1: \documentclass[aps,pre,amssymb,twocolumn,showpacs]{revtex4}
2: %\documentclass[aps,pre,preprint,groupedaddress,showpacs]{revtex4}
3: %*************************************
4: \usepackage{graphicx}
5: \usepackage{epsfig}
6: \usepackage{subfigure}
7: \usepackage{graphics}
8: \usepackage{setspace,bm}
9: %**************************************
10: \begin{document}
11:
12: \renewcommand{\bottomfraction}{0.8}
13: \setcounter{bottomnumber}{3}
14:
15: \bibliographystyle{aip}
16:
17: \title{Infrared actuation in aligned polymer-nanotube composites}
18: %
19: \author{S. V.~Ahir, A.M.~Squires, A.R.~Tajbakhsh and E.M.~Terentjev}
20: \affiliation{Cavendish Laboratory, University of Cambridge, J.J.
21: Thomson Avenue, Cambridge CB3 OHE, U.K. }
22:
23: \begin{abstract}
24: \noindent Rubber composites containing multi-walled carbon
25: nanotubes have been irradiated with near infrared light to study
26: their reversible photo-mechanical actuation response. We
27: demonstrate that the actuation is reproducible across differing
28: polymer systems. The response is directly related to the degree of
29: uniaxial alignment of the nanotubes in the matrix, contracting the
30: samples along the alignment axis. The actuation stroke depends on
31: the specific polymer being tested, however, the general response
32: is universal for all composites tested. We conduct a detailed
33: study of tube alignment induced by stress and propose a model for
34: the reversible actuation behavior, based on the orientational
35: averaging of the local response. The single phenomenological
36: parameter of this model describes the response of an individual
37: tube to adsorption of low-energy photons; its experimentally
38: determined value may suggest some ideas about such a response.
39: \end{abstract}
40:
41: \pacs{42.70.Gi, 71.35.Gg, 73.22.Lp, 81.07.-b, 82.35.Np}
42:
43: \maketitle
44:
45: \section{Introduction}
46:
47: Many structures are able to change their mechanical properties and
48: dimensions when an appropriate stimulus is applied. This
49: phenomenon is commonly called actuation. The energy from an
50: external source triggers changes in the internal state of the
51: system, leading to a mechanical response much larger than the
52: initial input. This ability to unlock internal work in a solid
53: state structure is of key importance for many actuator
54: applications. Actuators with differing characteristics and
55: mechanisms have been widely adopted by industry to fill a variety
56: of technological requirements~\cite{Huber1997} with some having a
57: one-way response, while others providing an equilibrium,
58: reversible response to the given stimulus. Shape-memory
59: alloys~\cite{refAlloys} or polymers~\cite{Langer01} are good
60: examples of such smart actuating systems. However, in most cases a
61: shape memory system works only in one direction, requiring a reset
62: after the actuation. Only very few systems can reversibly actuate
63: and then return back to the equilibrium shape once the stimulus is
64: removed. So far only liquid crystal elastomers~\cite{Eugenebook}
65: have proven to be a truly equilibrium reversible actuating system.
66:
67: A polymer benign to external stimulus can also be made to actuate
68: when blended with of one or more distinctly different materials to
69: impart a new physical response leading to the actuation process. A
70: recent article has demonstrated one such system, based on a common
71: silicon rubber filled with a low concentration of aligned carbon
72: nanotubes, actuating in response to infrared
73: radiation~\cite{Ahir2005}. Apart from actuation itself, the
74: stimulation of functionalized nanotubes by infrared (IR) radiation
75: is also proving an effective technique, e.g. in biomedical
76: applications~\cite{Dai2005}. Clearly, there are rich prospects and
77: much motivation to understand nanotube action and the actuation
78: behavior under IR irradiation when they are embedded in a polymer
79: matrix.
80:
81: The work presented here focuses on the use of multi-walled carbon
82: nanotubes (MWCNTs) to impart equilibrium mechanical actuation in
83: the rubbery matrix. The properties of multi-walled nanotubes has
84: been well documented for over a
85: decade~\cite{Ajayan1997,Forro2001,PhysicalNanotubesBook}. Their
86: behavior in polymer composites is less well understood but some
87: reviews have recently appeared in the
88: literature~\cite{Breuer2004,Andrews2004,Harris2004,Thostenson2001,Nalwa2005}.
89: For mechanical applications, the interface between the tube
90: surface and the host polymer is of critical importance and most of
91: the studies have focussed on this aspect. In contrast, the nature
92: of the active response of nanotubes within a polymeric matrix has
93: yet to be fully understood. The complex behavior of tubes is often
94: simplified and analogies are made with aligned rigid rods. It is
95: unclear whether such analogies are always valid, especially when
96: the tubes do not necessarily form rigid rods in a polymer matrix
97: and certainly do not align unless an external field is
98: present~\cite{Nalwa2005}.
99:
100: The actuating properties of MWCNTs have recently being elucidated
101: upon with the possibility of designing nanoelectromechanical
102: (NEMS) systems~\cite{Cumings2000}. The actuator properties of
103: individual bending MWCNTs under an applied electric field have
104: been studied experimentally~\cite{Poncharal1999}. The torsional
105: actuation behavior of multi-walled tubes has also been
106: reported~\cite{Willams2002,Fennimore2003}. These works are
107: important but we note that all these studies focus on individual
108: tubes and not a collection of tubes, nor their properties within a
109: continuous elastic matrix. The massive elastic response of
110: single-walled nanotube bundles, when stimulated by light, was very
111: effectively demonstrated by Zhang and
112: Iijima~\cite{Zhang1999ElasticLight}, although little work has
113: followed from their discovery. They showed the bundles responding
114: to visible light and a near IR laser radiation by elastically
115: changing their dimensions; examining the figures in
116: \cite{Zhang1999ElasticLight} we deduce that the induced strain
117: must be about 20\%. In the context of this paper, we shall refer
118: to the actuation stroke as the change in strain, when an external
119: stimulus is applied.
120:
121: There are several reports of actuation behavior of
122: polymer-nanotube
123: composites~\cite{Landi2002,Koerner2004,Tahhan2003}. These works
124: have focussed on accentuating the already present features of the
125: host matrix by adding nanotubes. The tubes act to exaggerate the
126: response by either improving electromechanical properties or
127: increasing heat transfer efficiency due to the inherent high
128: conductivity~\cite{Naciri2003,{Nalwa2005}} that originates from
129: their delocalized $\pi$-bonded skeleton. Recent work has departed
130: from this traditional `improvement' scheme and asked whether it is
131: possible to blend nanotubes with benign polymers to create new
132: composite actuator properties, that otherwise would not occur in
133: that system. Such effects have been observed by Courty~\textit{et
134: al.}~\cite{Courty2003} where electric field stimulation of liquid
135: crystal elastomers with embedded MWCNTs lead to mechanical
136: contraction.
137:
138: Similarly, the photomechanical response from MWCNTs when embedded
139: in a silicone rubber (PDMS) matrix~\cite{Ahir2005} is a new
140: effect. The pristine elastomer shows no response to near IR
141: radiation, yet the presence of nanotubes causes a strong
142: reversible response that can be tailored by manipulating the
143: degree of alignment the tubes experience. The present work expands
144: on such a simple polymer nanocomposite system and goes on to show
145: that the effect can exist independently of the host polymer matrix
146: which, by the presence of MWCNTs, produces a mechanical response
147: to the IR irradiation. We show that both a compression and an
148: extension response can be achieved (depending on the external
149: uniaxial strain applied to the composite sample), but that the
150: magnitude of the actuation stroke strongly depends on the host
151: polymer used. We also develop a simple model that considers the
152: orientational ordering of nanotubes in the matrix along with their
153: individual and bulk actuating behavior.
