J. MATER. CHEM., 1994, 4(8), 1205-1213 Monolayer Behaviour and Langmuir-Blodgett Film Properties of some Amphiphilic Phthalocyanines: Factors influencing Molecular Organisation within the Film Assembly Michael J. Cook,*aJim McMurdo,a David A. Miles," Richard H. Poynter,a John M. Simmons,a Simon D. Haslam,tb Robert M. Richardsonb and Kevin Welford" a School of Chemical Sciences, University of East Anglia, Nomvich, UK NR4 7TJ School of Chemistry, University of Bristol, Cantock's Close, Bristol, UK BS8 ITS DRA, St. Andrews Road, Malvern, Worcs, UK WR743PS The monolayer and LB film-forming properties of 20 structurally related amphiphilic octa-substituted phthalocyanine derivatives have been assessed. The molecular packing within examples of the films has been probed by visible-region spectroscopy and low-angle X-ray diffraction methods.Among derivatives where the aliphatic substituents are attached by ether linkages, there is a variation in the behaviour according to the length of the chains and whether or not the chains are branched. Films are not highly ordered but may contain domains of ordered structure, giving rise to ri red-shifted absorption band in the visible spectrum. Analogues where the chains are attached by carbon-carbon bonds show superior monolayer behaviour and are excellent materials for deposition as LB films. Furthermore, there is good evidence from the spectra that change in the length of the alkyl chains provides a means of controlling the type of molecular packing within the films. Hughes' and Alexander2 investigated monolayer behaviour of phthalocyanine (Pc) derivatives in the 1930s, but it was Roberts and co-workers' research3 in the early 1980s which marked the start of the extensive interest in both Pc monolayer and LB films evident over the last ten year^.^,^ Most of the attention during this period has centred on substituted derivatives, those bearing three or four substitu- ent groups having been particularly well studied. Not all give good monolayer or deposition behaviour, but those that do often give rise to LB films with a promising degree of molecular order. The most common type of order involves the molecules stacking more or less cofacially in aligned columns whose axes are parallel to the substrate surface. This type of packing leads to a characteristic blue shift of the visible absorption band relative to that observed in the solution phase.6 Our own interests have centred on the use of octa-substituted compounds.Unlike the tetra-substituted com-pounds, which are normally formed as a mixture of isomers and used as such, the octa-substituted derivatives can be synthesized isomerically pure, a feature which should encour- age molecular ordering. In earlier work we noted the contrasting behaviour of the octa-alkoxy series 1 and the octa-alkyl series 2. The former give LB films with some degree of order depending upon chain length.7 The latter form rigid compressed aggregates at the air/water interface which cannot be deposited onto substrates8 Series 3, analogues of 2 but containing both hydrophilic and hydro- phobic groups, behaved very differently.The introduction of amphiphilic character encourages ordering at the water surfaceg and the monolayers can be deposited as highly ordered LB The molecular packing is quite different from that in the films of the tri-and tetra-substituted compounds; the visible spectrum shows both a red- and blue-shifted band characteristic of Davydov splitting, implying the presence of translationally non-equivalent mol- ecules within the unit cell.lo,ll The present paper reports an investigation of a range of t-Present address: DRA, St. Andrews Road, Malvern, Worcs, WR14 3PS. RC M la R = -(CH2),CH3; M = H,H b R = --(CH2),CH*(CH&; M = H,H c R = --(CH2)2CH.(CH3),; M = Ni R' 2 R=R'=alkyl; M=H,HorCu 3 R = alkyl; R' = -(CH2),C02H; M = H ,H or Cu new amphiphilic octa-substituted compounds, 4-14, closely related in terms of their structures.They were chosen to appraise how ring substituents, in particular, chain-lengt h and functionality, can influence both monolayer and LB film-forming properties.12 The structural packing within examples of the films has been probed by visible spectroscopq and low-angle X-ray diffraction methods. J. MATER. CHEM., 1994, VOL. 4 CI CI OH HO OH 4a R = -(CH,),CH,; M = H,H b R = -(CH2)4CH3; M = CU c R = -(CH2)4CH3; M = Ni n= 1 5a R = -(CH2),CH*(CH3),; M = H,H b R = -(CH2)2CHm(CH3)2; M = CU c R = -(CH,),CH.(CH,),; M = Ni n= 1 6 R = -(CH2)2CH*(CH3)2; M = CU n=2 7e R = -(CH2)2CH*(CH3)2; M = H,H b R =-(CH2)2CH*(CH3)2; M = CU n=3 9 R = -(CH,)&H3; M = H,H n= 1 Experimental Materials Metal-free Alkoxy Amphiphiles The metal-free amphiphilic compounds of series 4-9 were obtained in yields of 2-5% by allowing the appropriately substituted 3,6-bis-alkoxyphthalonitrileand 3,6-di( hydroxy- a1koxy)phthalonitrile to react in a 9 : 1 ratio.These precursors were prepared by reaction of the appropriate bis-cyano- hydroquinone with an alkylating agent using standard conditions, cf. ref. 13. The required amphiphilic phthalocyan- ine product was separated chromatographically from the octaalkoxyphthalocyanine by-product of type 1 [in partic- ular, 1,4,8,11,15,18,22,25-octapentyloxyphthalocyanine,la, mp 109"C (lit.,I3 116-117.5 "C) and 1,4,8,11,15,18,22,25-octa-iso-pentyloxyphthalocyanine, lb, see Table 11.The general procedure, an adaptation of the synthesis of compounds of series 1 described earlier,13 is exemplified by the following. 