Polymerization DOI: 10.1002/anie.201203639 Control of a Living Radical Polymerization of Methacrylates by Light** Brett P. Fors and CraigJ. Hawker* The ability to precisely control molecular weight and molecular weight distributions, as well as gain sequence and architecture control in polymer synthesis is of considerable importance and has greatly impacted the advancement of science and technology. [1] Indeed, the development of con- trolled living polymerization methods has profoundly changed polymer research with strategies, such as nitroxide- mediated radical polymerization (NMP), [2] atom transfer radical polymerization (ATRP), [3] and reversible addition fragmentation chain transfer polymerization (RAFT), [4] allowing the facile synthesis of well-defined polymers that are diverse in both their structure and function. Recently there has been an effort to dramatically increase the scope of living radical polymerization through the development of strategies to regulate the activation and deactivation steps by using an external stimulus. [5–7] Arguably, the most successful strategy that controls both the initiation and growth steps has been the recent work of Matyjaszewski and co-workers who exploited the unique aspects of electro- chemistry to control the ratio of activator to deactivator in ATRP. [5] By selective targeting of redox-active catalytic species, the polymerization reaction could be turned “on” and “off” by adjusting parameters such as applied current, potential, and total charge passed. As with traditional radical polymerization, the most robust and widely used form of regulation is through photo- polymerization, which is a pervasive procedure in both academia and industry. [8] The ability to develop a photocon- trolled living radical polymerization would, therefore, repre- sent a significant breakthrough. Interestingly, one of the earliest attempts to develop a living radical polymerization involved iniferter polymerization using a dithiocarbamate under UV irradiation. [9] However, the procedure was intrinsi- cally limited and poor control and broad molecular weight distributions were obtained. Subsequently, photoinitiation of ATRP, [7] NMP, [10–13] and RAFT [14–17] polymerizations have been developed, though in all cases only the initiation step was photocontrolled and all subsequent growth steps could not be photoregulated. As a result, the development of a highly responsive photocontrolled living radical procedure, which affords control over the chain growth process, is both a major opportunity as well as challenge for the future of living polymerizations. The key to addressing this challenge was recent work by the research groups of Macmillan, [18] Yoon, [19] Stephenson, [20] and others [21] who have exploited the power of photoredox catalysts for organic transformations that are mediated by visible light. [22–25] We envisaged that the unique properties of these photoredox catalysts would allow for the development of a highly responsive photocontrolled living radical poly- merization. Our proposed mechanism for this process is shown in Scheme 1. The fac-[Ir(ppy) 3 ](1, Figure 1), a com- mercially available complex utilized previously by Macmillan and co-workers, has been shown to absorb visible light to afford fac-[Ir(ppy) 3 ]*. [26, 27] We anticipated that this excited Ir III * species would reduce an alkyl bromide initiator to give the desired alkyl radical, which could initiate polymerization of the monomer. The key to this process is that the highly oxidizing Ir IV complex could then react with the propagating radical to afford the initial Ir III complex in the ground state, as well as a dormant polymer chain with a bromo end group. Having regenerated the starting Ir III complex, homolysis of Scheme 1. Proposed mechanism of a visible-light-mediated living radi- cal polymerization using an Ir-based photoredox catalyst. P n = polymer chain. Figure 1. The photoredox catalyst fac-[Ir(ppy) 3 ] (ppy = 2-pyridylphenyl). [*] Dr. B. P. Fors, Prof. Dr. C. J. Hawker Materials Research Laboratory, University of California Santa Barbara, CA 93106 (USA) E-mail: [email protected][**] We thank the MRSEC program of the National Science Foundation (DMR 1121053, C.J.H.) and the Dow Chemical Company through the Dow Materials Institute at UCSB (B.P.F.) for financial support. B.P.F thanks the California NanoSystems Institute for the Elings Prize Fellowship in Experimental Science. Supporting information for this article is available on the WWW under http://dx.doi.org/10.1002/anie.201203639. . Angewandte Communications 8850 # 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim Angew. Chem. Int. Ed. 2012, 51, 8850 –8853
Transcript
PolymerizationDOI: 10.1002/anie.201203639
Control of a Living Radical Polymerization of Methacrylates byLight**Brett P. Fors and Craig J. Hawker*
The ability to precisely control molecular weight andmolecular weight distributions, as well as gain sequence andarchitecture control in polymer synthesis is of considerableimportance and has greatly impacted the advancement ofscience and technology.[1] Indeed, the development of con-trolled living polymerization methods has profoundlychanged polymer research with strategies, such as nitroxide-mediated radical polymerization (NMP),[2] atom transferradical polymerization (ATRP),[3] and reversible additionfragmentation chain transfer polymerization (RAFT),[4]
allowing the facile synthesis of well-defined polymers thatare diverse in both their structure and function.
Recently there has been an effort to dramatically increasethe scope of living radical polymerization through thedevelopment of strategies to regulate the activation anddeactivation steps by using an external stimulus.[5–7] Arguably,the most successful strategy that controls both the initiationand growth steps has been the recent work of Matyjaszewskiand co-workers who exploited the unique aspects of electro-chemistry to control the ratio of activator to deactivator inATRP.[5] By selective targeting of redox-active catalyticspecies, the polymerization reaction could be turned “on”and “off” by adjusting parameters such as applied current,potential, and total charge passed.
As with traditional radical polymerization, the mostrobust and widely used form of regulation is through photo-polymerization, which is a pervasive procedure in bothacademia and industry.[8] The ability to develop a photocon-trolled living radical polymerization would, therefore, repre-sent a significant breakthrough. Interestingly, one of theearliest attempts to develop a living radical polymerizationinvolved iniferter polymerization using a dithiocarbamateunder UV irradiation.[9] However, the procedure was intrinsi-cally limited and poor control and broad molecular weightdistributions were obtained. Subsequently, photoinitiation ofATRP,[7] NMP,[10–13] and RAFT[14–17] polymerizations havebeen developed, though in all cases only the initiation stepwas photocontrolled and all subsequent growth steps could
not be photoregulated. As a result, the development ofa highly responsive photocontrolled living radical procedure,which affords control over the chain growth process, is botha major opportunity as well as challenge for the future ofliving polymerizations.
The key to addressing this challenge was recent work bythe research groups of Macmillan,[18] Yoon,[19] Stephenson,[20]
and others[21] who have exploited the power of photoredoxcatalysts for organic transformations that are mediated byvisible light.[22–25] We envisaged that the unique properties ofthese photoredox catalysts would allow for the developmentof a highly responsive photocontrolled living radical poly-merization. Our proposed mechanism for this process isshown in Scheme 1. The fac-[Ir(ppy)3] (1, Figure 1), a com-
mercially available complex utilized previously by Macmillanand co-workers, has been shown to absorb visible light toafford fac-[Ir(ppy)3]*.[26, 27] We anticipated that this excitedIrIII* species would reduce an alkyl bromide initiator to givethe desired alkyl radical, which could initiate polymerizationof the monomer. The key to this process is that the highlyoxidizing IrIV complex could then react with the propagatingradical to afford the initial IrIII complex in the ground state, aswell as a dormant polymer chain with a bromo end group.Having regenerated the starting IrIII complex, homolysis of
Scheme 1. Proposed mechanism of a visible-light-mediated living radi-cal polymerization using an Ir-based photoredox catalyst. Pn = polymerchain.
Figure 1. The photoredox catalyst fac-[Ir(ppy)3] (ppy = 2-pyridylphenyl).
[*] Dr. B. P. Fors, Prof. Dr. C. J. HawkerMaterials Research Laboratory, University of CaliforniaSanta Barbara, CA 93106 (USA)E-mail: [email protected]
[**] We thank the MRSEC program of the National Science Foundation(DMR 1121053, C.J.H.) and the Dow Chemical Company throughthe Dow Materials Institute at UCSB (B.P.F.) for financial support.B.P.F thanks the California NanoSystems Institute for the ElingsPrize Fellowship in Experimental Science.
Supporting information for this article is available on the WWWunder http://dx.doi.org/10.1002/anie.201203639.
the C�Br bond, addition of monomer to the radical chain endand recapping with bromide can occur with the overall cyclicprocess being mediated by visible light.[28, 29] Significantly, thiswould result in a photochemically controlled, living ATRP-like process. However, the most advantageous aspect of thistype of system, and what sets it apart from traditional Cu-mediated ATRP methods, is the ability to reversibly activateor deactivate the polymerization with visible light. Specifi-cally, when light is removed from this reaction no IrIII* will bepresent, and the polymerization will rest at the dormant, andstable, bromo chain-terminated species. Further, upon reex-posure to visible light, IrIII* will be formed, reactivating thepolymerization allowing for true control over polymerstructure and architecture.
Initially, the polymerization of methyl methacrylate(MMA) was examined by using the Ir complex 1 as thecatalyst, ethyl-a-bromophenylacetate (2) as the initiator, anda 50 W fluorescent lamp as the light source. In keeping withthe simple, user friendly nature of traditional ATRP, complex1 was selected for preliminary studies owing to its stability,highly reducing nature, and commercial availability.[30]
Encouragingly, initial results employing 0.2 mol % catalystdid afford polymer; however, the reaction displayed little tono control, with a Mw/Mn value of 2.76 (Table 1, entry 1). We
reasoned that our photoredox catalyst 1 was only acting as aninitiator in this reaction, and to gain control over this systemthe radical concentrations needed to be lowered, which couldbe achieved by simply reducing the catalyst loading. Insupport of this hypothesis, it was found that significantlydecreasing the loading of 1 to 0.005 mol% led to excellentcontrol over the polymerization with a molecular weightdistribution of 1.25 being obtained (Table 1, entry 2); this lowcatalyst loading is a highly desirable feature. Further, underthe optimized reaction conditions subsequent polymeri-zations demonstrated that the molecular weight of thepolymer could be efficiently controlled by changing the
monomer/initiator ratio with close agreement between theexperimental and theoretical molecular weights (Table 1,entries 2–5).