154:
155: This paper is organized as following: after giving details of
156: preparation and basic composite characterization, we concentrate
157: on the analysis of tube orientation induced by stretching of the
158: host polymer matrix, section III. We then turn to the
159: IR-stimulated actuation, section IV, and study different
160: nanocomposite systems in some detail (although the majority of our
161: studies remain on the PDMS system). Section V presents a simple
162: theoretical model that might well describe the actuation mechanism
163: and compares it with our experimental data and the literature. We
164: conclude that two-way actuation behavior is dependent on nanotube
165: orientation, but is independent of the chosen homogenous polymer
166: matrix and can occur in any rubbery solid, albeit with varying
167: magnitude. It is thought that no other materials of any class
168: (metal, polymer, ceramic) can display this behavior and to such
169: large effect, thus, the study of the underlying physics of such
170: systems is of clear scientific, medical and commercial importance.
171:
172: \section{Experimental}
173: \subsection{Materials}
174:
175: There are many different sources of carbon nanotubes on the market
176: today. After extensive searching and testing, we have settled on
177: nanotubes provided by Nanostructured \& Amorphous Materials, Inc.
178: (USA). These are multi-walled, with the core diameter between
179: 5-10nm, outer diameter of 60-100nm and length between 5-15
180: microns. Purity has been verified (with SEM) as $>$95\% in raw
181: form from the supplier, in agreement with specification. These
182: nanotubes were not surface-modified at any time during processing
183: and are used throughout this study for all polymers tested.
184: Chemical functionalization is necessary in many nanocomposite
185: fields, but in our work it has been avoided to reduce the number
186: of variables in the system. We share the views of other authors
187: that chemical functionalization of the tube walls will degrade the
188: properties of the tubes overall due to further introduction of
189: sp$^3$ hybridized carbon defects~\cite{Garg1998,Sinnott2002}.
190:
191: Three types of polymer have been tested; PDMS rubber (crosslinked
192: polydimethylsiloxane), SIS (styrene-isoprene-styrene) triblock
193: thermoplastic elastomer and a nematic liquid crystal elastomer
194: (LCE, in both mono and polydomain form). Each type of polymer has
195: a unique preparation method outlined in the following sections.
196: Where possible, similarities in processing have been kept.
197: Table~\ref{tableMats} lists the composites made and their
198: abbreviations.
199: \begin{table}[b]
200: \caption{List of host polymer materials, nanotube loading and the
201: abbreviations of resulting composites.}
202: \label{tableMats}
203: \begin{center}
204: \begin{tabular}{|c|p{0.65in}|p{1.3in}|}
205: \hline
206: \textbf{Host} & \textbf{Tube loading (wt\%)} & \textbf{Abbreviation} \\
207: \hline\hline
208: PDMS & 0, 0.02, 0.3, 0.5, 1, 2, 3, 4 \& 7 & PDMS, PDMS0.02,
209: PDMS0.3, PDMS0.5, PDMS1 .. \& PDMS7 \\
210: \hline
211: Mono LCE & 0 \& 0.2 & MLCE \& MLCE0.2 \\
212: \hline
213: Poly LCE & 0 \& 0.15 & PLCE \& PLCE0.15 \\
214: \hline
215: SIS & 0.01 & SIS0.01\\
216: \hline
217: \end{tabular}
218: \end{center}
219: \end{table}
220:
221: \subsubsection{PDMS composite preparation}
222:
223: The PDMS (Sylgard 184$^{TM}$) silicone elastomer system was
224: obtained from Dow Corning, USA, in the form of the main compound
225: and the hydrosilane curing agent (crosslinker). In pristine
226: conditions, the mixing and crosslinking procedure gives a uniform
227: solvent-free elastomer. We have verified (with SEM on
228: cryo-microtomed and freeze-fractured surfaces) that the resulting
229: polymer network is pure crosslinked PDMS with no filler particles,
230: as sometimes is the case with supplied elastomer mixes.
231:
232: The nanotube-polymer composite was fabricated by first carefully
233: weighing the desired quantity of nanotubes and the polymer
234: compound. Calculations of weight percentage take into account the
235: weight of crosslinker, to be later used in the mixture. The highly
236: viscous fluid was sheared using an Ika Labortechnik mixer for a
237: minimum of 24 hours.
238:
239: The crosslinker was added to the mixture after 24 hours. The ratio
240: of crosslinker to PDMS was 1:10, according to Sylgard 184$^{TM}$
241: specification, ensuring negligible sol fraction after preparation
242: of the pristine network. The sample was then further sheared for
243: another 30 seconds before being placed in vacuum for 5 minutes to
244: degas, at all times remaining at ambient temperature to ensure
245: little crosslinking reaction takes place in this time. After
246: removing the air cavities, that unavoidably form during shear
247: mixing, the mixture was deposited in a specially designed reactor
248: (centrifuge compartment with PTFE film lining its inner wall) and
249: placed in a centrifuge at 5000rpm and 80$^{\circ}$C. At this
250: temperature the PDMS crosslinking is much faster and the
251: centrifugation achieves the uniform thickness and full
252: homogenization of resulting rubber composite samples.
253:
254: The subsequent processing depends on the target sample properties.
255: If we require a completely non-aligned nanotube dispersion, the
256: sample remains in the reactor for 24 hours, resulting in a
257: homogeneous elastomer composite. In some cases (as will be clear
258: from the text below) we aim to produce a sample with nanotubes
259: permanently pre-aligned. In this case the initial mix remains in
260: the reactor, at 80$^{\circ}$C, for 14 minutes (calculated from
261: separate measurements of crosslinking reaction rates). The
262: partially crosslinked network was then removed from the reactor
263: and aligned mechanically by applying uniaxial extension using
264: specially designed clamps. Removing the sample from the reactor
265: after what is a relatively short period of time ensures that it is
266: being mechanically aligned while still having over 50\% of
267: crosslinking to take place. Finally, while still constrained in
268: the clamps, the sample was placed in an oven at 70$^{\circ}$C for
269: a further 24 hours as it finished its crosslinking cycle under
270: stress. As a result a homogeneous elastomer was prepared where the
271: nanotubes had a preferred orientation induced by the processing
272: technique and are also well dispersed in the matrix. The degree
273: of nanotube alignment in each sample was quantified using X-ray
274: techniques (discussed below).
275:
276: There is a separate question of solvent and shearing conditions,
277: and the time required for the full MWCNT dispersion; systematic
278: studies of nanotube dispersion and re-aggregation rates are to be
279: published shortly. The quality of nanotube dispersion is monitored
280: throughout the processing with the use, initially, of optical
281: microscopes and later with a High-Resolution Scanning Electron
282: Microscope (HRSEM, Phillips XL 30 series) as aggregate sizes
283: reduce below optical resolution. We find that a shearing regime of
284: high-viscosity mixture, lasting 24 hours, is suitable in removing
285: nanotube aggregates. Samples are identified by the wt\% of MWCNTs
286: mixed with the PDMS and the abbreviations assigned to them in
287: table~\ref{tableMats}. Most experiments have been conducted on the
288: 0, 0.02, 0.3, 0.5, 1, 2, 3, 4 and 7wt\% MWCNTs in PDMS elastomer
289: films. A sample with 3wt\% carbon black instead of nanotubes has
290: also been made using the same procedure.