1,4-Di (3-hydroxypropyloxy)-S, 1 1,15,18,22,25-hexaisopentyl-oxyphthalocyanine, (5a):Using a typical procedure, a mixture of 3,6-di(isopenty1oxy)phthalonitrile (2.16 g, 9 mmol) and 3,6-di(3-hydroxypropyloxy)phthalonitrile (0.28 g, 1mmol) in dry pentan-1-01 (10 ml) was heated to reflux and lithium metal (0.1 g) added in small pieces. Reflux was continued for 45 min and then the mixture allowed to cool to room temperature (rt) when glacial acetic acid (10 ml) was added and stirring continued for 1 h.The solvents were removed under reduced pressure and the residue taken up in chloroform (50 ml) and washed with water (50 ml), saturated brine (50 ml), dried with (MgS04), filtered and the chloroform removed under reduced pressure to give a dark green viscous oil. This was chromato- graphed over silica gel Merck grade 7734 using as eluent a mixture of 40-60 light petroleum-THF 9: 1 to give a green solid which was recrystallised from THF-methanol to give 1,4,8,11,15,18,22,25-octaisopentyloxyphthalocyanine( 1b) (612 mg, 28%) as green needles mp 201 "C. (Found: C, 71.8; HO 10 R = --(CH,),CH,; M = H,H 11 R = -(CH,)&H,; M = H,H 12 R = -(CHp)$H3; M = H,H 13a R = --(CH,)&H$ b R=--(CH,)&H,; c R = -(CH2)&H3; M = H,H M=CU M = Ni 14a R = -(CH,),CH,; M = H,H b R = --(CH,)&H,; M = CU H, 8.2; N, 9.3YO.C,,H,,N,O, requires C, 7 1.5; H, 8.3; N, 9.3YO). 6H (60 MHz, CDC1,): 0.24 (s, 2H), 1.05 (d, 48H), 2.2 (m, 24H), 5.0 (t, 16H), 7.6 (s, 8H).Increasing the eluent polarity (40-60 light petroleum-THF 4 : 1)gave a second green fraction which was chromatographed a second time over silica gel (eluent cyclohexane-THF 3 :2). (In later preparations it was shown that chromatographic separations could be improved by modifying the eluent through the addition of triethylamine at 1YO.)Recrystal-lisation from THF-MeOH afforded dark green crystals of 1,4-di( 3-hydroxypropyloxy)-8,11,15,18,22,35-hexaisopenty1-oxyphthalocyanine [63 mg, 6% based on the 3,6-di(3-hydroxypropyloxy)phthalonitrile] mp 196-197 "C.(Found: C, 69.5; H, 7.6; N, 9.5%. C68H90N@10 requires C, 69.2; H, 7.7; N, 9.5%). 6, (400 MHz, CDCl,): 0.24 (s, 2H), 1.02 (d, 12H), 1.06 (d, 12H), 1.08 (d, 12H), 1.60 (m, 6H), 2.00 (m, 4H), 2.15 (m, 8H), 2.27 (m, 4H), 3.93 (m, 4H), 4.27 (t, 2H), 4.87 (m, 12H), 5.04 (t, 4H), 7.58 (d, 2H), 7.6 (s, 2H), 7.63 (d, 2H), 7.66 (s, 2H). Metal-free Alkyl Amphiphiles Compounds 10, 12 and 14a were available from a recent study of mesogenic phthalocyanines. l4 The homologues 11 and 13a were prepared similarly in yields of 8 and 6%, respectively. Metallated Phthalocyanines Metal-free phthalocyanines were converted into their metallated derivatives by reaction with the corresponding metal acetate, e.g.:1,4,8,11,15,18-Hexadecyl-22,25-di(4-hydroxy-butyl)phthalocyaninatocopper(n), ( 14b): In a typical experiment, a mixture of 1,4,8,11,15,18-hexadecyl-22,25-di-(4-hydroxybuty1)phthalocyanine(91 mg, 61 pmol) and cop- per@) acetate (400 mg, 2 mmol) in dry pentan-1-01 (20 ml) was heated under reflux for 30 min.The mixture was allowed to cool to rt and the pentan-1-01 removed under reduced J. MATER. CHEM., 1994, VOL. 4 Table 1 Characterisation of novel compounds compound C H N Q-band no. M substituent mp/"C" formula found (reqd.) LaJmmb lb 20 1 71.8 8.2 9.3 762 (l.52)c (71.85 8.3 9.3) 740 (1.31) lc 220 68.9 7.7 8.8 734 ( 1.73)d (68.6 7.7 8.9)4a 6 x OC,H1, 101-103 69.1 7.9 9.4 762 (1.76) 2 x O(CH2hOH (69.2 7.7 9.5) 738 (1.49) 4b 6 x OC5H11 149-150 65.8 7.3 8.9 740 (1.87) 2 x O(CH,),OH (65.8 7.15 9.0)4c 6 x OC5Hll 209-210 65.7 7.4 9.45 733 (1.20) 2 x O(CHz),OH (66.1 7.2 9.1)5a 6 (CH2)2CH(CH3 12 196-197 69.5 7.6 9.5 763 (1.60)2 x O(CH,),OH (69.2 7.7 9.5) 740 ( 1.36)5b 6 0(CH2)2CH(CH3)2 225-226 65.4 7.0 8.9 740 (2.10) 2 x O(CH,),OH (65.8 7.15 9.0)0(CH2)2CH(CH3)2 269-271 65.9 6.9 9.0 733 (1.74) 2 x O(CH2),0H (66.1 7.2 9.1)0(CH2)2CH(CH3)2 230-232 66.5 7.05 8.5 739 (2.111) 2 x O(CH2),0H (66.25 7.3 8.8)7a 6 0(CH2)2CH(CH3)2 65-67 70.4 8.2 8.6 762 (1.33) 2 x O(CH2)50H (70.0 8.0 9.0) 740 (1.59) 7b 6 x O(CH2)2CH(CH3)2 87-88 66.3 7.4 8.3 739 2 x O(CH2),0H (66.7 7.5 8.6)0(CH2)2CH(CH3)2 104 59.3 6.3 7.8 757 (1.78) 6 x C1; 2 x O(CH2),0H (58.9 6.1 8.1) 731 (1.56) 0(CH2)2CH(CH3)2 132-133 55.8 5.9 6.9 739 (1.38) 6 x C1; 2 x O(CH2),0H (56.1 5.75 7.3)9 6 x OC7H15 66-67 71.1 8.5 8.1 762 (1.43) 2 x O(CH2)30H (71.3 8.5 8.3) 738 (1.25) C7H15 K-Dl18 78.9 9.7 8.6 700 ( 1.25)" 2 x (CH,),OH D+I152 (78.9 9.5 9.0) 735 (1.72) 13a 6 CgHl9 K+D98 79.6 10.25 7.5 695 (1.32)" 2 x (CH2),0H D-+I122 (79.8 10.0 7.9) 730 (1.47) C9H19 K+D154 76.05 9.6 7.4 710 (2.17)' 2 x (CH2),0H D-I170 (76.4 9.55 7.6) 13c 6 CgHl9 K+D95 76.5 9.5 7.4 705 (1.01 )" 2 x (CH2),0H D-I123 (76.65 9.6 7.6) 14b 6 x C10H21 151-1 52 76.8 10.0 6.9 710 (1.98)" 2 x (CH,),OH I-D151 (76.8 9.9 7.2) D-K136 "Melting points and transition temperatures for liquid-crystalline materials.K, crystal state; D, discotic mesophase; I, isotropic liquid.bMeasured as solutions in toluene unless indicated otherwise, as for: 'in cyclohexane, din dichloromethane. 'in l,l,l-trichloroethane. pressure. The residue was chromatographed over silica gel available to the material at a rate of 50 cm2 min-' until (Merck grade 7734;eluent 40-60 light petroleum-THF 2:1) a compressed state was achieved at surface pressurcs of and recrystallised from THF-methanol to give as green crys- ca. 25-40 mN m-l depending upon the sample. The stability tals 1,4,8,11,15,18-hexadecyl-22,25-di(4-hydroxybuty1)phthal-of the compressed monolayer as a function of time was ocyaninatocopper(II), (84mg, 88%). (Found: C, 76.8;H, 10.0; assessed by monitoring the percentage decrease in area over N, 6.9%.C100H154CuN802requires: C,76.8;H, 9.9;N,7.2%). a particular timespan. Characterisation data for all novel compounds are presented LB films for X-ray diffraction and optical spectroscopic in Table 1. Mps were measured using a Linkam hot-stage studies were prepared by depositing compressed monolayers attached to an Olympus polarising microscope. Mesophase (typically at surface pressures of 30-35 mN m-') onto silicon behaviour was observed for compounds 10-14. and glass slides cleaned as described earlier." These were rendered hydrophobic by silanising the slides in a 2-4% solution of dichlorodimethylsilane in l,l,l-trichloroet hane Monolayer Behaviour and LB Film Deposition for 30 min. Experiments were performed on either a Joyce-Loebl con-Vertical dipping through the molecular monolayer was stant-perimeter trough or a NIMA 622 dual-compartment undertaken with the surface pressure maintained constant trough.Water purification procedures, equipment and pro- during the dipping process. Dipping speeds were 10 mm niin-' cedure for preparing multilayer LB films have been discussed for compound la and 8 or 15mm min-' for all others. A previou~ly.',~ drying time was frequently incorporated into the dipping Material was transferred to the water surface as solutions sequence depending upon the material being deposited, see in l,l,l-trichloroethane at known concentrations of ca. Discussion. 