To confirm the proposed mechanism, control polymeri-zations were conducted without added catalyst or in theabsence of light, and in both examples no reaction wasobserved (Table 1, entries 6 and 7). Additional supportingevidence was obtained through fluorescence studies.[26] When1 was combined with various concentrations of MMA, nochange in the fluorescence of the Ir complex was detected.However, when the same experiment was performed with theinitiator, ethyl-a-bromophenylacetate, a concentration-de-pendent fluorescence quenching was observed. These resultssuggest that the excited IrIII* complex is undergoing a redoxprocess with the initiator and not reacting with the monomer(see the Supporting Information for more details).
The lack of any reaction in the absence of visible lightsuggests that a true “on”–“off” living photopolymerizationsystem could be developed. To demonstrate this possibility,monomer, initiator, and catalyst were initially combined inthe absence of light and after one hour no polymerization wasobserved. The reaction was then exposed to visible light fortwo hours at room temperature, which resulted in approx-imately 15 % monomer conversion. In demonstrating truetemporal control, removal of the light source stops thepolymerization immediately and no conversion was observedduring the dark period (1 hour). Exposure to light fora second two-hour period “turns” the polymerization backon and this “on”/“off” cycle can be repeated numerous timeswithout observable reaction in the absence of irradiation(Figure 2a). These results demonstrate that this system ishighly responsive to our external stimulus and when light isremoved from the system polymerization stops almostimmediately. This high degree of temporal control illustratesthe efficient nature of the fac-[Ir(ppy)3] catalyst for reversibleactivation and deactivation of the bromo chain end.
To clearly demonstrate that existing chain ends arereactivated during these “on”/“off” cycles with no newchains being initiated during the polymerization, plots ofln([M]0/[M]t) versus total exposure time (Figure 2b), and Mn
versus conversion (Figure 2c) both gave linear relationships.Significantly, this data proves that when the light is turned offand polymerization stops, termination of the chain ends is notoccurring and in the absence of light, the dormant species isthe stable bromo chain end. In analogy with traditionalATRP, when these dormant chain ends are reexposed to lightin the presence of the iridium catalyst, efficient reactivation ofthe chain ends is achieved. These features provide compellingevidence that this process is a photocontrolled living radicalpolymerization, which is highly responsive to visible light asan external stimulus.[31]
To further probe the living nature of this system, as well asprovide additional evidence for the presence of active bromogroups at the chain ends, block copolymers were preparedusing sequential photocontrolled living radical procedures(Figure 3). Initially, irradiation of a mixture of methylmethacrylate and the initiator 2 in the presence of the iridiumphotocatalyst 1 (0.005 mol%) afforded a well-definedPMMA derivative, 3, with controlled molecular weight and
Table 1: Molecular weight and polydispersities for the visible-light-mediated polymerization of methyl methacrylate using [Ir(ppy)3] .
[a] Reaction conditions: MMA (1 equiv), 1 (0–0.2 mol%), and 2 (0.002–0.20 equiv) in DMF (0.37 mLmmol�1 of MMA) at room temperature withirradiation from a 50 W fluorescent lamp (Mn = number averagemolecular weight; Mw = weight average molecular weight). [b] Thereaction was run in the absence of visible light.
low polydispersity. Use of 3 as a macroinitiator in thepolymerization of benzyl methacrylate and in this case,0.01 mol% of 1, proved to be a well-behaved process leadingto the desired poly(methyl methacrylate)-b-(benzyl metha-crylate) diblock copolymer 4, with size exclusion chromatog-raphy showing little or no starting macroinitiator (Figure 3).This efficient block copolymer formation further illustratesthat minimal termination is occurring during the polymeri-zation process.
A major difference between the reported photocontrolledliving radical polymerization and traditional ATRP proce-dures is the stability of the catalyst, with the Ir-based system
being extremely tolerant to a variety of functional groups. Todemonstrate this added versatility, the synthesis of homo-polymers of methacrylic acid (MAA), which is notoriouslydifficult to polymerize under ATRP conditions, and randomcopolymers of MAA and benzyl methacrylate (BnMA) wereexamined. Under standard conditions and at 10 % incorpo-ration of MAA, excellent control over molecular weight anda low polydispersity were observed. Moreover, the polydis-persity increased only slightly with higher MAA content(Table 2). These polymerizations in the presence of a freecarboxylic acid exemplify the robust nature of our Ir-basedcatalyst 1. Moreover, this excellent functional-group toler-ance is an additional advantage over a traditional Cu-basedATRP process.
In summary, we have developed a new controlled livingradical polymerization that displays an unprecedentedresponse to activation and deactivation of polymerizationthrough external visible light stimulation. The advantages ofthis approach lie in its highly responsive nature, facilereaction setup, use of only ppm levels of catalyst, andexcellent functional group tolerance. In analogy with thepervasive nature of traditional photopolymerization andATRP procedures, this photocontrolled living radical poly-merization offers a versatile platform for the preparation offunctional materials with applications in sustainability, elec-tronics, and health.
Figure 2. Polymerization of MMA using catalyst 1 while cycling thereaction’s exposure to visible light. a) Conversion vs. time; b) time oflight exposure vs. ln([M]0/[M]t), with [M]0 and [M]t being the concen-trations of monomers at time points zero and t, respectively; c) con-version vs. Mn (*) and conversion vs. Mw/Mn (D).
Figure 3. Synthesis of a poly(methyl methacrylate)-b-(benzyl methacry-late) diblock copolymer. Size exclusion chromatogram with the grayand black traces corresponding to 4 and 3, respectively, is shown onthe right.
Table 2: Synthesis of random copolymers and the homopolymer ofmethacrylic acid (MAA).[a]
Entry MAA:BnMA Mn [kgmol�1] Mw/Mn
1 10:90 21 1.242 20:80 22 1.363 100:0 28 1.61
[a] Reaction conditions: MAA (0.1–1.0 equiv), BnMA (0–0.90 equiv),1 (0.005–0.13 mol%), and 2 (0.004 equiv) in DMF (0.37 mLmmol�1 ofmonomer) at room temperature with irradiation from a 50 W fluorescentlamp; for characterization the polymers were methylated after polymer-ization with TMSCHN2 (TMS= trimethylsilyl) to give the methyl ester.
[1] A. H. E. M�ller, K. Matyjaszewski, Controlled and LivingPolymerizations: Methods and Materials, Wiley-VCH, Wein-heim, 2009.
[2] C. J. Hawker, A. W. Bosman, E. Harth, Chem. Rev. 2001, 101,3661 – 3688.
[3] T. Pintauer, K. Matyjaszewski, Chem. Soc. Rev. 2008, 37, 1087 –1097.
[4] G. Moad, E. Rizzardo, S. H. Thang, Polymer 2008, 49, 1079 –1131.
[5] A. J. D. Magenau, N. C. Strandwitz, A. Gennaro, K. Matyjas-zewski, Science 2011, 332, 81 – 84.
[6] N. Bortolamei, A. A. Isse, A. J. D. Magenau, A. Gennaro, K.Matyjaszewski, Angew. Chem. 2011, 123, 11593 – 11596; Angew.Chem. Int. Ed. 2011, 50, 11391 – 11394.
[7] Y. Kwak, K. Matyjaszewski, Macromolecules 2010, 43, 5180 –5183.
[8] Y. Yagci, S. Jockusch, N. J. Turro, Macromolecules 2010, 43,6245 – 6260.
[9] T. Otsu, M. Yoshida, Makromol. Chem. Rapid Commun. 1982, 3,127 – 132.
[10] J. C. Scaiano, T. J. Connolly, N. Mohtat, C. N. Pliva, Can. J.Chem. 1997, 75, 92 – 97.
[11] A. Goto, J. C. Scaiano, L. Maretti, Photochem. Photobiol. Sci.2007, 6, 833.
[12] E. Yoshida, Colloid Polym. Sci. 2010, 288, 73 – 78.[13] Y. Guillaneuf, D. Bertin, D. Gigmes, D.-L. Versace, J. Lalev�e,
J. P. Fouassier, Macromolecules 2010, 43, 2204 – 2212.[14] M. A. Tasdelen, Y. Y. Durmaz, B. Karagoz, N. Bicak, Y. Yagci, J.
Polym. Sci. Part A 2008, 46, 3387 – 3395.
[15] S. Muthukrishnan, E. H. Pan, M. H. Stenzel, C. Barner-Kowol-lik, T. P. Davis, D. Lewis, L. Barner, Macromolecules 2007, 40,2978 – 2980.
Fabrication of Complex Three-Dimensional PolymerBrush Nanostructures through Light-Mediated LivingRadical Polymerization**Justin E. Poelma, Brett P. Fors, Gregory F. Meyers, John W. Kramer, andCraig J. Hawker*
Surface-initiated polymerizations (SIPs) have received sig-nificant attention as a robust and effective strategy for thefabrication of polymer brushes.[1, 2] These polymerizations aregenerally performed on substrates modified with a self-assembled monolayer of initiators, giving polymer chains thatare tethered to the substrate by one end. By employingcontrolled radical polymerization techniques, such as atom-transfer radical polymerization,[3] reversible addition-frag-mentation chain transfer polymerization,[4] and nitroxide-mediated polymerization,[5] brush architectures, such as blockcopolymer brushes and a variety of polymer brushes pat-terned in the x- and y-dimensions,[6] are accessible.[7] Theversatility of these synthetic routes has led to a wide range ofapplications, including antifouling coatings,[8, 9] chemical sens-ing,[10] biofunctional interfaces,[11, 12] and stimuli-responsivematerials.[13, 14]
While progress has been made toward advanced brusharchitectures, current surface initiation strategies lack tem-poral and spatial control and, therefore, rely on a prepatternedinitiator layer to template brush formation. Prepatterning hasbeen demonstrated on a variety of substrates using top-downlithographic techniques,[15] such as photo- and interferencelithography,[6] electron-beam lithography,[16–18] scanning-probe lithography,[19–21] and soft lithography.[22] In thesecases, polymerization only occurs in regions where theinitiator is present, resulting in patterned polymer brushes.[22]
With greater difficulty, the patterning of polymer brushescan be extended to three-dimensional (3D) nano- and micro-structures by patterning a concentration gradient of theinitiating species by lithographic techniques or throughcontrolled vapor deposition of initiators.[23–25] Steric interac-tions between chains cause densely packed areas to be highlyextended when compared to sparsely grafted areas, resultingin varying brush heights on the substrate. A key characteristicof these routes to patterned and gradient brushes is theuniform distribution of initiators on the substrate. Though lesscommon, gradient brushes have also been obtained froma uniform layer of the initiators by varying the chain densitythrough the time of exposure to UV light,[26, 27] or bymanipulating the contact time of the surface with themonomer and catalyst solution.[28] While these approachescan readily give rise to gradient surfaces, it is technicallychallenging to produce complex 3D structures.