291:
292: \subsubsection{Nematic elastomer composite preparation}
293:
294: There is a wealth of literature regarding liquid crystal elastomer
295: (LCE) preparation~\cite{Eugenebook}. For our purposes, we have
296: tested two specific types of LCE: polydomain and monodomain, with
297: uniaxially aligned nematic director. Control samples containing no
298: nanotubes were made, following the procedure introduced
299: in~\cite{Kupfer1991LCE} and widely used in the field since. The
300: procedure of nanocomposite preparation was detailed
301: in~\cite{Courty2003}. The polysiloxane backbone chains ($\sim 60$
302: monomer units long) had their Si$-$H bonds reacted, using platinic
303: acid catalyst, with the terminal vinyl groups of the mesogenic
304: rod-like molecule $4-$methoxyphenyl) $-4'-$buteneoxy benzoate
305: (MBB) and the two-functional crosslinker
306: $1$,$4-$di-$11-$undeceneoxy benzene (11UB), with the molar ratio
307: 18:1 (thus achieving the 9:1 ratio of substituted groups on each
308: chain, or the effective 10\% crosslinking density). The
309: crosslinking was initiated by a combination of adding the catalyst
310: and heating to 80${}^{\rm o}$C in the already described
311: centrifugation reaction chamber. The subsequent procedure of
312: two-stage crosslinking, with intermediate stretching to induce
313: director alignment, is similar to the procedure of PDMS alignment
314: above.
315:
316: Polydomain control samples were made identically with the single
317: exception that no uniaxial extension is applied during the
318: crosslinking cycle. This avoids orientational bias being
319: introduced during processing.
320:
321: For LCE nanocomposites, a minor modification is made. Before the
322: crosslinker and catalyst were added, MWCNTs were shear mixed into
323: the polymer to ensure homogenous dispersion. Due to the
324: sensitivity of the crosslinker and catalyst, shear mixing is
325: reduced to 4 hours at elevated temperatures ($\sim$50$^{\circ}C$.
326: This is acceptable as the nanotube concentration in such systems
327: was very small (0.15-0.2wt\%) while the nematic polymer is highly
328: viscous. A higher concentration in such a system is currently
329: unachievable due to catalyst and crosslinker sensitivity
330: limitations.
331:
332: \subsubsection{SIS composite preparation}
333:
334: The SIS nanocomposite was made by adding the desired quantity of
335: tubes (0.01wt\%) to the melt of SIS symmetric triblock copolymer
336: (14\% of polystyrene, obtained from Sigma Aldrich) in the presence
337: of small amount of toluene solvent. The solvent dilutes the
338: otherwise rubbery thermoplastic system and allows shear mixing at
339: 40$^{\circ}C$ for 24 hours. The solvent was added in small
340: portions during the mixing cycle to maintain the mixture in a
341: high-viscosity state. Once the dispersed state was achieved,
342: fibers could be drawn from the mixture and left to air dry. During
343: this period the PS micelles are formed in the usual way
344: \cite{SISbook} to form the elastic network surrounding and
345: encapsulating the nanotubes. We note that too high a loading of
346: MWCNT prevents physical crosslinks from occurring in the host
347: polymer and thus nanotube content was kept to a very low level.
348:
349: In all cases the sample dimensions were kept approximately
350: constant, 1.5mm $\times$ 3cm, with thickness 0.2mm.
351:
352: \subsection{Experimental techniques} \label{expt}
353:
354: The main part of this study, and purpose of this paper, is
355: concerned with the response of these materials to infrared
356: radiation and to that end a specially constructed rig was built to
357: test the actuator response. Two dynamometers were used in this
358: study; a 25g dynamometer for small sensitive measurements and a
359: larger 55g dynamometer allowing a larger range of responses to be
360: tested. The dynamometers (Pioden Systems Ltd) were housed in a
361: custom made thermal-control box with an open front end. The
362: device, together with an independent thermocouple, outputs data
363: via a DAQ card to a PC, see Fig.~\ref{appa}. The sample (S) was
364: clamped in the frame with its length controlled by the micrometer
365: (M), with $\pm$0.001mm accuracy, and the exerted force measured by
366: the dynamometer (D). Thermocouples ($T_1$ and $T_2$) were placed
367: in front and behind, on the sample surface. The actuation was
368: induced by the light source (IR), Schott KL1500 LCD, with quoted
369: peak power density at $\approx$675nm, 702 $\mu$W/cm$^{2}$ at 1m
370: distance. The source uniformly illuminated the sample from $\sim
371: 2$cm distance. Measuring the scaling of the intensity decay with
372: distance, we obtained that the power density delivered to the
373: sample was $\sim 1.5 \hbox{mW/cm}^2$ at 675nm. The rig was
374: enclosed in the thermally controlled compartment, and calibrated
375: with weights to give a direct measure of stress and strain.
376:
377: Figure~\ref{Absorbance} shows the spectral distribution of the
378: light source, as well as the nanotube absorbance. These
379: measurements were carried out on a Varian Cary 300 BIO UV-visible
380: spectrophotometer in the 190-1000nm range, adjusted for the
381: background. Absorbtion units {\sf Au}$=\log [I_0/I_{\rm
382: transmitted}]$ indicate that the PDMS control sample of given
383: thickness transmits $\sim 70$\% of light across the spectrum. In
384: contrast, the same thickness of low-loading PDMS0.3 composite
385: absorbs $> 97$\% of light across a range of wavelengths. The
386: strong absorbtion of light by nanotubes is a well-known effect,
387: although the relatively flat spectral distribution was a surprise
388: in our case, Fig.~\ref{Absorbance}.
389:
390: \begin{figure}%[h]
391: \centering \resizebox{0.22\textwidth}{!}{\includegraphics{fig1}}
392: \caption{Scheme of the apparatus; see text for detail. }
393: \label{appa}
394: \end{figure}
395:
396: \begin{figure}%[h]
397: \centering \resizebox{0.35\textwidth}{!}{\includegraphics{fig2}}
398: \caption{Spectral data of the light source (left axis, arb.
399: units), and the normalized absorbtion of the PDMS0.3 composite and
400: the control pristine PDMS elastomer (right axis). }
401: \label{Absorbance}
402: \end{figure}
403:
404: To standardize the results across all samples, pre-experimental
405: checks were undertaken to accurately find the zero strain value of
406: each experiment. The gradient of the stress-strain curve for a
407: buckled sample was equated with the gradient for the stress-strain
408: curve of the taut sample -- the meeting point of the two lines
409: designates the zero-point strain, with the length of sample
410: defined as $L_0$. The imposed extensional strain is calculated by
411: $\varepsilon= (L-L_0)/L_0$, with $L$ provided from the micrometer
412: reading.
413:
414: After a fixed pre-strain was applied to each sample, the stress
415: was allowed to relax for a minimum of 10 minutes. After this
416: relaxation period, readings of stress were taken for 1-2 minutes,
417: to verify that the material is equilibrated, and then the IR
418: source was switched on to full intensity. After a period of
419: exposure, the light source was switched off and further relaxation
420: data collected. After completion, the sample was relaxed and then
421: this protocol was repeated for a different applied pre-strain
422: $\varepsilon$. Each sample is tested under a range of applied
423: pre-strains between 2\% and 40\% ($0.02 \leq \varepsilon \leq
424: 0.4)$. In order to avoid a systematic influence of pre-strain,
425: through thermal history and possible degradation, we applied the
426: different values of $\varepsilon$ in random order, not
427: sequentially. The LCE composites have been tested for even larger
428: deformation as they can spontaneously undergo thermal strains of
429: hundreds of percent~\cite{Eugenebook}.
430:
431: Our attempts to rationalize the observed response, changing
432: qualitatively on increasing the applied pre-strain, invoke the
433: concept of increasing nanotube alignment under uniaxial
434: deformation. To monitor this, wide angle X-ray diffraction
435: measurements were carried out on a Phillips PW1830 Wide Angle
436: x-ray generator (WAXS) using Cu$_{K_{\alpha1}}$ radiation (1.54$
437: \AA $), running at 40kV and 40mA. A specially designed clamp was
438: used allowing measurement of the X-ray images as a function of the
439: applied strain during the experiment. Azimuthal scans of intensity
440: were generated, Fig.~\ref{xray}, and fitted with theoretical
441: models. The method of background correction employed is crucial.