0.5 mg ml-' with the trough barriers fully open. Monolayer 'Horizontal lifting' of monolayers of 10, 12 and 14a was behaviour was examined once the solvent had been deemed attempted onto glass slides.The slides were arranged horizon- to have evaporated (30 min). The barriers were then slowly tally and lowered onto the the monolayer at ca. 1 mm rnin-l closed to 5-15 mN m-' and then reopened. Monolayer behav- and then raised, the process then repeated to obtain the iour was then plotted as a PA isotherm by reducing the area required number of layers. Optical Spectroscopy Visible spectra were measured using either a Hitachi U2000 or U3000 spectrophotometer. The substrate was aligned such that it was normal to the sample beam. Polarised spectra were measured using Polaroid sheet polarisers placed between the incident beam and the sample holder and recorded with the incident light polarised both parallel and perpendicular to the dipping direction.X-Ray Diffraction Analysis of the films was performed using a high-resolution X-ray reflectometer which has been describcd elsewhere." Collimated Cu-Ka X-rays were used (A=1.54 A). The sample LB film was placed on a horizontal table and the angle of incidence and angle of reflection were scanned simultaneously so that they remained equal. The reflected intensity was recorded as a function of the scattering vector Q =4nsinO/A, where 6' is the angle (in radians) between the incident (or reflected) beam and the substrate. Results and Discussion Materials Of the compounds 4-14 all bar 10, 12 and 14a are novel, the synthesis of the last three having been described re~ent1y.l~ The remainder were prepared by analogous routes, gave satisfactory elemental analyses (Table 1) and, in the case of the metal-free and nickel derivatives, 'H NMR signals fully consistent with their structures.Proton signals in the aromatic region proved particularly diagnostic, compounds of series 4, 5, 7 and 9 showing two singlets and an AB pattern while an accidental equivalence of protons of 11 and 13 reduces the signals to two apparent singlets in the ratio of 3 : 1, cf. ref. 14. A detailed 'H and 13C NMR analysis of compounds 5a and 9 will be presented elsewhere.16 Previously we demonstrated that 10, 12 and 14a exhibit discotic columnar liquid-crystalline beha~iour.'~ Of the new materials, 11, 13a,b,c and 14b also give rise to a liquid-crystal phase and transition temperatures are summarised in Table 1.Monolayer Behaviour Material at the air/water interface was compressed and decom- pressed to give the conventional n-A isotherms. Examples obtained for the first compression/relaxation cycle are shown in Fig. 1. Area per molecule, Ao, data (Table 2) were obtained by extrapolating the region of the isotherm which corresponds to the condensed phase to zero pressure. Table2 also gives the results of monolayer stability tests, which are presented as the percentage decrease in surface area of films compressed to a specified pressure over a specified time. Octaalkoxyphthalocyanines, la, 1b, lc: Compound la was examined in our earlier work,7 but has been re-measured alongside the new derivatives lb and lc for comparative purposes. Data for la confirmed the finding that this com- pound gives isotherms which show substantial hysteresis during relaxation [Fig.1(a)]. Inspection of a spate-filling model of the compound suggtsts a width of ca. 22A and a packing thickness of ca. 4.5-5 A. Thus the measured A, value, 70 A, is inconsistent with monolayer behaviour. Inspection of the TC-Aisotherm indicates that there is some instability in the condensed phase at surface pressures >15 mN m-l and there is a significant decrease in surface area over time. Compound lb gives a similar type of isotherm. Again there is hysteresis but the condensed phase appears to be stable to ca.20 mN m-l, there is a much smaller decrease in area with J. MATER. CHEM., 1994, VOL. 4 201 n 10, , , , 0 30-z 20-E--.2 - 20- L2 Q 10-. 10- 0 1 0 \ 0 0 100 200 0 100 200 area per molecule/A* Fig. 1 Examples of n-A isotherms showing the initial compression/ relaxation cycle for (a) 1,4,8,11,15,18,22,25-octapentyloxyphthalo-cyanine (la), (b) 1,4-di(3-hydroxypropyloxy)-8.11,15,18,22,25-hexa-isopentyloxyphthalocyanine (5a), (c) 1,4,8,1 1,15,18-hexaheptyloxy- 22,25-di(3-hydroxypropyloxy)phthalocyanine(9), (d) 1,4,8,11,15,18-hexadecyl-22,25-di(4-hydroxybutyl)phthalocyaninatocopper (11) ( 14b), (e) 1,4-di(4-hydroxybuty1)-8,11,15.18,22,25-hexa~~ctylphthalocyanine (12) and (f) the n-A isotherm for 1,4-di(4-hydroxybutyl)-8,11,15,18,22,25-hexaoctylphthalocyanine(12) obtained for the second compression/relaxation cycle time, and the value for A, is 113 A.The latter is more consistent with a monolayer. Thus the behaviour of the branched-chain isomer at the air/water interface is substan- tially superior to that of the straight-chain compound. The nickel analogue lc behaves similarly but exhibits a small enhancement in monolayer stability. Alkoxy amphiphiles, 4-9: Compounds of series 4-7 are structurally related to la and lb differing insofar as two of the pentyloxy or isopentyloxy groups have been replaced by two hydroxylated alkoxy groups. The introduction of amphiphilic character provides a marked improvement in monolayer behaviour which is broadly shared by each member of the series.Relative to la and lb, compounds of series 4-7 show much less hysteresis in the n-A isotherm and the compressed states are more stable, retaining their integrity at pressures of ca. 30-35 mN m-l. The n-A isotherm for com- pound 5a is fairly typical and is shown in Fig. l(b). During compression of the film, the surface pressure starts to rise at a higher surface area per molecule than for la and lb. The slope is not steep but becomes steeper at cu. 20 mN