Recently, our group reported the living radical polymer-ization of methacrylates regulated by visible light using an Ir-based photoredox catalyst.[29, 30] In this system the propagatingpolymer chains are efficiently returned to their dormant statewhen the light source is removed, and can be reinitiated uponsubsequent exposure to light, affording temporal control overchain growth.
Herein, we demonstrate facile, temporally and spatiallycontrolled brush formation from a uniform initiating layerthrough a visible light-mediated radical polymerization, andillustrate a set of key differentiating features of this approachcompared with previous strategies (Figure 1 a). For example,the use of light to control the polymerization allows the brushheight to be determined by the exposure time. Through theuse of a traditional photomask, brush growth can also bespatially confined to exposed regions (Figure 1b), however,
the unexposed regions still contain active initiating speciesthat can be utilized for subsequent polymerizations. In priorstudies, the initiating groups were typically destroyed in areaswhere the polymer brushes were not grown. Finally, neutraldensity filters can be used to modulate the intensity ofincident light and, therefore, the kinetic rate of polymeri-zation from the surface. These factors allow the directformation of gradient brush structures and arbitrary 3Dfeatures in a single step over large areas (Figure 1c).[31] Incombination with the uniform density of initiating groups, thismethod leads to homogeneous stretching of the polymerchains with varying molecular weights, in direct contrast tothe variable stretching in prior studies. The unique propertiesof this process, which leads to nanoscale features that aremolecularly distinct from those achieved previously, offersignificant scope for applications ranging from photolithog-raphy to one-step, high-throughput fabrication of patternedsubstrates.
To demonstrate the capabilities of this new concept,silicon oxides were uniformly functionalized with trichloro-silane-substituted a-bromoisobutyrate-based initiators (Fig-
Figure 1. Patterning of polymer brushes from substrates uniformlyfunctionalized with trichlorosilane-substituted a-bromoisobutyrate-based initiators (a) using b) a photomask for patterns or c) a neutraldensity filter for gradient structures. DMF= N,N-dimethylformamide,ppy = 2-phenylpyridine.
[*] J. E. Poelma, Dr. B. P. Fors, Prof. C. J. HawkerMaterials Research Laboratory and Materials DepartmentUniversity of CaliformiaSanta Barbara, CA 93106 (USA)E-mail: [email protected]
Dr. G. F. Meyers, Dr. J. W. KramerThe Dow Chemical CompanyMidland, MI 48667 (USA)
[**] We thank the MRSEC program of the National Science Foundation(DMR 1121053, J.E.P. and C.J.H.) and the Dow Chemical Companythrough the Dow Materials Institute at UCSB for financial support.B.P.F. thanks the California NanoSystems Institute for the ElingsPrize Fellowship in Experimental Science. J.E.P. thanks the NSFGraduate Research Fellowship for funding.
Supporting information for this article is available on the WWWunder http://dx.doi.org/10.1002/anie.201301845.
ure 1a).[7] Initially, the relationship between film thicknessand irradiation time was determined by a series of separate,yet comparable experiments, in which exposure time to thelight of a commercial 26 Watt fluorescent lamp (availablefrom any hardware store) was varied for a solution of methylmethacrylate (MMA), DMF, and fac-[Ir(ppy)3] in contactwith silicon wafers uniformly functionalized with a covalentlybound initiating species. Film thickness increased linearlywith time upon continuous irradiation (Figure 2). Signifi-
cantly, the use of light as an external mediator of polymer-ization enables the control of film thickness without theaddition of a “sacrificial” untethered initiator or a deactivat-ing species to the monomer and catalyst solution. To furtherconfirm the facile nature of this process and establish thatbrush growth only occurs when irradiated by light, a series of“on”–“off” experiments were conducted. First, a sample wasirradiated for 20 minutes, left in the dark for 10 minutes, andthen re-exposed to light for 10 minutes, resulting in a totalexposure time of 30 minutes. Similarly, a substrate was cycledbetween three dark periods of 5 minutes and two intervals of5 minutes of exposure to light (10 minutes total exposure tolight). In all cases, brush thickness was determined only by thetotal irradiation time (Figure 2, &). This ability to “pause” and“restart” surface-initiated polymerizations has profoundimplications for patterning polymer brushes and clearlydemonstrates that iridium-based photocontrolled polymeri-zation affords excellent temporal control of brush growthfrom a surface.
One of the most attractive features of a photochemicallycontrolled route to polymer brushes is the potential for directspatial control over brush growth. As a simple illustration,initiator-functionalized substrates in a solution of MMA and[Ir(ppy)3] in DMF were irradiated through photomaskscontaining rectangular patterns of different sizes. Opticalmicrographs show clear patterning of the poly(methyl meth-acrylate) (PMMA) brushes (Figure 3), thus demonstratingspatial control over brush formation from a uniform initiating
layer and the ability to pattern a range of features over largeareas. Patterning could be achieved at submicrometer levelsand was only limited by the wavelength of the light. The keyto minimizing the impact of diffusion on the resolution in thissystem is the short excited state lifetime of the [Ir(ppy)3]catalyst(ca. 50 ns). Based on an upper limit for the diffusioncoefficient of the Ir catalyst, which is the self-diffusioncoefficient of water (D = 2.3 � 10�9 m2 s�1), the catalyst isexpected to diffuse less than approximately 20 nm during itsexcited-state lifetime. This distance is significantly shorterthan the wavelength of light and leads to the high degree offidelity observed in this system.
The ability to spatially control brush formation also opensup the intriguing possibility of combining spatial (x,y dimen-sions) with intensity (z dimension) modulation to producewell-defined three-dimensional nanostructures in a singlestep. When compared to previous strategies, this representsa more practical and versatile synthetic approach. As aninitial example, a grayscale photomask that contains an arrayof squares of varying optical density was used to probe therelationship between brush height and light intensity fora given exposure time. The brush height was found to beinversely proportional to the optical density of the mask(Figure 4). Regions of the substrate that are exposed to morelight experience an increase in the kinetic rate of brushformation, resulting in higher molecular weight brushes and,as a result, an increase in polymer brush thickness. Because
Figure 2. Brush height as a function of irradiation time measured byspectral reflectance. *: continuous irradiation, &: brushes that wereobtained by “on”–“off” cycles (see text for details).
Figure 3. Optical microscopy image of patterned PMMA brushesobtained using a negative photomask with a) 20 mm by 200 mm andb) 2.5 mm by 25 mm rectangles.
Figure 4. Brush height as a function of optical density of the photo-mask, as measured by profilometry.
the initiator density is uniform over the entire substrate,variances in brush thickness are a result of variations in themolecular weights of the polymer brushes in different regions.To support this hypothesis, the grafting densities weredetermined to be consistent (0.27� 0.02 chainsnm�2) fora range of different features/samples, as well as within thesame sample, regardless of light intensity or exposure time.[26]
These grafting densities compare favorably with reportedvalues for moderately dense polymer brushes.[32]
A powerful consequence of this novel mechanism is thepossibility to fabricate complex and arbitrary three-dimen-sional patterns by modulating the intensity of light to controlthe molecular weight of the brushes rather than the density ofthe initiator. Using a grayscale lithography mask, a variety offeatures could be prepared, including inclined planes, micro-prisms, gradients, and arrays of microlenses (Figure S3). Theoptical micrograph and 3D atomic force microscopy (AFM)image of an inclined plane (Figure 5) illustrate the compellingnature of this technique for patterning 3D polymer brushstructures from a uniform initiating layer in a single step.Importantly, the linear relationship between feature heightand the optical density of the mask is maintained, asevidenced by a height profile along the length of the structure(Figure 5c).
An additional attractive feature of this strategy is thenondestructive nature of the patterning process, which allowsretention of initiator activity after initial polymer brushformation. As a result, this technique represents a facile routeto patterned block copolymer brushes in which both chemicalfunctionality and surface topography can be tuned (Fig-ure 6a). To investigate this capability, a uniform PMMAbrush with a height of approximately 40 nm was initiallyprepared. Exposure of this surface through a TEM grid asa photomask was then used to grow poly(tert-butyl meth-acrylate) (PtBuMA) chains specifically in the irradiated areas.
This technique gives regions of PMMA-b-PtBuMA where thePtBuMA domains are 12 nm thicker than the PMMAinitiating layer with these block copolymer domains clearlybeing visible by optical microscopy (Figure 6 b).
The ability to selectively initiate the fabrication ofa second block shows that the first PMMA layer still containsactive alkyl bromide chain ends, which can be efficiently re-initiated to afford spatially defined block copolymer brushes.As with the temporal control observed by cycling betweenlight and dark periods, no untethered initiator or deactivatingspecies was required to maintain control over the polymerchain ends. To illustrate the variations of surface propertiesthat can be readily obtained, PtBuMA was converted topolymethacrylic acid (PMAA) by immersion in a 1:1 mixtureof dichloromethane and trifluoroacetic acid (TFA) for30 minutes. After deprotonation with a 0.1 molar aqueousKOH solution, selective wetting of the PMAA regions wasobserved by optical microscopy (Figure 6c), further verifyingthe presence of the patterned block copolymers.[1,2, 33] Theability to pattern block copolymer brushes through sequentialpolymerization of monomers from a uniform initiating layerrepresents significant progress in the fabrication of 3Dfeatures for tuning surface properties.[3,34]
In summary, a facile approach to patterned polymerbrushes has been developed by taking advantage of thetemporal and spatial control afforded by a “living” visiblelight mediated radical polymerization. Through modulationof the light intensity, complex and arbitrary 3D structures canbe fabricated. Furthermore, patterned block copolymerstructures can be formed for tuning surface properties.