442: Two to three different regions were selected from an image to
443: gather an average of background noise which is then subtracted
444: from the azimuthal curves generated. This is repeated for all
445: scattering images before order parameter was calculated.
446:
447: \begin{figure}%[h]
448: \centering \resizebox{0.44\textwidth}{!}{\includegraphics{fig3}}
449: \caption{(a) \ The X-ray scattering image showing key reflections;
450: the outer ring (3.4${\AA}$) is the signal from the multiwall
451: nanotubes. The inner ring (7.5${\AA}$) represents the PDMS mesh
452: size, see section~\ref{xrayData}. The arrow shows the direction of
453: the uniaxial aligning strain. \ (b) \ The typical azimuthal
454: intensity variation, $I(\beta)$, at a scattering angle of
455: 3.4${\AA}$ reflection. The data is fitted by the model
456: \cite{Deutsch1991}.} \label{xray}
457: \end{figure}
458:
459: With the IR irradiation, the question always exists, whether the
460: response is due to photon absorption, or the trivial heating of
461: the materials (which does take place during irradiation). The
462: technique used to measure temperature involves two thermocouples
463: and we were reasonably sure that the measured increase in
464: temperature ($\sim$15-20${}^{\circ}$C) is a true temperature
465: across the sample. A separate study was conducted using
466: thermocouples on the surface and embedded within the sample which
467: showed similar values throughout for any relevant time-scale. The
468: samples were kept purposely thin to ensure very quick heat
469: conduction. To compare the effects, the same experiment was
470: carried out on the PDMS1 sample, with the infrared source replaced
471: by a mica-insulated heater (Minco Products Inc.) mounted
472: approximately 10mm away from the sample. Temperature was regulated
473: through an integrating controller using thermocouples mounted on
474: the sample. The maximum temperature reached was 15-20$^\circ$
475: above ambient, and although thermo-mechanical response was
476: present, it was much slower and almost an order of magnitude
477: smaller than the direct IR-irradiation effect.
478:
479: \section{Material characterization} \label{mat-sec}
480: \subsection{Elastic strength}
481:
482: Figure~\ref{modu} shows a summary of the linear mechanical
483: response of our nanocomposites for different nanotube loadings in
484: the crosslinked PDMS matrix. As the concentration of MWCNTs is
485: increased the rubbery network becomes stiffer and Young modulus
486: $Y$ (the response to static linear extension) of the composites
487: increases. This is expected and in line with literature
488: findings~\cite{Gorga2004MechAlign,Harris2004}. An account for
489: subtle variations in measured moduli could be obtained from the
490: analysis of the polymer-nanotube interface and relaxation of local
491: stress in the composites. This is not the focus of the work
492: presented here.
493:
494: \begin{figure} %[h]
495: \centering \resizebox{0.3\textwidth}{!}{\includegraphics{fig4}}
496: \caption{Young modulus $Y$ for PDMS nanocomposites at increasing
497: MWCNT loading. The arrow points at the value for control PDMS
498: rubber. } \label{modu}
499: \end{figure}
500:
501: At very low nanotube loading one might expect that large regions
502: of rubbery network are still pristine. However, even with the
503: lowest nanotube loading ($\sim$0.02wt\%), a small but significant
504: increase in modulus was observed suggesting that the presence of
505: the tubes even in tiny quantities has an immediate mechanical
506: effect. Starting from $\sim$0.02wt\% the linear increase of the
507: modulus was observed, characteristic of non-interacting inclusions
508: in the elastic matrix. We observe an almost three-fold linear
509: increase in the elastic modulus from 0-7wt\%.
510:
511: It can be argued that at higher loading the concentration
512: dependence must become non-linear, quadratic at first indicating
513: the pair interactions between nanotube inclusions, etc. This may
514: signify the onset of a `mechanical' percolation within the
515: composite system. One may be tempted to make a connection between
516: the onset of this non-linear regime and the separately determined
517: electric percolation threshold, when the composite becomes
518: conducting through nanotube contacts. Unfortunately, it is
519: difficult to make an unambiguous connection in a crosslinked
520: system: an increase in tube concentration would undoubtedly
521: increase the modulus $Y$ -- but would also cause a reduction in
522: the crosslinker concentration (the presence of nanotubes has an
523: inhibitive effect on siloxane reactions). Overall, the Young's
524: modulus of such a nanocomposite would not be able to directly
525: reflect the nanotube interactions.
526:
527: For completeness, let us quote the measured Young modulus values
528: for the other nanocomposite systems under study -- MLCE0.2:
529: $Y\approx 0.2$~MPa, PLCE0.15: $Y\approx 0.2$~MPa, SIS0.01:
530: $Y\approx 0.6$~MPa.
531:
532:
533: \subsection{Nanotube orientation by strain}\label{xrayData}
534: \subsubsection{X-ray data analysis}
535: As a crucial part of
536: material characterization, before and during the main actuation
537: experiment, we need a more quantitative analysis of the nanotube
538: orientation in the matrix. It is a key element in our model of the
539: actuation mechanism, but also has its own merit considering the
540: high interest in all aspects of polymer nanocomposite studies.
541: Wide angle X-ray diffraction is used as a method to determine the
542: average tube orientation as a function of increasing applied
543: uniaxial strain. Figure~\ref{xray}(a) shows characteristic
544: features of the diffraction image. The image is for PDMS7,
545: initially non-aligned, stretched by $\varepsilon = 0.33$ (33\%).
546: The scattering reflection at an angle corresponding to MWCNT [002]
547: layer periodicity (inter-shell spacing~\cite{Charlier1993}) of
548: 3.4$\AA$ allows calculation of the tube orientation distribution
549: from the corresponding azimuthal intensity variation,
550: Fig.~\ref{xray}(b).
551:
552: In this separate study of deformation-induced alignment we used a
553: 7wt\% loaded PDMS7 composite simply to enhance the X-ray contrast
554: and enable using a desktop X-ray generator (as opposed to the
555: synchrotron study required for very low-loading composites). Note
556: that the scattering intensity at 3.4$\AA$ is still relatively low,
557: because of the small contrast between the nanotubes and PDMS
558: matrix.
559:
560: A question must arise about the bright scattering ring
561: corresponding to the length scale $\sim 7.5\AA$. This is a very
562: interesting feature, but totally irrelevant for our work: this
563: scattering is exactly the same in the pristine PDMS rubber
564: prepared in the same batch. In the PDMS network, with no solvent,
565: the only X-ray contrast may arise due to the difference between
566: the chains and crosslinks. A very clear scattering length must be
567: an indication of crosslink density fluctuations (in other
568: terminology called clustering). As the extensive theory of this
569: clustering phenomenon suggests~\cite{rabin}, at the given chain
570: lengths and crosslinking density the network is well below the
571: `crosslink saturation threshold' and the correlation length of
572: clustering should be of the order of network mesh size. The length
573: scale of $\sim 7.5\AA$ is very accurately this size and,
574: accordingly, we believe this scattering to be produced by very
575: small scale crosslink density fluctuations. These should not
576: affect macroscopic properties, or even the local MWCNT embedding.
577:
578: Intensity variation along the azimuthal arcs, $I(\beta)$ in
579: Fig.~\ref{xray}(b), is the signature of the orientational
580: distribution function. When $I(\beta)$ is approximated as a
581: Legendre polynomial series in $\cos \beta$, it gives a measure of
582: the orientational order parameter S$_d$:
583: \begin{equation}\label{orientation}
584: S_d \equiv \langle P_2 \rangle
585: ={\textstyle{\frac{3}{2}}}(\langle\cos^2\beta \rangle-1),
586: \end{equation}
587: where the averaging is performed with $I(\beta)$ as the
588: distribution function. This is called the Herman's orientation
589: parameter and it adequately describes the true orientational
590: ordering at very small bias, when $S_d \ll 1$.