m-l and continues to increase as such, at least until a surface pressure of 35mN m-l is reached. The less steep region may corre- spond to the liquid-expanded region or a phase in which the molecules are reorganising from a flat-on to a more perpen- dicular arrangement.Earlier workg on a longer-chained hom- ologue of la indicated that the molecules are lying flat on the water surface at low surface pressures, an orientation presum- J. MATER. CHEM., 1994, VOL. 4 Table 2 Monolayer and deposition behaviour stability %dec. (surface deposition transfer dipping compound A,-,/A2 a pressure/mN m-’)’ tYPe ratio rate/mm inin-’ la lb IC 4a 4b 4c 5a 5b 5c 70 113 134 120 131 123 135 155 140 lO(25) 10( 15)* 6( 15)* < l(30) <l(30) <1(30) 1(35)<l(35) < l(35) Z Y Y Y-z Y-z Y+Z Y Y Y 0, 0.6 0.9, 1 0, 0.8 0, 0.9 0, 1 0.8, 1 0.7, 1 0.9, 1 C 10 10 10 15 15 15 15 15 15 6 7a 7b 8a 8b 129 125 146 115 140 5(25) 1(35) 1(30) < l(35) < l(30) Y Y Y Y d 0.8, 1 1, 1 0.9, 1 0.9, 1 15 15 15 15 9 > 150 26( 15)* e 10 124 5-10( 30)* Y 0.9, 0.9 8 11 12 13a 121 124 124 <2( 30)* < 1(30)* <1 (30)* Y Y Y 1, 1 1, 1 0.9, 1 8 8 8 13b 13c 14a 14b 137 117 137 125 <2(30)* < 1(30)* <1 (30)* <1 (30)* Y Y Y Y 1, 1 1, 1 1, 1 1, 1 8 8 8 8 “Value for the average area per molecule in the compressed state, extrapolated to zero surface pressure.’Stability of monolayer measured as the percentage decrease in surface area at the surface pressure indicated in parentheses () over 20 min or, in the case of measurements noted by *, over 1 h. ‘The monolayer collapsed during deposition, rendering measurement impossible. dDeposition not attempted. ‘No deposition observed.ably encouraged by weak interaction between the oxygen atoms in the ether links and the surface water molecules. As the surface pressure is increased the molecules are compressed up towards the vertical. Similar behaviour is postulated for the series 4-7, though the values for A. are sometimes larger than expected for molecules fully perpendicular to the water surface. Where comparisons are possible it appears that monolayer behaviour is marginally improved for the metallated species and that straight-chain pentyloxy compounds, 4, give some- what less stable monolayers than their isomers of series 5. The length of the hydrophilic chain, which is varied through series 5-7, appears not to be a factor in controlling monolayer properties in a significant way.The introduction of six chlorine substituents onto the ring system, as in 8, also has little apparent effect on the monolayer properties, indicating that the factors which dominate the behaviour are the isopentyloxy groups and the presence of the hydroxy group at the end of the other two chains. The longer-chain homologue of 4a, the heptyloxy analogue 9, behaves less well. The n-A isotherm, Fig. l(c), is of the same general form as those of the other ‘alkoxy amphiphiles’ but the surface pressure starts to rise at a higher surface area per molecule. The inflection occurs at ca. 15 mN m-’ and there is rather more hysteresis on decompression from 25 mN m-l. Above 25 mN m-l the film collapses. It is unlikely that the molecules reach the fully perpendicular orientation because the value for A, extrapolated from the isotherm in Fig.1(c)is too large to be consistent with this. The compressed film is much less stable over a period of time than the films of 4a and 4b, suggesting that the length of the alkoxy chain is rather critical in controlling monolayer behaviour. This is in accord with our earlier work on series 1 which showed that the octapentyloxy derivatives were better behaved than those with either shorter or longer chains.’ Alkyl amphiphiles, 10-14: Replacement of the ether linkages by methylene groups has a very marked effect on the appear- ance of the n-A isotherm. These compounds, with the excep- tion of 12, behave very similarly.They show well defined transitions from the two-dimensional gas phase to the condensed phase, the latter characterised by a near-vertical region in the isotherm, e.g. Fig. l(d), with no evidence tor the intervening state apparent for the alkoxy amphiphiles. Values for A, indicate that the molecules are oriented with their planes more or less perpendicular to the surface. The isotherms are very similar in form to those recorded earlier for series 3. An earlier X-ray reflectivity study of a monolayer of the latter indicated’ that the molecules are essentially perpendicular even in the uncompressed film and the new compounds may behave similarly. Compound 12 gave rise to a slightly different isotherm which showed a transitional state prior to the formation of the fully condensed phase, Fig.l(e). This inter- mediate state is not observed on the second compression/ decompression cycle, Fig. l(f),which gives an isotherm com- parable to that of the other compounds on the first cycle, Fig. l(d). Comparisons of 4a and 10, and of 9 and 12, are particularly interesting. These pairs of compounds are ‘isosteric’, having the same number of linking atoms in the chains, arid the superior monolayer behaviour of the two alkyl amphiphiles illustrates how replacement of the ether links by methylene groups markedly affects the monolayer properties. A property which distinguishes the alkyl amphiphiles from the (ilkoxy compounds is that they give rise to a liquid-crystal phase in the bulk material.The reason for this difference between the two series is unclear at the present time but may be the key to the differences in the monolayer behaviour. Thus the alkyl amphiphiles are examples of amphotropic material^'^ in that they may self-assemble through both their capacity to form a mesophase and through their amphiphilic character. ,4s the surface area is reduced the molecules may well mimic the behaviour observed in the liquid-crystal state wherein the aromatic cores align in columns. The n-A isotherms, which show minimal hysteresis, suggest that such columns break up as readily as they are formed when the surface pressure decreases. Compound 10 in the compressed state showed a small decrease in surface area over time, which was unexpected in view of its otherwise good monolayer behaviour.To investi- gate this further, the monolayers of each of the alkyl amphi- philes were compressed to the point where they collapsed. The surface pressure required to cause collapse of the films was similar in each case at ca. 50 mN m-l. This suggests that the monolayer of 10 is not intrinsically less stable, and we propose that the reduction in surface area reflects a gradual reorganisation of the molecules in the monolayer to give a more closely packed monolayer. It may be significant that the deposited films of 10 contain a different type of molecular assembly to the others in this series, vide infra. Deposition Behaviour Octaalkoxy phthalocyanines, la, lc: The present experiments on the deposition of the simple octapentyloxy compound la gave rise to 2-type deposition in contrast with Y-type observed in our earlier series of experiment^.