Received: March 5, 2013Revised: April 17, 2013Published online: June 3, 2013
Figure 5. a) Optical micrograph of nanoscale-inclined plane formedfrom a 3D polymer brush; b) 3D AFM image of nanoscale-inclinedplane, and c) height along hashed line across feature as shown in (a).
Figure 6. a) Schematic of patterned block copolymer brushes andconversion of PMMA-b-PtBuMA to PMMA-b-PMAA. b) Optical micro-graph of PMMA-b-PtBuMA brushes patterned from a uniform PMMAinitiating layer. c) Selective wetting of PMMA-b-PMAA regions afterexposure to water vapor.
[1] S. Edmondson, V. L. Osborne, W. T. S. Huck, Chem. Soc. Rev.2004, 33, 14 – 22.
[2] R. Barbey, L. Lavanant, D. Paripovic, N. Sch�wer, C. Sugnaux, S.Tugulu, H.-A. Klok, Chem. Rev. 2009, 109, 5437 – 5527.
[3] J. Pyun, T. Kowalewski, K. Matyjaszewski, Macromol. RapidCommun. 2003, 24, 1043 – 1059.
[4] M. Baum, W. J. Brittain, Macromolecules 2002, 35, 610 – 615.[5] M. Husseman, E. E. Malmstrçm, M. McNamara, M. Mate, D.
Mecerreyes, D. G. Benoit, J. L. Hedrick, P. Mansky, E. Huang,T. P. Russell, C. J. Hawker, Macromolecules 1999, 32, 1424 –1431.
[6] C. Schuh, S. Santer, Adv. Mater. 2009, 21, 4706 – 4710.[7] K. Matyjaszewski, P. J. Miller, N. Shukla, B. Immaraporn, A.
Gelman, B. B. Luokala, T. M. Siclovan, G. Kickelbick, T. Vallant,H. Hoffmann, Macromolecules 1999, 32, 8716 – 8724.
[8] H. Ma, J. Hyun, P. Stiller, A. Chilkoti, Adv. Mater. 2004, 16, 338 –341.
[9] J. L. Dalsin, B.-H. Hu, B. P. Lee, P. B. Messersmith, J. Am. Chem.Soc. 2003, 125, 4253 – 4258.
[10] S. Kurosawa, H. Aizawa, Z. A. Talib, B. Atthoff, J. Hilborn,Biosens. Bioelectron. 2004, 20, 1165 – 1176.
[11] R. R. Bhat, B. N. Chaney, J. Rowley, A. Liebmann-Vinson, J.Genzer, Adv. Mater. 2005, 17, 2802 – 2807.
[12] N. A. Hutter, M. Steenackers, A. Reitinger, O. A. Williams, J. A.Garrido, R. Jordan, Soft Matter 2011, 7, 4861 – 4867.
[13] M. A. C. Stuart, W. T. S. Huck, J. Genzer, M. M�ller, C. Ober, M.Stamm, G. B. Sukhorukov, I. Szleifer, V. V. Tsukruk, M. Urban,et al., Nat. Mater. 2010, 9, 101 – 113.
[14] I. Luzinov, S. Minko, V. V. Tsukruk, Soft Matter 2008, 4, 714 –725.
[15] T. Chen, I. Amin, R. Jordan, Chem. Soc. Rev. 2012, 41, 3280 –3296.
[16] S. J. Ahn, M. Kaholek, W. K. Lee, B. LaMattina, T. H. LaBean, S.Zauscher, Adv. Mater. 2004, 16, 2141 – 2145.
[17] N. Ballav, S. Schilp, M. Zharnikov, Angew. Chem. 2008, 120,1443 – 1446; Angew. Chem. Int. Ed. 2008, 47, 1421 – 1424.
[18] M. N. Khan, V. Tjong, A. Chilkoti, M. Zharnikov, Angew. Chem.2012, 124, 10449 – 10452; Angew. Chem. Int. Ed. 2012, 51, 10303 –10306.
[19] W.-K. Lee, M. Patra, P. Linse, S. Zauscher, Small 2007, 3, 63 – 66.[20] X. Zhou, X. Wang, Y. Shen, Z. Xie, Z. Zheng, Angew. Chem.
[21] X. Zhou, Z. Liu, Z. Xie, X. Liu, Z. Zheng, Small 2012, 8, 3568 –3572.
[22] F. Zhou, Z. Zheng, B. Yu, W. Liu, W. T. S. Huck, J. Am. Chem.Soc. 2006, 128, 16253 – 16258.
[23] R. R. Bhat, M. R. Tomlinson, T. Wu, J. Genzer, N. Carolina, Adv.Polym. Sci. 2006, 198, 51 – 124.
[24] S. Schilp, N. Ballav, M. Zharnikov, Angew. Chem. 2008, 120,6891 – 6894; Angew. Chem. Int. Ed. 2008, 47, 6786 – 6789.
[25] M. Steenackers, R. Jordan, Adv. Mater. 2009, 21, 2921 – 2925.[26] B. P. Harris, A. T. Metters, Macromolecules 2006, 39, 2764 – 2772.[27] A. W. Harant, V. S. Khire, M. S. Thibodaux, C. N. Bowman,
Macromolecules 2006, 39, 1461 – 1466.[28] M. R. Tomlinson, J. Genzer, Chem. Commun. 2003, 1350 – 1351.[29] B. P. Fors, C. J. Hawker, Angew. Chem. 2012, 124, 8980 – 8983;
Angew. Chem. Int. Ed. 2012, 51, 8850 – 8853.[30] F. A. Leibfarth, K. M. Mattson, B. P. Fors, H. A. Collins, C. J.
Photon Control of Liquid Motion on Reversibly PhotoresponsiveSurfaces
Dongqing Yang,† Marcin Piech,‡ Nelson S. Bell,‡ Devens Gust,§ Sean Vail,§Antonio A. Garcia,⊥ John Schneider,⊥ Choong-Do Park,§ Mark A. Hayes,§ and
S. T. Picraux*,†,#
School of Materials, Department of Chemistry and Biochemistry, and Harrington Department ofBioengineering, Arizona State UniVersity, Tempe, Arizona, 85287, Sandia National Laboratories,
Albuquerque, New Mexico, 87185, and Center for Integrated Nanotechnologies, Los Alamos NationalLaboratory, Los Alamos, New Mexico 87545
ReceiVed May 23, 2007. In Final Form: July 24, 2007
The movement of a liquid droplet on a flat surface functionalized with a photochromic azobenzene may be drivenby the irradiation of spatially distinct areas of the drop with different UV and visible light fluxes to create a gradientin the surface tension. In order to better understand and control this phenomenon, we have measured the wettingcharacteristics of these surfaces for a variety of liquids after UV and visible light irradiation. The results are used toapproximate the components of the azobenzene surface energy under UV and visible light using the van Oss-Chaudhury-Good equation. These components, in combination with liquid parameters, allow one to estimate thestrength of the surface interaction as given by the advancing contact angle for various liquids. The azobenzenemonolayers were formed on smooth air-oxidized Si surfaces through 3-aminopropylmethyldiethoxysilane linkages.The experimental advancing and receding contact angles were determined following azobenzene photoisomerizationunder visible and ultraviolet (UV) light. Reversible light-induced advancing contact-angle changes ranging from 8to 16° were observed. A large reversible change in contact angle by photoswitching of 12.4° was achieved for water.The millimeter-scale transport of 5µL droplets of certain liquids was achieved by creating a spatial gradient invisible/UV light across the droplets. A criterion for light-induced motion of droplets is shown to be consistent withthe response of a variety of liquids. The type of light-driven fluid movement observed could have applications inmicrofluidic devices.
Introduction
Microfluidic systems are of interest for a variety of potentialapplications, such as delivering analyses in lab-on-chip environ-ments, bioassays in drug discovery, and chemical analyses.1-4
These applications are driving interest in handling increasinglysmaller volumes of liquids in order to transport, store, mix, react,and analyze small amounts of liquids in miniaturized analyticalsystems. Benefits include reducing sample size, decreasing assaytime, and minimizing reagent volume. However the physics ofscaling does not permit a simple miniaturization of macroscopicpumps for liquid transport because of the increasing importanceof interfacial forces relative to inertial forces as the surface-to-volume ratio increases. Thus, alternative methods to control themotion of small liquid volumes that scale favorably withdecreasing liquid sample size are needed.
Some of the most promising new methods for manipulatingliquid droplets on surfaces involve changing the interfacialproperties of materials to control surface wetting by liquids. Thekey concept of these methods is that a gradient in surface tensioncan generate a net force for manipulation of the droplet on a
surface. Such nonmechanically induced flow arising from theaction of a surface tension gradient can be created by thermal,5-7
electrochemical,8-10 and chemical11 methods. The use of lightas a driving force offers unique opportunities.12-17 Advantagesanticipated for microfluidic systems driven by light include (i)the technology is amenable to miniaturization; (ii) no mechanicalmoving parts are needed; (iii) the use of light does not requirecontact of any additional materials with sensitive biologicalsolutions; and (iv) chemical reactions can be performed on a tinyscale without the need for containers or channels.
Azobenzene is a well-known photochromic organic moleculethat can be easily and reversibly photoisomerized from the transto the cis form by UV irradiation, and from cis to trans by visible
* To whom correspondence should be addressed. E-mail: [email protected].
† School of Materials, Arizona State University.§ Department of Chemistry and Biochemistry, Arizona State University.⊥ Harrington Department of Bioengineering, Arizona State University.‡ Sandia National Laboratories.# Los Alamos National Laboratory.(1) Yager, P.; Weigl, B. H.Science1999, 283, 346.(2) Hayes, M. A.; Polson, N. A.; Phayre, A. N.; Garcia, A. A.Anal. Chem.
(6) Sindy, K.; Tang, Y.; Mayers, B. T.; Vezenov, D. V.; Whitesides, G. M.Appl. Phys. Lett. 2006, 88, 061112.