591:
592: At higher degree of alignment (such as, for instance, in nematic
593: liquid crystals) the orientational distribution function
594: significantly deviates from the measured $I(\beta)$. The analytic
595: treatment of the problem of X-ray scattering from orientationally
596: biased medium is developed by Deutsch~\cite{Deutsch1991}, mainly
597: in the context of nematic liquid crystals. Instead of using the
598: full theory, we have derived an interpolating analytical
599: approximation to the complete results of~\cite{Deutsch1991}. With
600: that, the orientational order parameter is given by
601: \begin{eqnarray}\label{DeutschOrientation}
602: S_d &=& 1-\frac{3}{2N} \int_{0}^{\pi/2}I(\beta) \sin \beta \bigg\{
603: \sin\beta \\
604: && \qquad \qquad + \cos^{2}\beta \ln
605: \left[\frac{1+\sin\beta}{\cos\beta} \right] \bigg\} d\beta
606: \nonumber
607: \end{eqnarray}
608: $${\rm with} \qquad N=\int_{0}^{\pi/2}I(\beta)d\beta .$$
609: This expression also properly accounts for nontrivial geometric
610: factors involved in projecting the 3D orientational distribution
611: onto a 2D detector plane. Experimental data was analyzed using
612: both Herman's approximation and Deutsch's interpolated analytic
613: result. We conclude that in the range of parameters we are working
614: with both expressions were in agreement qualitatively but slightly
615: differ quantitatively. We favor the Deutsch analytical method and
616: used it exclusively in this study.
617:
618: \subsubsection{Induced orientation of nanotubes} \label{orie}
619:
620: Figure~\ref{order} presents the results of the calculation of
621: orientational order parameter $S_d$, acquired as a function of
622: sample strain applied to the PDMS7 sample, as well as the
623: prediction of the theoretical model discussed below. As the
624: applied strain is increased, the initially disordered nanotubes
625: align along the strain axis resulting in bias in the azimuthal
626: curve $I(\beta)$. This phenomenon has recently been confirmed by
627: synchrotron experiments~\cite{Kelarakis2005} although it should be
628: noted that the focus of the work by Kelarakis \textit{et al.} was
629: not on nanotube reorientation in a rubbery matrix. Our composites,
630: with no significant initial alignment, on subsequent stretching
631: reached substantial values of induced orientational order.
632: Furthermore, the change in orientation on stretching was
633: reversible, i.e. equilibrium, which is discussed later. To our
634: knowledge, this is the first time nanotube \emph{reorientation}
635: has been reported and analyzed in a semisolid/rubbery sample.
636:
637: \begin{figure}%[h]
638: \centering \resizebox{0.32\textwidth}{!}{\includegraphics{fig5}}
639: \caption{The change in the orientational order parameter $S_d$ of
640: nanotubes in PDMS7 composite, as a function of imposed uniaxial
641: strain, obtained from the X-ray scattering data ($\circ$ - data
642: points; dashed line is a guide to the eye). Solid line shows the
643: affine rigid rod model prediction. } \label{order}
644: \end{figure}
645:
646: As will be described in section~\ref{ir-sec}, there is good
647: evidence that much better nanotube alignment can be achieved if
648: dispersed in a monodomain liquid crystal elastomer during
649: processing -- the mesogenic moieties act to align the tubes. A
650: similar effect has been demonstrated for pure liquid
651: crystals~\cite{Lynch02,Dierking2004LC}, and also is well-known in
652: the field of ferronematics~\cite{ferronematic}. X-ray diffraction
653: of such a system is not reported due to the continuing problem of
654: poor contrast between the two species and only low nanotube
655: concentrations studied.
656:
657: There is an issue, well argued in the
658: literature~\cite{Somoza2001,Islam2004}, about whether a truly
659: isotropic nanotube dispersion can be obtained. Regarding the tubes
660: as rigid rods with extremely high aspect ratio, well dispersed in
661: an amorphous medium, the Onsager transition to the steric
662: orientational ordering could start at very low concentrations as
663: has been recently reported~\cite{Song2005}. We have as yet
664: observed no clear indication of truly nematic liquid crystalline
665: architecture in our system, although this could be due to a number
666: of factors including matrix viscosity and sample preparation.
667:
668:
669: \subsubsection{Affine model of induced orientation} \label{affQ}
670:
671: Let us compare the observed induced orientational order parameter
672: $Q(\varepsilon)$ with a simple model prediction based on the
673: affine deformation of the rubbery matrix. The most straightforward
674: approach is to evaluate the average orientational bias resulting
675: from an imposed uniaxial extension of such a matrix, in which the
676: ensemble of rigid rods is initially embedded isotropically. The
677: direction, known as the director $\bm{n}$, is the average axis
678: along which nanotubes can and do align. This is a local property
679: of the system obtained as a result of averaging of individual
680: particle axes, $\bm{u}_i$, over the macroscopically infinitesimal
681: volume. This averaging applies equally well for rigid rod-like
682: particles and for the segments of semi-flexible chains, e.g. in
683: the study of nematic polymers \cite{Eugenebook}. The corresponding
684: local orientational order parameter is a second-rank tensor
685: $Q_{\alpha \beta}$ which for the uniaxial alignment (reflecting
686: the quadrupolar symmetry breaking) is defined as:
687: \begin{equation}\label{orientation0}
688: Q_{\alpha \beta} \equiv {\textstyle{\frac{3}{2}}} Q (n_\alpha
689: n_\beta - {\textstyle{\frac{1}{3}}} \delta_{\alpha \beta} )
690: \equiv \left(
691: \begin{array}{ccc} -\frac{1}{2}Q & 0 & 0 \cr 0 &-\frac{1}{2}Q & 0
692: \cr 0 & 0 & Q
693: \end{array} \right),
694: \end{equation}
695: where the principal axes are aligned with $z$ along the uniform
696: ordering direction $\bm{n}$, cf. Fig.~\ref{defo}. The value of the
697: local scalar order parameter is indeed the average of the second
698: Legendre polynomial of orientation of embedded rods,
699: \begin{equation}
700: Q \equiv S_d = \int_{0}^{\pi} [{\textstyle{\frac{3}{2}}} \cos^2
701: \theta -{\textstyle{\frac{1}{2}}}] \, P(\theta) \, \sin\theta
702: d\theta d\varphi . \label{defQ}
703: \end{equation}
704: Here $(\bm{n}\cdot \bm{u}_i) \equiv \cos \theta_i$ for each rod,
705: and $ P(\theta)$ is the orientational probability distribution,
706: normalized such that $\int P(\theta) \sin \theta d\theta d\varphi
707: = 1$. Let us assume the initial state is un-aligned, and thus
708: characterized by the flat distribution $P_0(\theta) = 1/(4\pi)$.
709:
710: \begin{figure} %[h]
711: \centering \resizebox{0.3\textwidth}{!}{\includegraphics{fig6}}
712: \caption{The scheme of an affine incompressible extension,
713: changing the orientation of an inflexible rod embedded in the
714: medium.} \label{defo}
715: \end{figure}
716:
717: The uniaxial extension of an incompressible elastic body is
718: described by the matrix of strain tensor
719: \begin{equation}
720: \bm{\Lambda} = \left(
721: \begin{array}{ccc} 1/\sqrt{\lambda} & 0 &
722: 0 \cr 0 & 1/\sqrt{\lambda} & 0 \cr 0 & 0 & \lambda
723: \end{array} \right) , \label{lambda}
724: \end{equation}
725: where the axis of stretching is taken as $z$ and the magnitude of
726: stretching is $\lambda = 1+\varepsilon \equiv L/L_0$ is the ratio
727: of the stretched and the initial sample length along $z$,
728: Fig.~\ref{defo}. This tensor describes the affine change of shape,
729: which could also be visualized as locally transforming an embedded
730: sphere (representing the orientational distribution $P_0$) into
731: the ellipsoid (representing the induced orientational bias) of the
732: same volume and the aspect ratio $R_\| / R_\bot = \lambda^{3/2}$.