^ We have no explanation for this inconsistent behaviour.A characteristic of the films was the manner in which they were wetted during the emersion process. To alleviate this, a drainage time of up to 20 min was incorporated into the cycle after each dip. Films were deposited with an irregular transfer ratio and were of variable quality and often visibly non-uniform. The behaviour of the branched-chain derivative, lc, was strikingly different. Y-type transfer was observed under conditions where the straight- chain compounds deposited 2-type and there was apparently less wetting after emersion.Transfer ratios were satisfactory (0.8-1.0) and the films appeared uniform on visual inspection. Alkoxy amphiphiles, compounds 4a, 4b, 4c, deposited Y- type over the first dipping cycle with subsequent 2-type deposition on subsequent cycles. An attempt to deposit 4b onto hydrophilic rather than hydrophobic substrates gave 2- type deposition throughout. In common with the symmetri- cally substituted compounds, la and lc, the films were found to have been wetted after the upstroke. Transfer ratios ranged from 0.7-1.0, which suggested an even coverage of the sub- strate, but visual inspection showed that the films were usually non-uniform. The longer-chain analogue 9 failed to deposit. Unlike the straight-chain compounds, the isopentyloxy substituted amphiphiles, 5-8, deposited in a regular Y-type manner, giving even green films.Transfer ratio data indicated consistent coverage of the substrate throughout the deposition experiment. The presence of three, four or five methylene units in the hydrophilic tails of these compounds, 5, 6 and 7, respectively, made no apparent difference to the film-forming properties. The chlorinated derivatives, 8, behaved in a similar fashion to the unchlorinated compounds. The uniformity of transfer of a representative example, 5b, was investigated by optical spectroscopy, plotting absorbance against number of layers and, in a separate study, bY capacitance methods. Both techniques showed that the build- up of the film is uniform.There was limited wetting of these films during dipping, suggesting that the branched chains are more hydrophobic than the straight chains. Alkyl amphiphiles, 10-14: Each compound was deposited by the vertical dipping method and 10, 12 and 14a were also transferred to a substrate by the horizontal lifting technique. The former method of transfer gave very even films by Y-type deposition with constant transfer ratios close to unity. There was little apparent wetting of the films after emersion. However, it was found to be advantageous to allow the films to drain for 1h after the first cycle. Subsequent dipping was carried out without incorporating a drying time between cycles. Plots of absorbance against number of dips were linear at least for films of up to 20 dips.J. MATER. CHEM., 1994, VOL. 4 Films deposited by the vertical-lifting method at 30 mN m-' gave a transfer ratio greater than unity, viz. 10, 1.90; 12, 1.80; 14a, 1.42. While X-type films are expected by this technique, there are precedents for the fabrication of Y-type structures, molecules apparently turning over and being drawn in underneath as the substrate is lifted. For the present compounds, lower values for the transfer ratios were obtained when the surface pressure was 25 mN m-'. Characterisation of LB Films Visual inspection revealed that the most even films were obtained from the metallated isopentyloxy amphiphiles and the alkyl amphiphiles. Where comparisons were possible it appeared that films obtained of the latter by the horizontal- lifting method approached the quality of those obtained by the vertical-dipping procedure.Alkoxy amphiphiles: Compounds 5b and 7b deposited to give particularly even films and were evaluated by X-ray reflectivity. Neither film gave any Bragg peaks, suggesting a lack of layer ordering. However, the diffraction from the film of 7b does exhibit thickness fringFs which transform to a coherent film thickness of c?. 280A. This corresponds to an amount of material ca. 22 A thick deposited during each Y- type dipping cycle. The electronic spectra of phthalocyanines show a character- istic absorption in the visible region referred to as the Q-band. Non-aggregated metal@) Pcs in the solution or gas phase show a single main absorption assigned to the doubly degener- ate transition alu-eg, see Fig.2. For metal-free Pcs the lower symmetry of the system lifts the degeneracy and the Q-band is split into two components. In the solid state the spectra are rendered more complex through exciton coupling which broadens peaks and leads to shifts in the band positions which are dependent upon molecular pa~king.