012), doi:10.1021/ cr200297f. [67] Y. Chen, J. Sun, H. Xianyu, S.-T. Wu, X. Liang, H. Tang, J. Display Technol. 7 (2011) 478. [68] M. Tolan, X-ray Scattering from Soft-matter Thin Films, Springer, Berlin, 1999. [69] O. Heavens, Optical Properties of Thin Films, Butterworth, London, 1955. [70] A.J. Nelson, Appl. Crystallogr. 39 (2006) 273. [71] http://www.wavemetrics.com. [72] J.L. Hutter, J. Bechhoefer, J. Cryst. Growth. 217 (2000) 332. [73] P.B.V. Prasad, Cryst. Res. Technol. 19 (1984) 1549. [74] S.C. Jain, S.A. Agnihotry, V.G. Bhide, Mol. Cryst. Liq. Cryst. 88 (1982) 281.