733:
734: After such a deformation, every element of length in the body
735: changes affinely according to the matrix product
736: $\bm{L}'=\bm{\Lambda}\cdot \bm{L}$, which in our case of uniaxial
737: incompressible extension means that $L_z'=\lambda L_z$ and
738: $L_\bot'= (1/\sqrt{\lambda}) L_\bot$. This corresponds to the new
739: angle of the rod, $\theta'$ such that $\tan \theta' =
740: L_\bot'/L_z'=(1/\lambda^{3/2}) \tan \theta$. Therefore, to obtain
741: the new (now biased) orientational distribution function we need
742: to convert the variable $\theta$ into the new (current) variable
743: $\theta'$, which gives (after some algebraic manipulation)
744: \begin{eqnarray}
745: \theta
746: & \rightarrow & \arctan (\lambda^{3/2} \tan \theta'); \nonumber \\
747: \sin \theta \, d\theta & \rightarrow & \frac{\lambda^3}{(\cos^2
748: \theta' + \lambda^3 \sin^2 \theta')^{3/2}}\sin \theta' d\theta' .
749: \end{eqnarray}
750: This defines the expression for the normalized orientational
751: distribution function
752: \begin{equation}
753: P(\theta')=\frac{\lambda^3}{4\pi (\cos^2 \theta' + \lambda^3
754: \sin^2 \theta')^{3/2}} , \label{ptheta}
755: \end{equation}
756: which is an explicit function of the uniaxial strain applied to
757: the body and can be used to calculate the induced order parameter
758: $Q$:
759: \begin{equation}
760: Q(\varepsilon) = \frac{3}{2} \int \frac{\cos^2 \theta'
761: [1+\varepsilon]^3 \sin \theta' d\theta'd\varphi}{4\pi (\cos^2
762: \theta' + [1+\varepsilon]^3 \sin^2 \theta')^{3/2}} - \frac{1}{2}.
763: \label{orderaffine}
764: \end{equation}
765: Analytical integration of this expression gives a function
766: $Q(\varepsilon)$, which is plotted as a solid line in
767: Fig.~\ref{order}:
768: \begin{eqnarray}
769: Q(\varepsilon) &=& \frac{3+2\varepsilon
770: (3+3\varepsilon+\varepsilon^2)}{2\varepsilon
771: (3+3\varepsilon+\varepsilon^2)} \label{orderfull}\nonumber \\
772: && + \frac{3(1+2\varepsilon
773: (3+3\varepsilon+\varepsilon^2))^3}{4\varepsilon
774: (3+3\varepsilon+\varepsilon^2)\sqrt{1-(1+\varepsilon)^3}}\ln
775: B(\varepsilon)\nonumber
776: \end{eqnarray}
777: $ \rm{where} \ B(\varepsilon)=\left[
778: \frac{-1+(1+\varepsilon)^3+\sqrt{1-(1+\varepsilon)^3}}
779: {1-(1+\varepsilon)^3+\sqrt{1-(1+\varepsilon)^3}} \right].$
780:
781: \noindent At relatively small strains, it approaches the linear
782: regime: $Q \approx \frac{3}{5}\varepsilon - \frac{6}{35}
783: \varepsilon^2+ ...$.
784:
785: The experimental data displays a lower order parameter than that
786: predicted by the affine model, although has the same qualitative
787: trend. One must remember that the model presented here does not
788: account for tube flexibility. Also, some proportion of the tubes
789: would be unable to orientate affinely due to the entanglements.
790: The experimental data reflects this and, accordingly, gives
791: slightly lower values of order parameter.
792:
793:
794: \section{Infrared Actuation} \label{ir-sec}
795:
796: \subsection{Typical observations}
797:
798: The detailed response to infrared stimuli is presented in
799: Figs.~\ref{steps}, showing the stress measured in the PDMS1
800: sample. Results for all composites are qualitatively similar. We
801: shall later examine the dependence on the host polymer and the
802: tube concentration. Composites, initially un-aligned, are
803: subjected to an increasing extension that we call pre-strain
804: $\varepsilon$. At each $\varepsilon$, the IR-irradiation takes
805: place and the stress response recorded. The complexity of the
806: plots necessitates more detailed description of what takes place.
807:
808: We begin with a 2\% pre-strain ($\varepsilon=0.02$) applied to it
809: initially. At $t=2$mins the light source is switched on and the
810: stress reading changes downwards, meaning that the sample natural
811: length $L_0$ has expanded on actuation (recall that the actual
812: length $L$ is fixed through $\varepsilon=L/L_0-1$). After a period
813: of constant irradiation, at $t=15$mins, the light source is
814: switched off -- and the stress reading returns to its original
815: value. This experiment is then repeated with the same sample
816: pre-strained at different values, up to 40\%, as shown by the
817: sequence of stress-reading curves in Fig.~\ref{steps}.
818:
819: \begin{figure} %[h]
820: \centering \resizebox{0.4\textwidth}{!}{\includegraphics{fig7}}
821: \caption{The response of a 1wt\% nanocomposite PDMS1 to IR
822: radiation at different levels of pre-strain $\varepsilon$. Stress
823: is measured at fixed sample length (different pre-strain curves
824: labelled on the plot).} \label{steps}
825: \end{figure}
826:
827: The data in Fig.~\ref{speed} is assembled to demonstrate the speed
828: of the actuation process more clearly, while Fig.~\ref{steps2wt}
829: helps differentiate between the light- and heat-driven actuation
830: response. In this case the data is for a PDMS3 composite; as was
831: mentioned above, all materials exhibit the same qualitative
832: features. We plot the change in stress and change in temperature,
833: normalized by their maximal value at saturation in the given
834: experiment; plotted in this form, all the results (for different
835: tube loading and different pre-strain) appear universal.
836:
837: \begin{figure} %[h]
838: \centering \resizebox{0.36\textwidth}{!}{\includegraphics{fig9}}
839: \caption{The speed of actuation response, illustrated by plotting
840: the actuation stress in PDMS3 nanocomposite, $\Delta \sigma$ in
841: kPa, as a function of time for different pre-strain values
842: (labelled on the plot). } \label{speed}
843: \end{figure}
844:
845: The change in temperature by IR-heating is unavoidable and reaches
846: $\Delta T \sim 15^\circ$C maximally on the sample surface, in our
847: setting (thermocouples placed below the surface and embedded in
848: the center of the sample may report the temperature change of up
849: to $20^\circ$C depending on nanotube concentration, but we avoided
850: disturbing the sample in mechanical experiments). This highlights
851: an important question as to whether the mechanical response is due
852: to the photon absorption or plain heat. Figure~\ref{steps2wt}
853: shows that the stress reaches its peak and saturation in $\sim
854: 0.5$min, while the thermal takes over 2min to reach its peak.
855: Although the difference in rates is not very dramatic, the fact
856: that the stress response is faster suggests that its mechanism is
857: not caused by the trivial sample heating. In a separate study (not
858: shown) we reach the conclusion that thermo-mechanical effects do
859: exist (i.e. the MWCNT-loaded composite has a stronger mechanical
860: response to heating than a pristine polymer) but their magnitude
861: is almost an order of magnitude smaller than the direct IR-photon
862: absorption mechanism.