(7) Paik, P.; Pamula, V. K.; Chakrabarty, K.2004 Inter Society Conferenceon Thermal Phenomena; IEEE: Los Alamitos, CA, 2004; pp 649-654.
(8) Whitesides, G. M.; Laibinis, P. E.Langmuir1990, 6, 87.(9) Gallardo, B. S.; Hwa, M. J.; Abbott, N. L.Langmuir1995, 11, 4209.(10) Gallardo, B. S.; Gupta, V. K.; Eagerton, F. D.; Jong, L. I.; Craig, V. S.;
Shah, R. R.; Abbott, N. L.Science1999, 283, 57.(11) Chaudhury, M. K.; Whitesides, G. M.Science1992, 256, 1539.(12) Rosario, R.; Gust, D.; Hayes, M.; Jahnke, F.; Springer, J.; Garcia, A. A.
Langmuir2002, 18, 8062.(13) Rosario, R.; Gust, D.; Garcia, A. A; Hayes, M.; Taraci, J. L.; Clement,
T.; Dailey, J. W.; Picraux, S. T.J. Phys. Chem. B2004, 108, 12640.(14) Bunker, B. C.; Kim, B. I.; Houston, J. E.; Rosario, R.; Garcia, A. A.;
Hayes, M.; Gust, D.; Picraux, S. T.Nano Lett.2003, 12, 1723.(15) Garcia, A. A.; Cherian, S.; Park, J.; Gust, D.; Jahnke, F.; Rosario, R.J.
Phys. Chem. A2000, 104, 6104.(16) Rosario, R.; Gust, D.; Hayes, M.; Springer,J.; Garcia, A. A.Langmuir
2003, 19, 8801.(17) Berna, J.; Leigh, D. A.; Lubomska, M.; Mendoza, S. M.; Perez, E. M.;
light. A solid surface modified with a photochromic azobenzeneis a promising system for the control of surface free energy withlight.18-19 The photoinduced geometry change of azobenzenebetween trans and cis isomers results in a change in the molecularshape and dipole moment, which in turn gives rise to reversibleswitching of surface wettability through intermolecular interac-tions.20-21Thus, the reversible switching behavior of azobenzeneupon exposure to light allows for optical control of surface wettingproperties, and thereby control over fluid motion.
In this paper, we describe the photoinduced wetting/dewettingof azobenzene-functionalized surfaces for various liquids. Thiswork follows our earlier studies of photoinduced movement ofwater drops on photochromic surfaces13and the light-driven liquidmotion studies by Ichimura and colleagues18-19 who formedphotoresponsive azobenzene surfaces usingO-octacarboxym-ethylated calix[4]resorcinarene (H-CRA-CM) absorbed ontofused silica surfaces with pendentp-octylazobenzene units asthe photoresponsive adsorbate. In the present work, the azoben-zene was covalently bound to a smooth, air-oxidized Si surfacethrough pretreatment with 3-aminopropylmethyldiethoxysilane(ADES) followed by the formation of an amide bond with anazobenzene carboxylic acid.
The focus of this study was to examine the role of chemistryon the surface-liquid interaction and the change in droplet contactangle following UV-visible light irradiation of the photochrome.First, we measured contact angles using a variety of liquids forthe azobenzene surface in both the cis and trans forms. Thesedata were used in conjunction with the van Oss-Chaudhury-Good equation to estimate the surface energy components of thetwo solid surface states. We then examined the criteria forphotoinduced droplet movement,11,19 which requires that thechange in contact angle upon photoisomerization exceeds thecontact-angle hysteresis. The results were used to predict whetherphotoinduced movement of liquid drops could be achieved forvarious liquids on the basis of their wetting characteristics. Thesepredictions were then tested by determining which liquid dropscould be moved experimentally with light over millimeterdistances.
1. Experimental Section
Synthesis.The synthetic route for azobenzene-acyl chloride (5)used in this study is illustrated in Figure 1 and was carried out asfollows:
4-[4′-(Pentylphenyl)azo]phenol (1).This compound was preparedaccording to the procedure previously described.22 4-Pentylaniline(20.2 g, 124 mmol) was dissolved in a mixture of concentratedhydrochloric acid (HCl) (30 mL) and distilled water (40 mL). Sodiumnitrite (8.6 g, 125 mmol) in distilled water (25 mL) was addeddropwise to the above solution at 0°C. The resulting solution ofdiazonium salt was added slowly to a mixture of phenol (11.6 g, 123mmol) and sodium carbonate (21.4 g, 258 mmol) in distilled water(75 mL) at 0°C. Pure1 was obtained in nearly 100% yield.
Isopropyl 11-Bromoundecanoate (2).This compound was preparedin a manner similar to that described for a related compound.23 Asample of 11-bromoundecanoic acid (25 g, 94 mmol) was reactedwith oxalyl chloride (18 g, 141 mmol) in dichloroethane (DCE)(150 mL) at 70°C for 30 min under N2. DCE and residual oxalylchloride were distilled under reduced pressure. Dichloromethane(90 mL), triethylamine (10.1 g, 100 mmol), and a catalytic amount
ofN,N-dimethylaminopyridine were added to the solution. Anhydrousisopropanol (11.3 g, 190 mmol) was subsequently added to themixture at 5°C. The reaction was warmed to room temperature andallowed to react for an additional 30 min. The reaction mixture wasextracted with 5-10% (wt/wt) aqueous HCl, and the liquid wasremoved under reduced pressure. Compound2 was isolated in 99%yield. 1H NMR (CDCl3, 400 MHz) 1.15 (6H, d,-CH(CH3)2), 1.21(10H, m, -CH2-(CH2)5-CH2-), 1.36 (2H, m,-CH2-CH2-CH2-), 1.55 (2H, m,-CH2-CH2-CH2-), 1.77 (2H, m,-CH2-CH2-CH2-), 2.18 (2H, t, -CH2-CO-), 3.33 (2H, t, Br-CH2-CH2-), 4.93 (1H, m,-OCH(CH3)2).
(8 g, 58 mmol), dicyclohexano-18-crown-6 catalyst (0.35 g, 1.0mmol) and dimethylformamide (DMF) (100 mL) were combined ina 500 mL flask. A portion of1 (12.6 g, 46.8 mmol) was then added,along with an additional 100 mL of DMF. The mixture was heatedto 100°C and allowed to react for 5 h under dry air. The liquid wasremoved by distillation under reduced pressure. After addingdichloromethane (200 mL) to the residue, the solution was washedwith water (200 mL) and dried over Na2SO4. Volatiles were distilledunder reduced pressure to leave pure product in 98% yield.1H NMR(CDCl3, 400 MHz) 0.88 (3H, t,CH3-CH2-) 1.20 (6H, d,-CH-(CH3)2), 1.29 (14H, m,-CH2-(CH2)7-CH2-), 1.44 (2H, m,-CH2-CH2-CH2-), 1.61 (4H, m, CH3-(CH2)2-CH2-), 1.79 (2H, m,-CH2-CH2-CH2-), 2.24 (2H, t,-CH2-CO-), 2.65 (2H, t, Ph-CH2-), 4.01 (2H, t, Ph-OCH2-), 4.98 (1H, m, -OCH(CH3)2), 6.97(2H, d, Ph-H), 7.26 (2H, d, Ph-H), 7.78 (2H, d, Ph-H), 7.88 (2H,d, Ph-H).
4-[(11-Carboxyundecyl)oxy]-4′-pentylazobenzene (4). A portionof ester3 (23 g, 46.4 mmol) was dispersed in 500 mL of 80/20vol/vol ethanol/water solution. Potassium hydroxide (KOH)(22.5 g, 400 mmol) and 2 mL of 40 wt % aqueous tetra-n-butylammonium hydroxide (n-Bu4NOH) were then added. Themixture was heated at reflux overnight, cooled to 0°C, and acidifiedto pH 2 with 6 N hydrochloric acid. The aqueous layer was thenextracted with dichloromethane and washed twice with 1 N HCl.Evaporation of the organic liquid under reduced pressure yielded18 g of product (85.5% yield), which was used without furtherpurification.1H NMR (CDCl3, 400 MHz) 0.88 (3H, t,CH3-CH2-),1.31 (14H, m,-CH2-(CH2)7-CH2-), 1.43 (2H, m,-CH2-CH2-CH2-), 1.64 (4H, m, CH3-(CH2)2-CH2-), 1.79 (2H, m,-CH2-CH2-CH2-), 2.33 (2H, t, -CH2-COOH), 2.65 (2H, t, Ph-CH2-), 4.01 (2H, t, Ph-OCH2-), 6.97 (2H, d, Ph-H), 7.28 (2H,d, Ph-H), 7.77 (2H, d, Ph-H), 7.86 (2H, d, Ph-H).
Azo-acid Chloride (5). A portion of acid4 (5 g, 11 mmol) wasreacted with oxalyl chloride (2.9 g, 22.5 mmol) in DCE (250 mL)
(18) Oh, S. K.; Nakagawa, M.; Ichimura, K.J. Mater. Chem.2002, 12, 2262.(19) Ichimura, K.; Oh, S. K.; Nakagawa, M.Science2000, 288, 1624.(20) Siewierski, L. M.; Brittain, W. J.; Petrash, S.; Foster, M. D.Langmuir
1996, 12, 5838.(21) Natansohn, A.; Rochon, P.;Chem. ReV. 2002, 102, 4139.(22) Siewierski, L. M.; Lander, L. M.; Leibmann, A.; Brittain, W. J.; Foster,
M. D. SPIE Proc.1995, 2441, 1.(23) Sekkat, A.; Wood, J.; Geerts, Y.; Knoll, W.Langmuir1996, 12, 2976.
Figure 1. Synthetic route for azobenzene-acid chloride5.