~ The visible absorption spectra were recorded of films of 4b, 4c,5b, 5c and 7b and each gave a remarkably similar spectrum; data are collected in Table 3 and the spectrum of the LB film of 5b is shown as Fig. 2(a). In each case the major component of the Q-band is broad and red shifted relative to the toluene solution spectrum.There is a second, lower-intensity absorp- tion to higher energy. The spectra of films of 4b and 4c show no dichroism. However, films of 5b,5c, and 7b give rise to dichroism such that the absorption intensity is lower when the electric field vector is polarised perpendicular to the dipping direction, El than parallel to it, Ell. The dichroic Table3 Visible spectral data for LB films of some alkoxy and alkyl amphiphiles compound i,,,/nm (El : El,)" lc 4b 4c 5b 5c 7b 10 11 12 13a 13b 13c 14a 14b 752 (1.0) 758 (1.0) 752 (1.0) 758 (0.75) 753 (0.96) 750 (0.77) 740 (1.7) 768 (0.92) 767 (0.93) 770 (0.71) 769 (0.64) 743 (0.74) 771 (0.75) 767 (0.78) 637 (1.36) 634 (1.23) 638 (1.32) 635 (1.07) 638 (1.02) 639 (1.12) 632 (1.0) "Main absorption band(s) in the visible region.Data in parentheses () show the dichroic ratio, (El:El,),the ratio of the absorbances recorded with the electric field vector, E, perpendicular and parallel to the dipping direction. J. MATER. CHEM., 1994, VOL. 4 1211 0.500T (a 1.000 @) I A I Q,0 0.500 0.500 -e (d) v)a a 0.400 0.400 -0.300 0.300-0.200 0.200-0.100 0.100-0.00~ o.ooo+ wavelengthlnm Fig. 2 Visible spectra of LB films (glass slides) of (a) 1,4-di( 3-hydroxypropyloxy)-8,11,15,18,22,25-hexaisopentyloxyphthalocyaninatoco~~per(11) (5b), (b) 1,4,8,11,15,18-hexahexyl-22,25-di(4-hydroxybuty1)phthalocyanine(lo), (c) 1,4-di(4-hydroxybutyl)-8,11,15,18,22,25-hexanonylphthalo-cyaninatocopper(rr) (13b) and (d) 1,4-di(4-hydroxybuty1)-8,11,15,18,22,25-hexanonylphthalocyaninatonickel(11)(13c). Spectra are recorded with the electric field vector polarized: A, perpendicular to the dipping direction, El; and B, parallel to the dipping direction, Ell.The third line, C, in each plot corresponds to the solution phase spectrum and is shown for comparison. ratio is essentially constant over the visible region with El: Ell =0.75, 0.94 and 0.77 for 5b, 5c and 7b, respectively. In general, the observation of dichroism points to an overall anisotropic arrangement of the molecules in the film. The greater absorption observed for films of 5b, 5c and 7b when E is parallel to the dipping direction indicates that the rings are preferentially aligned such that the mean angle of the planes of the rings relative to the dipping direction is <45 "C.The absence of well defined layer spacing deduced from the X-ray reflection study, even in the films which do show dichroism, points to a lack of long-range order through the film, i.e. perpendicular to the substrate. This does not, however, preclude short-range in-plane molecular order and it is this which may give rise to the red-shifted absorption. Using Kasha's 'dimer' model," based on a pair of interacting parallel transition dipoles, the molecular exciton theory predicts a red-shifted absorption if the transition dipoles are offset by a specific amount. Extrapolated in terms of a columnar arrange- ment of phthalocyanines, a red-shifted absorption is predicted for a column in which the planes of the molecules form an angle, 8, to the stack axis which is <54.7 "C.The absence of dichroism observed in the films of the other alkoxy amphiphiles points to a lack of anisotropic alignment, though note that an arrangement of the molecules with the planes of the rings parallel to the substrate surface will not produce dichroism either. We suggest that, as a series, the alkoxy derivatives deposit to form films with varying degrees of short-range order. The domains of order may contain molecules arranged with their planes offset from their nearest neighbours; where dichroism is observed, the domains are overall anisotropically ordered. Alkyl amphiphiles: X-Ray reflectivity studies were per-formed on 30-layer films (15 vertical dips) of 10, 12, 14a and 14b which were taken as representative examples.These gave results different from those of the alkoxy amphiphiles. At higher angles they show two Bragg peaks, indicative of molecular layers parallel to the substrate, e.g. Fig. 3. Tht: layer spacing, d, was calculated by applying Bragg's law io the second-order peaks (i.e. nA=2d sin 0, with n =2), Table 4. Earlier experience suggests that the position of these peaks (unlike the position of the 001 peaks) is not significantly distorted by X-ray refractive index effects in the film, arid the 8L6 I I , I ,A 0 0.1 0.2 0.3 0.4 0.5 QIA Fig. 3 X-Ray reflections recorded for a 30-layer film of 1.4-di(4-hydroxybutyl)-8,11,15,18,22,25-hexaoctylphthalocyanine (12) on glass Table 4 Summary of XRR data measured for LB films (15 dips) of compounds 10, 12, 14a and 14b deposited on silicon (unless indicated otherwise) d-spacing frqm thickness calculate4 compound 002 peak/A (from fringes)/A thickness/A" ~~ 10 35 540-620 525 lob 35 540-620 525 12 40 540-620 600 14U 41 640-710 615 14b 42 630/658' 630 a Calculated from the proposed number of layers and observed bilayer spacing.bOn glass substrate, not silicon. 'Value obtained from fit (see text ). layer spacings quoted are considered to have an accuracy of +0.5 A. From molecular models we estimate the distance, w, from the OH groups to the end of fully extended alkyl chain: on the opposite side of the molecules to be 23, 24 and 25 A for 10, 12 and 14a,b, respectively.For each compound the measured value for d is greater than w, consistent with a bilayer assembly deposited by Y-type deposition. However, for each film d <2w, and considerably so in the case of 10, which implies that the molecules are to some extent tilted and/or there is interdigitation of the chains of molecules in adjacent layers. In each case the second-order peak was far stronger than the first-order peak, indicative of a bilayer structure with low electron density in the middle of the bilayer. The most intense Bragg peaks were observed for the copper-containing com-pound, 14b. This is expected due to the increased scattering contrast provided by the ion.For the metal-free derivatives, the intensities of the peaks decrease as the chains are length-ened. This may be associated with a genuine increase in disorder (layers intermix and become less well defined) or a loss in contrast in the X-ray scattering profile induced by greater interdigitation of substituent chains. All of the films exhibit Keissig fringes implying