863:
864: \begin{figure} %[h]
865: \centering \resizebox{0.37\textwidth}{!}{\includegraphics{fig8}}
866: \caption{ The normalized stress response plotted alongside the
867: normalized change in temperature, as functions of time (PDMS3,
868: pre-strain $\varepsilon = 20\%$), see text for discussion.}
869: \label{steps2wt}
870: \end{figure}
871:
872: \subsection{Analysis of IR-actuation response}
873:
874: Of great interest is the observation that this response changes
875: sign at a certain level of pre-strain (at $\varepsilon \sim 10\%$
876: in Fig.~\ref{steps}). In other words, relaxed or weakly stretched
877: composites show the reversible \textit{expansion} on irradiation,
878: while the same sample, once strained more significantly,
879: demonstrates an increasing tendency to \textit{contract} (hence
880: the increase in the measured stress). This is our key finding.
881:
882: Figure~\ref{sum} summarizes the magnitude of the IR-actuation
883: effect by plotting the stress step at saturation ($\Delta
884: \sigma_{\sf max}$) in the IR-on state, at different levels of
885: pre-strain and for samples with increasing MWCNT loading. Although
886: this is not explicitly measured in our (isostrain, $L=$const)
887: experiment, we can directly calculate the change of the underlying
888: natural length $L_0$(IR) of the samples on actuation from the
889: known Young modulus values. This is shown on the right axis of the
890: same plot, highlighting the regions of expansion and contraction.
891: Remarkably, all samples with different nanotube loading appear to
892: have a crossover at the same point, around 10\% pre-strain.
893:
894: \begin{figure} %[h]
895: \centering \resizebox{0.4\textwidth}{!}{\includegraphics{fig10}}
896: \caption{The magnitude (in kPa) of exerted actuation stress (the
897: height of steps in Fig.~\ref{steps}, $\Delta \sigma_{\sf max}$),
898: as function of pre-strain. Different PDMS composites are labelled
899: on the plot by their wt\% value. The right $y$-axis shows the
900: corresponding actuation stroke: the change in natural length
901: $L_0$(IR).} \label{sum}
902: \end{figure}
903:
904: For comparison, the pristine PDMS rubber in the same experiment,
905: shows no discernible stress response at all. Also, the response of
906: the PDMS composite with a 3wt\% of carbon black is much lower.
907: Indeed, this 3wt\% carbon-black composite closely follows the
908: low-concentration PDMS0.02 composite. We believe the response is
909: due to trace amounts of nanotubes that can often be found in
910: commercially supplied carbon black. Hence the very small response
911: from such a highly loaded sample. The shift in transition
912: pre-strain may well be due to the trace nanotubes having their
913: alignment hindered by the activated carbon black.
914:
915: \begin{figure} %[h]
916: \centering \resizebox{0.33\textwidth}{!}{\includegraphics{fig11}}
917: \caption{The magnitude of the actuation stroke at
918: $\varepsilon=40\%$ as a function of filler concentration $n$. The
919: maximum of the response at $\sim$2wt\% is evident. The single
920: square symbol gives the value for 3wt\% carbon black filler in
921: PDMS. } \label{Interplay}
922: \end{figure}
923:
924: The interaction between filler particles is also evident when the
925: nanotube concentration is increased beyond 2wt\% loading. Above
926: this value, the magnitude of the actuation stroke decreases
927: sharply. Figure~\ref{Interplay} displays the effect clearly by
928: plotting the maximal change in natural length $L_0$ on IR
929: irradiation, at a fixed $\varepsilon=40\%$, for all PDMS-nanotube
930: composites, and the 3wt\% carbon black system for comparison. A
931: rapid increase in the stroke is observed with increasing
932: concentration, which then peaks at 2wt\% nanotube loading. The
933: reason for the subsequent decline is not obvious. There may well
934: be a number of factors that interplay to reduce the stroke
935: magnitude. At high concentrations entanglements between the long
936: tubes could take place. Note that through conservation of volume,
937: a contraction in the $z$-axis of the tube will be concomitant with
938: an expansion in the $x$-$y$ plane; such expansion may be hindered
939: for a significant number of nanotubes by their nearest neighbors
940: (another representation of entanglement). There may also be an
941: issue of photon screening at higher concentration which is
942: difficult to avoid.
943:
944:
945: \subsection{Observations in other host polymers}\label{hosts}
946:
947: Other polymers acting as a crosslinked host matrix for the
948: low-concentration nanocomposite display the same qualitative
949: behavior as PDMS systems. Figure~\ref{PolydomainCNT} summarizes
950: the response of LCE and SIS composites. The direction of the
951: actuation, changing from expansive to contractive mode with
952: increasing MWCNT alignment, as observed in PDMS-nanotube samples,
953: is unambiguously reproduced for vastly different materials.
954:
955: The magnitude of the actuation stroke is shown in
956: Fig.~\ref{PolydomainCNT}, in comparison with some of the PDMS
957: composites. The value of actuation stress is different for various
958: polymeric systems considered in this work, which is due to the
959: different Young modulus (which we use to calculate the stroke from
960: the measured stress $\Delta \sigma$). We see that the stroke
961: magnitude in these differing materials is in the same range of
962: magnitudes. SIS0.01 has a much lower filler concentration, and
963: again its stroke is comparable to that of a similarly loaded
964: PDMS0.02. This important finding demonstrates the universality of
965: multi-walled nanotubes behaving as photo-actuators regardless of
966: the soft matrix they are in.
967:
968: \begin{figure} %[h]
969: \centering \resizebox{0.3\textwidth}{!}{\includegraphics{fig12}}
970: \caption{Summary of IR actuation stroke from LCE composites
971: (PLCE0.15 and MLCE0.2) and a SIS0.1 composite, as a function of
972: applied pre-strain. Note that MLCE data does not have a crossover
973: at $\varepsilon \sim 10$\%, since the tubes are aligned there at
974: preparation. For comparison, the similar data for two PDMS
975: samples, 0.5 and 0.02, is shown by dashed lines.}
976: \label{PolydomainCNT}
977: \end{figure}
978:
979: The response of nematic liquid crystal elastomers to heat is well
980: documented~\cite{Eugenebook}. Because they can be
981: thermo-responsive materials, the data in Fig.~\ref{PolydomainCNT}
982: is obtained by a complex procedure of subtraction of such
983: background effects. We do not go into its details, as this is
984: irrelevant to the main points of the present paper, however, must
985: emphasize that the plotted response highlights the effect of
986: nanotubes within the given matrix.
987:
988: In Fig.~\ref{PolydomainCNT} we note that the MLCE0.2 sample shows
989: no expansive actuation and a crossover, while the similar
990: polydomain (un-aligned) composite does. This is clearly because
991: the two-step crosslinking at preparation of the monodomain
992: material involves aligning the mesogenic
993: groups~\cite{Kupfer1991LCE}. The embedded nanotubes align strongly
994: under such conditions, as others have found in ordinary liquid
995: nematics~\cite{Lynch02,Dierking2004LC}. As already discussed, the
996: expansive mode of actuation will only occur when the degree of
997: nanotube alignment is very low. It is important that the crossover
998: occurs at $\varepsilon^* \sim 10$\% for all studied materials with
999: nanotubes not aligned before pre-strain.
1000:
1001:
1002: \section{Modelling the mechanism}
1003: There are two main questions to answer: what mechanism is
1004: responsible for such a large photo-mechanical response, and why
1005: does it reverse its direction on sample extension?
1006:
1007: We shall try to deduce the actuation behavior of individual tubes
1008: from the macroscopic observations detailed above. We believe the
1009: change of actuation direction on increasing sample extension is
1010: due to the nanotube alignment induced by pre-strain, as described
1011: in section~\ref{xrayData} and before. In the whole region of our
1012: pre-strains, the orientational order induced in the MWCNT
1013: distribution is, to a good approximation, a linear function of the
1014: strain: $S_d \approx 0.6\varepsilon$ in the affine model. At the
1015: crossover strain $\varepsilon^* \approx 0.1$, the value of the
1016: order parameter would be $S_d^* \sim 0.06$. We now apply the same
1017: ideas about the induced orientational bias and averaging of the
1018: (hypothetical) individual nanotube response.