Control of Liquid Motion on PhotoresponsiVe Surfaces Langmuir, Vol. 23, No. 21, 200710865
at 70°C for 1 h under N2. The liquid and residual oxalyl chloridewere removed by distillation to give pure product in 99% yield.1HNMR (CDCl3, 400 MHz) 0.90 (3H, t,CH3-CH2-), 1.34 (14H, m,-CH2-(CH2)7-CH2-), 1.48 (2H, m,-CH2-CH2-CH2-), 1.69(4H, m, CH3-(CH2)2-CH2-), 1.82 (2H, m,-CH2-CH2-CH2-),2.67 (2H, t, Ph-CH2-), 2.88 (2H, t,-CH2-COCl), 4.04 (2H, t,Ph-OCH2-), 6.96 (2H, d, Ph-H), 7.30 (2H, d, Ph-H), 7.80 (2H,d, Ph-H), 7.89 (2H, d, Ph-H). MALDI-MS (terthiophene matrix):calculated 470.27, found 471.25 [M+ H].
Materials. Polished silicon (100)n-type 4 in. wafers were obtainedfrom WaferNet, Inc., and their surfaces were confirmed to be quitesmooth (atomic force microscopy gave root-mean-square (rms)roughnesses in the subnanometer range). The ADES was obtainedfrom Gelest, Inc. Toluene was obtained from Mallinckrodt Baker,Inc. and was double distilled prior to use. Deionized (DI) water usedfor contact-angle measurements was obtained from a BarnsteadNanopure system. All probe liquids used for contact-angle measure-ments were obtained from Alfa Aesar or Aldrich and used as received.
Substrate Monolayer Functionalization.Si (100)n-type sub-strates with an air-oxidized surface were cut into rectangular samplesof about 1.5 cm× 3 cm. These samples were cleaned with a UVozone treatment (UV ozone cleaner, model 42, Jelight, Inc.) for 30min, rinsed extensively with DI water, and dried in an oven at140 °C for 15 min. After cleaning, the substrate was immersed ina solution of 100µL ADES in 10 mL of double-distilled toluene for30 min at room temperature, washed well three times in toluene toremove excess reagent, and heated in an oven at 140°C for 3 h. Theaminosilane-coated substrate was incubated in an oven at 80°C ina toluene solution of4 (20 mM) in the presence of 4-dimethylami-nopyridine (200 mM) for 8 h. The substrate was then removed fromsolution, washed sequentially with toluene, dichloromethane, andtetrahydrofuran for 5 min, and dried with nitrogen gas. This treatmentresults in azobenzene tethered to the SiO2 surface through a siloxane(ADES) as the linker molecule (ADES-AZO) as shown schematicallyin Figure 2. Glass substrates with tethered azobenzene were alsoprepared for UV-vis absorption studies. The glass slides were cleanedin a warm sulfuric acid and hydrogen peroxide (2:1) solution for 30min, rinsed in DI water, and then cleaned in a warm hydrogen peroxideand hydrochloric acid solution (1:1) for 30 min and finally rinsedwith DI water. After this cleaning treatment, the same procedure asthat described above for functionalizing the azobenzene monolayeron the surface was used.
Physical Measurements.The UV-vis absorption spectrum ofan ADES-AZO monolayer prepared on glass is presented in Figure3. Although trans and cis azobenzene exhibit overlapping absorption,
the two forms can be readily distinguished.24 The photoirradiationof the ADES-AZO monolayer with UV (366 nm) light results in theformation of the cis isomer, with an absorption band at 445 nm. Theprominent peak at 340 nm observed following exposure of the ADES-AZO monolayer to visible light is attributed to thenfπ* electronictransition for the trans isomer. Curves 1-4 in Figure 3 illustrate thechange in the absorption bands during switching with various UVand visible light exposure times. These changes observed in theUV-vis absorption spectrum confirm the reversible photoisomer-ization of the ADES-AZO monolayer between cis and trans forms.Because both isomers absorb at both of the wavelengths used forirradiation, at equilibrium under either light the sample is presentas a photostationary distribution of the two isomers, enriched in oneform.
Contact-Angle Measurements. The dynamic contact-anglemeasurements were performed using a Rame´-Hart standardautomated goniometer, model 200-00. This equipment included theDROP image standard software that uses a curve-fitting routine forcontact-angle calculations. The dynamic advancing and recedingcontact angles were measured while gradually increasing anddecreasing the volume of liquid with the dispensing needle embeddedin the sessile drop. All measurements were carried out at roomtemperature. Figure 4 illustrates the contact-angle measurementmethod. The volume of the droplet is increased by adding liquidfrom t0 to t2 and decreased by removing liquid fromt2 to t4. As thevolume of the droplet is increased fromt0 to t1, the contact angle(Figure 4, upper curve) is seen to increase, and the droplet maintainsthe same contact area with the surface. Att1, the droplet begins toincrease in surface contact area, and, as the volume of droplet increasesfrom t1 to t2 (as indicated by the droplet width in Figure 4, middlecurve), the area continues to advance over the unwetted area whilekeeping the same maximum contact angle. Thus, the advancingcontact angle,θadv, is measured by averaging this maximum angleover a large number of measurements while increasing the liquidvolume fromt1 to t2. Fromt2 to t3, the volume of the droplet decreases,and the contact angle decreases while maintaining the same contactarea. At t3, the contact area begins to decrease with a constantminimum angle fromt3 to t4 as the volume of liquid in the dropletis decreased. The receding contact angle,θrec, is obtained by averagingthis minimum angle over a corresponding large number of measure-
Figure 2. Molecular structure of azobenzene tethered to the surfacevia ADES, illustrating the trans and cis states after visible and UVirradiation, respectively.
Figure 3. UV-visible absorbance spectra for an ADES-AZO surfaceunder a sequence of visible and UV irradiation; inset shows the 445nm peak region enlarged. Spectra were taken in the following order:curve 1 (black line) ADES-AZO surface exposed to UV light for5 min; curve 2 (red line) after visible irradiation for 5 min; curve3 (blue line) after a second UV irradiation for 5 min; curve 4 (greenline) after visible irradiation for 10 min.
10866 Langmuir, Vol. 23, No. 21, 2007 Yang et al.
ments betweent3 to t4 as liquid is removed from the droplet. Byaveraging over an large number of measurements in this way andby repeating these measurements on multiple locations (typicallyfive) on the azobenzene-functionalized surfaces for the dynamicadvancing and receding contact angles for each liquid studied, theaccuracy was improved over single contact-angle measurements ateach location. The contact-angle hysteresis is defined as the differencebetween theabove-measureddynamicadvancingand recedingcontactangles.
Photoirradiation. The UV light source was a model UVGL-25Mineralight lamp with 254 and 366 nm wavelengths. We used the366 nm light for 5 min for the UV irradiations with a flux of 2 mWcm-2 at the sample surface. The visible light source was a model190 Fiber optic illuminator from Dolan-Jenner Industries (Lawrence,MA), coupled with a 410 nm-long wave length pass filter, resultingin a flux of 10 mW cm-2 on the sample surface. An exposure timeof 5 min was also used for the visible light irradiations.
2. Results and Discussion
Photoswitchable ADES-AZO Surface.We prepared azoben-zene-functionalized monolayer surfaces on smooth air-oxidizedSi substrates in order to use the photochromic property ofazobenzene to reversibly switch between trans and cis states,thereby altering the surface energy. Under visible light, azoben-zene is present mainly as thetransisomer, and the drops exhibita higher contact angle; under UV light, azobenzene is presentmainly as the cis isomer, and drops display a lower contact angle.Figure 5 shows reversible water wettability after visible andUV irradiation on the ADES-AZO surface. The water contact-angle changes from 100-102° under visible light to 88-89°under UV light. The reversible photoinduced advancing contact-angle change was therefore 11-13°. This observed contact-
angle switching is approximately 50% greater than values notedin earlier studies. Oh et al. used anO-octacarboxymethylatedcalyx resorcinarene with pendentp-octylazobenzene units as aphotoresponsive adsorbate (H-CRA-CM), and observed contact-angle switching of 8° using visible and UV illumination.18
Siewierski and co-workers prepared monolayer assemblies onsilicate substrates by direct deposition of alkyltriethoxysilanesfollowed by covalent attachment of azobenzene derivatives tothe functionalized layers. Their contact-angle switching rangedfrom 4° to 9°.20The larger contact-angle switching values reportedhere suggest that the present method of tethering the azobenzeneto the surface allows for greater interaction between the liquidand azobenzene moieties while maintaining sufficient free volumefor photoisomerization.
Contact Angles of Different Liquids on ADES-AZOSurfaces.To investigate the dependence of liquid-solid interac-tions on liquid properties, we measured the advancing (θadv) andreceding (θrec) contact angles for various liquids on theazobenzene-modified surface following both visible and UV (366nm) irradiation. The liquids were diiodomethane, 1-bromonaph-thalene, 1-methylnaphthalene, DMF, acetonitrile, benzonitrile,formamide,andethyleneglycol.Weprepared twoor threesamplesfor each liquid and measured five different contact areas for eachsample. Contact-angle results for each liquid after visible/UVirradiation of the surface are summarized in Table 1.
Surface Energy of Azobenzene.The contact angle on theazobenzene-modified surfaces increases approximately linearlywith the surface tension of the probe liquids after both UV andvisible irradiation, as shown in Figure 6. These trends suggestthat results for the various probe liquids can provide insight intothe energetics of the azobenzene-fluid interface and, conse-quently, the fluid transfer properties at this interface.
The surface free energy is an important parameter that playsa major role in a variety of interfacial physicochemical processes.Although there are several methods for estimating surface
Figure 4. The dynamic advancing and receding contact-anglemeasurement method is illustrated for water, with the top curve (a)being a plot of the contact angle and the middle curve (b) being thewidth of the sessile drop. These values are continuously measuredas a function of time while adding liquid (timet0 to t2) or removingliquid (time t2 to t4) from the drop. The advancing contact angle isdetermined by the average value betweent1 andt2, where the contactangle has reached it maximum value and the contact area of the dropis continuously growing in size. The receding contact angle is givenby the average value betweent3 andt4, where the contact angle hasreached its minimum value and the contact area of the drop iscontinuously decreasing in size.
Figure 5. (a) Reversible advancing contact angle for water asmeasured on the ADES-AZO surface, showing reversible switchingas a function of sequential visible and UV light irradiation. (b) Side-view photograph of the water droplet after irradiation with visiblelight. (c) Side view photograph of the water droplet after irradiationwith UV light. A comparison of panels b and c illustrates the decreasein advancing contact angle with UV irradiation.