that they are reasonably uniform. The fringes are not well defined, except for the film of 14b, but fringe intensity does seem to increase as the substituent chain length is increased, suggesting that more uniform films are formed. The thicknesses of the films esti-mated from the positions of the fringes using the formula AQ =27+ are given in Table 4.AQ is the separation of adjacent fringes in the reflectivity profile and t is the thickness of the film. The calculated thicknesses do not, in general, correlate well with the thicknesses calculated from the number of dips (15 'bilayers') and the observed bilayer spacings. This may be due to the poor definition of the fringes or interference between the secondary maxima associated with the 001 peak 3 0 0 0 0 0.05 0.10 0.15 0.20 0.25 QIA Fig. 4 Fitted X-ray reflectivity profile of the LB Film of 14b J. MATER. CHEM., 1994, VOL. 4 and the fringes. This would produce a shift in fringe position. Each film exhibits a number of fringes between the critical angle and the 001 peak that supports the development of 15 bilayer films.An attempt was made to model the reasonably well defined fringes yielded by the film of 14b by calculating and fitting the reflectivity using an optical matrix formalism. This would be expected to give a more accurate determination of the film thickness. The resultant fit and the associated parameters are given in Table 5. The data were only fitted up to but excluding the first-order Bragg peak using a model of a single film of uniform electron density. The principal variable parameters in the fitting were the mean electron density of the film, the film thickness and the roughnesses of the air/film and film/substrate interfaces. The other parameters (scaling factor, background, beam divergence and electron density of air and the substrate) were kept constant at physically re$sonable values.The value of 658 & 1 A implies a film thickness of 15.7 bilayers. The large drop in reflected intensity after the 001 peak is usually indicative of the formation of a non-integral number of repeat units due to interference between the 001 peak and the Keissig fringes. It therefore appears that, con-sidering the X-ray reflectivity data and the fitted parameters, an 'extra' layer has been deposited onto the film. Interfacial roughnesses of films of phthalocyanines have not been reported before, but the roughnesses are very high when compared to those found by Mu~grove'~[or LB filmsooffatty acids deposited onto silicon (ualf.= 1.9 A, uf,,=4.4 A). The value of the mean electron density is sensible when compared to those of othet closely packed aromatic systems (for example, Pbe.,zene =0.281 A comparison of the data yielded by films of 10 on glass and silicon gives interesting results.The 'bilayer' spacings are the same, but the first-order Bragg peak is stronger in the case of the film on silicon. This implies a slightly different packing of the molecules to generate a bilayer with greater electron density at its centre. The visible absorption spectra were recorded of a film of each compound (Table 3). Spectra of LB films of 10 are somewhat similar to those of the alkoxy-substituted materials but show a higher degree of structure and a much sharper main band which replaces the split Q-band of the solution phase, Fig.2(b). The interpretation of the spectral bandshape is not straightforward and will not be pursued at the present time. The dichroism, with E, :E,,x 1.3 over the whole of the visible region, is indicative of a preferential alignment of the planes of the rings towards the normal to the dipping direction rather than along the dipping direction. In contrast, LB films of 11 through to series 14 give spectra with prominent red-and blue-shifted absorptions relative to the solution phase spectrum [Table 3 and Fig. 2(c) and (d)]. With one exception, the spectrum of the nickel derivative, 13c, where the separation of the main absorptions is smaller, Fig. 2(d), the spectra are closely similar in terms of band position to those obtained earlier for LB films of series 3.All show a similar dichroism such that E,:EI,> 1 for the blue-shifted band and E,:EII<l for the red-shifted band. As discussed in our earlier work, the splitting of the absorption Table 5 Parameters used to fit fringes of the LB film of 14b parameter value ~ layer thickness/A 658k 1 electron density/e k3 0.30f0.01 roughness of filmlsubstrate interface, u,,,/A 15.3f0.01 roughness of air/film interface, u,,,/A 11.7f0.01 J. MATER. CHEM., 1994, VOL. 4 band and the dichroism is indicative of Davydov splitting arising from the presence of translationally non-equivalent molecules within the 'unit cell', as would be found for a 'herring-bone' arrangement of the molecules within columns. On the basis of the simplified exciton model for an oblique arrangement of transition dipoles described by Kasha," the observed polarisation would point to the column axes aligned preferentially perpendicular to the dipping direction.Precedents for this type of column alignment where molecules are fully cofacial have been interpreted in terms of deposition of preformed columnar structures within the monolayer at the air/water interface.21 Conclusion The n-A isotherms and deposition characteristics observed for the alkoxy derivatives show that monolayer behaviour and film deposition are highly sensitive to the type of alkoxy group attached to the ring, and whether or not the compounds have been rendered amphiphilic through the modification of two of the side chains with terminal hydroxy groups.Metallated derivatives showed consistently better behaviour than the metal-free derivatives, with the latter giving LB films which appeared patchy on visual inspection. Introduction of amphiphilic character and incorporation of isopentyloxy groups in place of straight-chain pentyloxy substituents proved beneficial. Indeed, the best monolayer and deposition properties were exhibited by compounds 5b, 5c, 6, 7b and 8b. The poorest behaviour was shown by the amphiphilic com- pound 9, which suggests that the longer heptyloxy chains are disruptive to surface ordering. We believe these results are significant in the light of our recent X-ray crystallographic study of three compounds of series 1.22Compound lb shows an ordered structure in which the molecules stack in offset columns, similar to the proposed type of assembly within the ordered domains in the LB films.However, the nickel analogue of the straight-chain octapentyloxy derivative and octahep- tyloxy phthalocyanine both recrystallise to form solids which do not give a significant diffraction pattern, i.e. there is no long-range order. Thus it appears that the more compact isopentyloxy groups help to confer order, whereas the straight- chain alkoxy groups disfavour order. Replacement of the alkoxy functionalities by alkyl groups, as in the alkyl amphiphiles 10-14, gives rise to a further improvement in monolayer and deposition behaviour. LB films of the alkyl amphiphiles are more highly ordered, as judged by the X-ray diffraction data which provide evidence of bilayer spacing and dichroism within the visible spectra.The spectra also show that the molecular packing within the film is dependent upon the length of the hydrophobic chains, with that for the shortest-chain derivative, 10, differing from that for the others. The contrast in behaviour within the pairs of isosteric compounds from the alkyl amphiphile and alkoxy amphiphile series, 10 and 4a, 12 and 9, is remarkable. It reveals that the lack of order associated with the straight- chain alkoxy groups, referred to above, is not merely a function of their length. Presumably the presence of the ether linkages introduces local interactions which are sufficient to disrupt ordered molecular packing which, in phthalocyanines, is normally associated with interactions between the aromatic cores of adjacent molecules within columnar stacks.This effect may provide the basis for an explanation as to why the alkyl amphiphiles exhibit columnar mesophase behaviour whereas the alkoxy amphiphiles do not, core: core interactions within the former overwhelming the disruptive effects of the mobile alkyl side chains. The present results illustrate the potential for controlling molecular packing within phthalocyanine LB films through modification of the substituents on the ring and the assemblies described here clearly differ from the fully cofacial packing commonly observed for the tri- and tetra-substituted deriva- tives referred to earlier.The ability to control the molecular packing in phthalocyanine LB films is potentially important for the application of these formulations in devices. Dif'ferent types of molecular packing may show different interactions with gases, of potential value in sensing devices and, :is has been demonstrated, will have different electronic absorption signatures which may prove to be of value in research into laser addressed data storage systems. The authors thank SERC for a research grant to support J.McM., CASE studentships with DRA for D.A.M, K.H.P. and S.D.H. and an SERC-IT studentship for J.M.S. References 1 A. Hughes, Proc. R. Soc. London, Ser. A, 1936,155,710. 2 A. E. Alexander, J. Chem. Soc., 1937,1813.3 S. Baker, M. C. Petty, G. G. Roberts and M. V. Twigg, Thibi Solid Films, 1983,99, 53; S. Baker, G. G. Roberts and M. C. Petty, IEE Proc. Solid State Electron. Devices, 1983, 130, 260; J. Batey, M. C. Petty, G. G. Roberts and D. R. Wight, Electronic: Lett., 1984,20,489. 4 A. Ulman, An Introduction to Ultrathin Organic Films from Langmuir-Blodgett to Self-assembly, Academic Press, San Diego, 1991. 5 M. J. Cook, in Spectroscopy of New) Materials, ed. R. J. H. Clark and R. E. Hester, Wiley, Chichester, 1993, pp.87-150. 6 e.g. M. Yoneyama, M. Sugi, M. Saito, K. Ikegami, S.-I. Kuroda and S. Iizima, Jpn. J. Appl. Phys., 1986, 25, 961; K. Ogawa, H. Yonehara, T. Shoji, S-I. Kinoshita, E. Maekawa, H. Nahahara and K. Fukuda, Thin Solid Films, 1989, 178, 439; K.Clgawa, S-I. Kinoshita, H. Yonehara, H. Nakahara and K. Fiikuda, J. Chem. SOC., Chem. Commun., 1989,477; M. Fujiki, H. Tatlei and S. Imamura, Jpn. J. Appl. Phys., 1987,26, 1224; M. Fujiki, H Tabei and T. Kurihara, Langmuir, 1988,4,1123. 7 M. J. Cook, A. J. Dunn, M. F. Daniel, R. C. 0. Hart, R. M. Richardson and S. J. Roser, Thin Solid Films, 1988,159,395. 8 N. B. McKeown, M. J. Cook, A. J. Thomson, K. J. Harrison, M. F. Daniel, R. M. Richardson and S. J. Roser, Thin Solid Films, 1988,159,469. 9 N. Dent, M. J. Grundy, R. M. Richardson, S. J. Roser, N. B. McKeown and M. J. Cook, J. Chim. Phys., 1988,85,1003. 10 M. J. Cook, M. F. Daniel, K. J. Harrison, N. B. McKeovn and A. J. Thomson, J. Chem. SOC., Chem. Commun., 1987,1148. 11 M. J. Cook, N. B. McKeown, J. M. Simmons, A. J. Thomson, M. F. Daniel, K. J. Harrison, R. M. Richardson and S. J. Roser, J. Muter. Chem., 1991, 1, 121. 12 For preliminary accounts of aspects of this study see: (a) M. A. Chesters, M. J. Cook, S. L. Gallivan, J. M. Simmons and D. A. Slater, Thin Solid Films, 1992, 210/211 538; (b)S. Mukopadhyay, A. K. Ray, M. J. Cook, J. M. Simmons and C. A. Hogarth, J. Muter. Sci., Muter. Electron., 1992,3, 139. 13 M. J. Cook, A. J. Dunn, S. D. Howe, A. J. Thomson and K. J. Harrison, J. Chem. SOC., Perkin Trans. 1, 1988,2453. 14 I. Chambrier, M. J. Cook, S. J. Cracknell and J. Mchlurdo, J. Muter. Chem., 1993,3, 841. 15 R. M. Richardson and S. J. Roser, Liquid Crystals, 1987,2, 797. 16 K. Bergesen, L. Haugland, M. J. Cook and J. McMurdo, in preparation. 17 H. Ringsdorf, B. Schlarb and J. Venzmer, Angew. Chem., lnt. Ed. Engl., 1988,113,27; A. Laschewsky, Angew. Chem., Int. Ed Engl., 1989,28,1574. 18 M. Kasha, in Spectroscopy of the Excited State, NATO Ad\. Stud. Ser. B, Physics, No 12, ed. B. D. Bartolo, Plenum Press, New York, 1976, p. 337. 19 R. J. Musgrove, Ph.D. Thesis, University of Bristol, 1991. 20 A. Zarbahksh, personal communication. 21 H. Itoh, T. Koyama, K. Hanabusa, E. Masuda, H. Shiri and T. Hayakawa, J. Chem. SOC., Dalton Trans., 1989,1543. 22 M. J. Cook, J. McMurdo and A. K. Powell, J. Chem. SOC., Chem. Commun., 1993,903. Paper 4/01269K; Received 2nd March, 1994