Control of Liquid Motion on PhotoresponsiVe Surfaces Langmuir, Vol. 23, No. 21, 200710867
energies, the Lifshitz-van der Waals/acid-base (LWAB)approach suggested by van Oss and co-workers26,27,29 isstraightforward and has proven to be of fairly general applicabil-ity.28,30According to the van Oss theory, the surface free energyγS can be expressed by eq 1:
whereγSLW is the nonpolar Lifshitz-van der Waals contribution,
and γSAB is the polar electron-donor/electron-acceptor (Lewis
acid-base) component of the energy. Equation 2 states thatγSAB
can be represented in terms of electron-accepting (γS+) and
electron-donating (γS-) contributions. Using this approach, the
work of adhesionWA for a liquid/solid system can be expressedas
where the subscripts S and L denote solid and liquid, respectively,andθ is the contact angle.
Equation3 (thevanOss-Chaudhury-Goodequation)providesa simple way to calculate the solid surface free energy fromcontact-angle measurements using three probe liquids, two ofwhich are polar and one of which is nonpolar. By measuringcontact angles for three different liquids and then solving threesimultaneous equations, we can calculate the three unknownsolid parameters for substitution into eqs 2 and 1. The liquidparameters,γL
LW, γL-, andγL
+, are available from the literature.The total surface energyγS for the azobenzene surface was
obtained by solving eqs 1-3. The three components ofγS weredetermined using eq 3 from the advancing contact angles of thefollowing probe liquids: water, ethylene glycol, and formamideas polar liquids, and diiodomethane and 1-bromonaphthalene asnonpolar liquids. Table 2 lists the literature values of the totalsurface energy and components for each probe liquid studied.We used four different combinations of three liquids to estimatethe surface energy components. Water was employed in eachcalculation in order to achieve the largest range of liquidparameters and because of its obvious importance for possibleapplications. The resulting surface free-energy components areshown in Table 3 and Figure 7. The total surface energy (γS)
(26) van Oss, C. J.Interfacial Forces in Aqueous Media; Marcel Dekker:New York, 1994.
(27) van Oss, C. J.Colloids Surf., A1993, 78, 1.(28) Weisberg, D. S.; Dworkin, M.Appl. EnViron. Microbiol. 1983, 45, 1338.(29) van Oss, C. J.; Giese, R. F.; Wu, W.J. Adhes.1997, 63, 71.(30) Jurak, M.; Chibowski, E.Langmuir2006, 22, 7226.
Table 1. Contact Angles on an ADES-AZO Surface for Various Liquidsa
contact angle (deg): trans contact angle (deg): cis
a The measured standard deviations are reported in Table 1 of the Supporting Information.b The light-induced contact-angle change,∆θs, is definedas the difference between the advancing contact angles for visible irradiation minus that for UV irradiation:∆θs ) θadv
trans - θadvcis . c The hysteresis
for the trans state (after visible light irradiation),∆θh, is given by∆θhtrans) θadv
trans- θrectrans. d Reference 25.e Reference 26.f Reference 27.g Reference
28.
Figure 6. Advancing contact angle measured on the ADES-AZOsurface vs surface tension for the various probe liquids studied after(a) visible and (b) UV light irradiation. The liquids as numbered inthe plots are (1) DI water, (2) formamide, (3) diiodomethane, (4)ethylene glycol, (5) 1-bromonaphthalene, (6) benzonitrile, (7)1-methylnaphthalene, (8) DMF, and (9) acetonitrile. Lines are linearbest fits with slopes of 1.89°/mN/m for panel a and 1.93°/mN/m forpanel b.
values obtained from different fluid combinations agree reason-ably well, given the limitations of the approach. The averagetotal surface energy after UV light exposure (predominantly cissurface) is 40 mJ/m2, whereas the average value after visiblelight irradiation (predominantly trans surface) is 35 mJ/m2. Wenote that, in this treatment, the trans state of the surface showsthe most variation for the different liquid combinations, with thecombinations containing the nonpolar probe liquid 1-bro-monaphthalene giving higher values than that for the combinationsusing the nonpolar liquid diiodomethane.
Chemical Nature of the Surface.The van Oss treatmentshows that the surface energy is dominated byγS
LW. The acid andbase component contributes on average less than 4% of the totalenergy. Thus, the interactions of this surface with liquids aremainly via weak induced dipole-induced dipole dispersioninteractions, dipole-induced dipole forces, and dipole-dipoleforces rather than specific Lewis acid-base or hydrogen bondinginteractions. This conclusion is supported by the fact that theaverageγS for the surface enriched in thetrans-azobenzene (35mJ/m2), which has a net dipole moment near zero,31 is somewhatless than that for the surface enriched in the cis isomer (40 mJ/m2). Azobenzene in the cis form has a dipole moment of 3.1 D.31
Additional information concerning the nature of the interactionsbetween the azobenzene surface and the probe liquids can begleaned from Figure 6, which shows the advancing contact angleas a function of the liquid surface tension. Liquids that are bothhydrogen bond donors and acceptors lie on or above the least-squares lines, while those lacking this property lie on or belowthe lines. In fact, if the hydrogen-bonding solvents water, ethyleneglycol, and formamide are excluded, the slopes of the lines are
much less. The contrast is easily seen by comparing ethyleneglycol with diiodomethane. Although these have comparableliquid surface tensions, the contact angles for ethylene glycol are∼20° larger than those for diiodomethane. Ethylene glycol (andthe other hydrogen bonding solvents) can self-associate viahydrogen-bonding networks, giving a reasonably high surfacetension. Hydrogen bonding to the azobenzene surface is muchweaker, and the contact angle is thus higher. The energetic costof spreading on the azobenzene surface is not overcome byattractive interactions with the surface. Diiodomethane molecules,on the other hand, do not hydrogen bond strongly with one anotherto form extended networks, and this liquid owes its surface tensionmore to dispersion and dipole-dipole forces. Thus, di-iodomethane interacts much more strongly with the azobenzenesurface than does ethylene glycol. The free energy is loweredfor diiodomethane spreading due to the creation of dipole-dipole and induced-dipole interactions with the azobenzenesurface.
Light-Driven Movement of Liquids. By introducing asufficiently large spatial gradient in the surface tension acrossa liquid droplet, the drop can be made to move on a surface. Thismotion due to a light-induced wettability gradient across thedroplet requires that the forces overcome droplet pinning on thesmooth surface as characterized by the contact-angle hysteresis,which is the difference between advancing and receding contactangles. Figure 8 illustrates top and cross section views of a liquiddrop placed on a surface with one-half of the droplet irradiatedwith UV light and the other half with visible light to create aspatial gradient in the surface free energy. The unbalancedYoung’s force for a section of the drop of width dx is given by11
whereγSV andγSL are the surface free energies of the solid-vapor and solid-liquid interfaces, respectively, after UV or visibleirradiation of the surface. Ifθvis and θUV represent the localcontact angles at the edges on the two sides of the droplet, theneq 7 can be represented as
with γLV being the surface free energy of the liquid-vaporinterface. For a drop to move, the force on the drop induced bythe gradient in the surface tension represented by the difference(31) Yager, K. G.; Barrett, C. J.J. Photochem. Photobiol. A2006, 182, 250.
Table 3. Surface Free-Energy Components (mJ/m2) of theADES-AZO Surface Obtained from Eqs 1-3
a Liquid combinations used for determination ofγS: (a) DI water,ethylene glycol, diiodomethane; (b) DI water, formamide, diiodomethane;(c) DI water, ethylene glycol, 1-bromonaphthalene; (d) DI water,formamide, 1-bromonaphthalene.
Figure 7. Total surface tension of an ADES-AZO surface aftervisible and UV irradiation calculated by the LWAB method usingthe following three-liquid combinations: (a) DI water, ethyleneglycol, diiodomethane; (b) DI water, formamide, diiodomethane;(c) DI water, ethylene glycol, 1-bromonaphthalene; and (d) DI water,formamide, 1-bromonaphthalene.
Figure 8. Schematic of UV irradiation on the right side of a dropletand visible irradiation on the left side to induce a gradient in thesurface tension, which can lead to light-driven motion in the directionof the arrow on the ADES-AZO substrate: (a) top view; (b) crosssection view.
dFY ) [(γSV - γSL)UV - (γSV - γSL)vis]dx (7)
dFy ) γLV(cosθadv - cosθrec)dx (8)
Control of Liquid Motion on PhotoresponsiVe Surfaces Langmuir, Vol. 23, No. 21, 200710869
in the contact angles must exceed the hysteresis in the contactangles. Specifically, for the drop to be pulled in the direction ofthe UV irradiation, the advancing angleθadv
UV in the cis state ofthe ADES-AZO surface must be lower than the receding angleθrec
vis in the trans state. Without considering hydrodynamicforces,32the approximate net forceFyon the drop can be expressedby the difference in the advancing and receding contact angleson the two sides of the drop and by integrating eq 8 over thewidth of the drop,w, giving for the case of Figure 8:
From eq 9, for example, the initial light-induced force on a1 mm wide benzonitrile droplet irradiated by UV light on oneside and visible light on the other would be 2.3µN based on themeasured values forθadv
UV andθrecvis from Table 1. The force is in
the direction of the smaller contact angle.Thus, for the motion of a liquid drop to be directed by light
on a photoswitchable surface, a minimum criterion11,19 is thatthe receding contact angle (trans isomerθrec
trans) under visible lightmust be larger than the advancing contact angle (cis isomerθadv
cis )under UV light. That is, the light-induced contact-angle changes(∆θs ) θadv
trans - θadvcis ) should be greater than contact-angle
hysteresis in the trans state (∆θh ) θadvtrans- θrec
trans). We define theparameterK ) θrec
trans - θadvcis ) ∆θs - ∆θh such thatK > 0
satisfies this criterion. By spatially controlling the UV/visibleirradiation, we anticipate that, for liquids withK > 0 on theADES-AZO surface, it may be possible to move droplets towardthe UV light direction.
To test this criterion and identify the liquids that can be movedon the smooth azobenzene-modified surfaces, we measuredadvancing and receding contact angles before and after UV lightirradiation, as was shown in Table 1. We illustrate these resultsin Figure 9 for the two contrasting cases of a photoinducedswitching angle∆θsbeing greater than and less than the hysteresis∆θh as defined above for (a) benzonitrile and (b) formamide,respectively. The measured advancing contact angles after visibleand UV irradiation for three cycles are shown, and the switchingangle and hysteresis (given in Table 1) for the two cases areindicated to the right of the figure. For benzonitrile,θrec
trans )26.2°, which is greater thanθadv
cis ) 17.0°, and thus this casefulfills the requirement for motion. As seen in Figure 9, theswitching angle for benzonitrile is large compared to thehysteresis, whereas, for formamide, the opposite is the case.Following the above criterion, light-induced motion should bepossible for the case of benzonitrile droplets on our ADES-AZOsurfaces, but not for the case of formamide droplets.
To investigate the light-guided movement of a droplet ofbenzonitrile, we placed a 5µL droplet on the photoresponsiveazobenzene-modified surface that had been set to the trans statewith visible light, and focused UV light on one edge of thedroplet to generate a gradient in the surface photoisomerization.With UV light irradiation, the contact angle of the droplet frontedge under UV light irradiation decreased, and the drop spreadat the UV-illuminated edge while the back edge of the dropletinitially did not move. After 30 s, the droplet undergoes sufficientelongation that the back edge began to move, after which transportof the entire droplet during of the UV irradiation occurred, as
predicted from the results of Figure 9a. This shape and positionof the droplet during light-driven transport is shown in Figure10 in a sequence of still photographs taken in side view for thebenzonitrile droplet illuminated on the left side with UV light.In contrast, similar irradiation for the case of formamide dropsdid not result in droplet movement, as expected from Figure 9b.
To further test this criterion we irradiated one side of a 5µLdrop with UV light to test for droplet motion for all the liquidslisted in Table 1 on ADES-AZO surfaces that had been set tothe trans state with visible light while imaging the drops in sideview. A summary of the photoinduced switching angle and thehysteresis for all the liquids studied along with the results fordrop movement is given in Figure 11, where “y” indicates thatthe droplet moves, and “n” means that it remains pinned.Benzonitrile, 1-bromonaphthalene, DMF, and 1-methylnaph-thalene fulfill the requirement ofK > 0 and were all found tomove by spatial irradiation of one side of the drop with UV light.Likewise, DI water, formamide, and ethylene glycol, for whichK < 0, did not move. For 1-methylnaphthalene, which had the
(32) Shankar, R. S.; Moumen, N.; McLaughlin, J. B.Langmuir 2005, 21,11844.
(33) Reichardt, C.SolVents and SolVent Effects in Organic Chemistry, 2nd ed.;Basel: Weinheim, Germany, 1988.
(34) Abboud, J.-L. M.; Notario, R.Pure Appl. Chem.1999, 71, 645.(35) Kalugin, O. N.; Lebed, A. V., Vyunnik, I. N.J. Chem. Soc. Faraday
Trans.1998, 94, 2103.
Fy ) wγLV(cosθadvUV - cosθrec
vis) (9)
Figure 9. Contact angle of (a) benzonitrile and (b) formamide liquidson an ADES-AZO surface after visible/UV light irradiation. Thefilled squares (and dashed lines) show the advancing contact angleafter visible and UV irradiation, and the dot-dashed line shows thereceding contact angle under visible irradiation.∆θh is the hysteresisof the ADES-AZO surface after visible irradiation, and is definedas the difference between the advancing and receding contact anglesin the trans state.∆θs is the switching angle induced by visible/UVlight illumination, and is defined as the difference between theadvancing contact angles after visible and UV light irradiation. Thisfigure illustrates the criterion that∆θs > ∆θh for a light-inducedgradient in surface tension to move droplets, as is found forbenzonitrile (a) but not for formamide (b).
10870 Langmuir, Vol. 23, No. 21, 2007 Yang et al.
smallest positiveK value (0.2°), droplets moved relatively smalldistances before becoming pinned, presumably because of somenonuniformities on the surface. Also, for the case of di-iodomethane, which had a very small but positiveK value (1.6°),droplet motion was not consistently achieved. Measurementswere not carried out for acetonitrile because of the rapidevaporation of the drop due to its high vapor pressure.Summarizing our results, it was possible to achieve light-inducedmotion of drops for liquids that had measuredK > 0, whereasall liquids with K < 0 were pinned. Additional images of themotion of droplets for 1-bromonaphthalene and DMF along witha tabulation of our measuredK values are shown in the SupportingInformation as Table 2.
Several correlations were explored between the strength ofthe driving force for moving liquid droplets with light gradientsand the solvent parameters for the liquids studied here. While,as anticipated for the complex case of a surface tension gradientinducing droplet motion, no strong correlations were observed,there were several trends. As shown in Figure 12a, liquids withlower values of surface tension tend to be easier to move. Thisis consistent with the generally lower hysteresis in the contactangle in the trans state also found here for liquids with lowersurface tensions (see Table 1 and the Supporting Information),and indicative that drops are more easily moved on a surfacewhen less energy is required to overcome the liquid surfacetension. Also, as shown in Figure 12b, the magnitude of the
optically induced switching angle tends to increase with the dipolemoment of the liquid. This switching of the contact angle providesthe driving force needed to overcome the hysteresis and movedroplets on a photochromic surface. The liquids ethylene glycol,formamide, and DMF appear as exceptions to this trend, pointingto the importance of the dipole-dipole interactions between the
Figure 10. Time sequence of side view photographs (a-f) of abenzonitrile droplet on an ADES-AZO surface, illustrating light-driven transport by UV light. The UV illumination is on the left edgeof the droplet, and the droplet is shown in a series of still framesas it moves to the left.
Figure 11. UV-visible contact-angle switching∆θsand the contact-angle hysteresisθh under visible light measured for different liquidson the ADES-AZO surface. Those liquids that were found to moveunder UV light irradiation on one side are labeled “y”, and thosethat did not are labeled “n”. The liquids are ordered from left to rightin decreasingK (K ) ∆θs - ∆θh) value where droplet motion isexpected for K> 0. The liquids as numbered on thex axis are (1)DI water, (2) formamide, (3) diiodomethane, (4) ethylene glycol, (5)1-bromonaphthalene, (6) benzonitrile, (7) 1-methylnaphthalene, (8)DMF, and (9) acetonitrile. For diiodomethane (marked by an asterisk),the droplet could not be consistently moved. Droplet motion wasnot measured for acetonitrile because of the rapid evaporation ofthis high vapor pressure liquid.
Figure 12. (a) K parameter values obtained from contact-anglemeasurements on the azobenzene-functionalized surface vs thesurface tension of the probe liquids. PositiveK values are requiredto satisfy the criterion for droplet motion. (b) Switching angle forUV-visible irradiation of the probed liquids as a function of theirdipole moment on the azobenzene-functionalized surface. Liquidsare (1) DI water,33(2) formamide,33(3) diiodomethane,34(4) ethyleneglycol,35 (6) benzonitrile,34 (7) 1-methylnaphthalene,34 (8) DMF,34
and (9) acetonitrile.34
Control of Liquid Motion on PhotoresponsiVe Surfaces Langmuir, Vol. 23, No. 21, 200710871
solvent liquid and azobenzene moiety relative to any hydrogen-bond-forming interactions, since the azobenzene undergoes adramatic change in dipole moment from approximately 0 to 3D upon switching from trans to cis states.
To further illustrate the potential for droplet control inmicrofluidics,wedemonstrated theoptically controlledcombiningof droplets for benzonitrile. In Figure 13 two drops of benzonitrileare shown in a series of still images for which UV lightsimultaneously irradiated the region between and at the edges
of two drops. The UV light irradiation resulted in the dropletsmoving together and mixing within 1 min. Light-driven liquidmotion thus provides an alternative nonmechanical method tomanipulate small volumes of liquids on surfaces withoutcontainers or channels.
3. ConclusionsThe present work demonstrates the use of a light-induced
gradient in the surface tension to move liquids on smooth surfacesfunctionalized with a photoresponsive monolayer. A syntheticsurface modification approach was developed to functionalizesubstrates with a photochromic azobenzene derivative, whichexhibited large and reversible changes in wetting angle after UVand visible light irradiation. The dynamic advancing and recedingcontact angles of droplets on this surface were measured for avariety of liquids for the surface optically switched into the cisand trans states. The surface free energy of these azobenzene-modified surfaces in the two optically switched states wascharacterized by the van Oss theory. Liquid droplet motion inthe direction of the UV light was achieved by irradiation witha UV-visible optical gradient across the drop. The driving forcefor the optically directed motion of liquids was related to themeasured advancing and receding contact angles, and the motionof liquid droplets on the photochromic surface was demonstratedand characterized for a variety of liquids. Good agreement wasachieved in comparisons to a contact-angle criterion for movingdroplets and the observation of droplet motion. These resultsprovide additional understanding of the ability to control dropletmotion with light for various solvents and have potentialapplicability to microfluidic systems for delivering analyses inlab-on-a-chip environments.
Acknowledgment. This work was supported in part by theU.S. Department of Energy, Sandia National Laboratories(Contract DE-AC04-94AL85000) and the Center for Nanotech-nologies at Los Alamos National Laboratory (Contract W-7405-ENG-36), and the National Science Foundation (DMR-0413523and CHE-0352599). Support was also provided by the Inter-disciplinary Network of Emerging Science and Technology(INEST).
Supporting Information Available: The error values for thedynamic advancing and receding contact angles, the measuredK values,and still photographs illustrating light-induced motion for 1-bromonaph-thalene and DMF. This information is available free of charge via theInternet at http://pubs.acs.org.
LA701507R
Figure 13. Time sequence of side-view photographs (a-f) showingUV light driving two droplets of benzonitrile on an ADES-AZOsurface to move toward each other and combine. The UV illuminationis in the middle, extending to the two interior edges of the droplets.