1019:
1020: Let us assume this individual nanotube response to the IR photon
1021: absorbtion is, in essence, a contraction -- because this is what
1022: our data shows the better-aligned composite response to be. It is
1023: easy to imagine why this could be for an initially rod-like tube:
1024: on photon absorption it could generate instabilities in the form
1025: of kinks, thus decreasing the net length due to the charge carrier
1026: separation. The resulting elastic deformation would be most
1027: pronounced in the already defect-dominated regions of the
1028: nanotube. Such an explanation, based on concentration of induced
1029: polarons~\cite{Verissimo2001,Perebeinos2005}, would also link with
1030: the earlier observation of a similar actuation response under DC
1031: electric field~\cite{Courty2003}. An alternative possibility is to
1032: suggest that large (and fast) local tube heating~\cite{ignite}
1033: causes the surrounding region of locally aligned elastomer to
1034: contract and ``crush'' the nanotube. This version of microscopic
1035: events does not contradict the discussion and the data in
1036: Fig.~\ref{steps2wt}, which shows the (slow) global thermal effect.
1037:
1038: At this stage we have to leave open the question of individual
1039: tube response to near IR radiation. Using an affine approach
1040: similar to the earlier analysis of ordering, let us assume that
1041: each nanotube undergoes a linear contraction by a factor $\Delta =
1042: R_\|({\rm IR}) /R_\|(0) < 1 $ (certainly proportional to radiation
1043: intensity, which was kept constant in our work but has previously
1044: been shown to effect the elastic
1045: response~\cite{Zhang1999ElasticLight}). This contraction must be
1046: accompanied by a transversely-isotropic volume conserving
1047: expansion $1/\sqrt{\Delta}$. This means that a local strain is
1048: created with the principal axes along the nanotube orientation (at
1049: angle $\theta$ to the macroscopic $z$-axis, see Fig.~\ref{affi})
1050: \begin{figure} %[h]
1051: \centering \resizebox{0.2\textwidth}{!}{\includegraphics{fig13}}
1052: \caption{The scheme illustrating how the distortion (kinking) of
1053: an individual tube, lying at an angle $\theta$ to the macroscopic
1054: alignment axis, projects on the $z$-axis to contribute to the
1055: average uniaxial strain, Eq.~(\ref{avL}). } \label{affi}
1056: \end{figure}
1057: $$
1058: \bm{\Lambda}_{\rm local} = \left(
1059: \begin{array}{ccc} 1/\sqrt{\Delta} & 0 &
1060: 0 \cr 0 & 1/\sqrt{\Delta} & 0 \cr 0 & 0 & \Delta
1061: \end{array} \right) .
1062: $$
1063: The projection of this local strain on the macroscopic axis of
1064: sample extension (and force measurement) is
1065: \begin{equation}
1066: \lambda_z({\rm IR}) = \Delta \cos^2 \theta + (1/\sqrt{\Delta})
1067: \sin^2 \theta . \label{lz}
1068: \end{equation}
1069: Averaging the local contribution with the probability to find the
1070: nanotube at this orientation, $P(\theta)$ obtained in
1071: section~\ref{xrayData}, gives an estimate of the effective stroke
1072: of the actuation. When multiplied by the corresponding Young
1073: modulus, the exerted stress of Fig.~\ref{sum} is also obtained:
1074: $\Delta \sigma = Y (\langle \lambda_z \rangle -1)$ at small
1075: deformations. Such a model is very crude indeed, ignoring a large
1076: number of undoubtedly important and delicate factors of continuum
1077: elasticity and nanotube response. However, it is elastically
1078: self-consistent and has only one parameter, $\Delta$ that
1079: presumably carries all the underlying complexity of the problem in
1080: it.
1081:
1082: The orientational averaging is given by using the distribution
1083: $P(\theta)$ with the projection of local strain in Eq.~(\ref{lz}):
1084: \begin{eqnarray}
1085: \langle \lambda_z \rangle &=& \int_{0}^{\pi} [\Delta \cos^2 \theta
1086: + (1/\sqrt{\Delta}) \sin^2 \theta ] P(\theta) \, \sin\theta
1087: d\theta d\varphi \nonumber \\
1088: &\approx & {\textstyle{\frac{1}{3}}} \left(\Delta
1089: +2/\sqrt{\Delta} \right) -{\textstyle{\frac{2}{5}}} \varepsilon
1090: \left( 1/\sqrt{\Delta} - \Delta \right) \label{avL}
1091: \end{eqnarray}
1092:
1093: Although the integral above has a full analytic form, it is more
1094: transparent to present its expansion to the linear order of small
1095: imposed pre-strain $\varepsilon$ as shown in the second line of
1096: Eq.(\ref{avL}). This demonstrates the key point: at very low
1097: pre-strain, $\varepsilon \rightarrow 0$, the average uniaxial
1098: deformation of the disordered nanocomposite is positive
1099: $(\lambda_z-1)$, i.e. the expansion of its natural length.
1100: However, above a threshold pre-strain $\varepsilon^*$ this average
1101: deformations transforms into the sample contraction along $z$. It
1102: is easy to find the crossover,
1103: \begin{eqnarray}
1104: \varepsilon^* \approx
1105: \frac{5(2-\Delta^{1/2}-\Delta)}{6(1+\Delta^{1/2}+\Delta)} ,
1106: \label{estar}
1107: \end{eqnarray}
1108: so that the prediction would be to observe the crossover at
1109: $\varepsilon^* \sim 0.1$ if the nanotube response factor $\Delta
1110: \sim 0.8$. That is, on IR-irradiation the nanotube contracts
1111: overall by $\sim$20\%. The value is higher than one might expect,
1112: considering early reports in the literature of nanotube strains of
1113: only 1-2\%. However, as Fig.~\ref{affi} indicates, our proposition
1114: is not that of the lattice strain of nanotube walls but a
1115: contortion of the tube as a whole. Although this has not been yet
1116: directly observed and reported in the literature, a similar effect
1117: of resonant undulation has been seen (in
1118: simulation~\cite{YakobsonPRL1} and in
1119: experiment~\cite{Poncharal1999}) in response to distortion beyond
1120: the linear regime. Furthermore, the more recent theoretical work
1121: on single-walled tubes supports the idea that massive $z-$axis
1122: contraction. Although in our system the multi-walled tubes respond
1123: under different conditions, being embedded in an elastic matrix
1124: under strain and absorbing the IR photons, the overall distortion
1125: factor of 20\% suggested by the model fit is perhaps not
1126: altogether unreasonable.
1127:
1128: Figure~\ref{theo} plots the full (non-expanded) result of
1129: orientational averaging of actuation stroke $(\langle \lambda_z
1130: \rangle - 1)$ from the integral in Eq.(\ref{avL}) to illustrate
1131: the points discussed in this section. The qualitative behavior (as
1132: summarized in Fig.~\ref{sum}) is reproduced here almost exactly,
1133: including the magnitude of the predicted actuation stroke [that
1134: is, the ratio $L_0({\rm IR})/L_0(0)-1$]. Note that we use only
1135: one parameter, $\Delta$ to match both the crossover
1136: $\varepsilon^*$ and the actuation stroke magnitude, so the
1137: conclusion is quite satisfactory and agrees with an apparent
1138: universality discussed in section~\ref{hosts}. It is very likely
1139: that the orientational feature of the effect, with its change of
1140: actuation direction at a critical level of induced alignment, $S_d
1141: \sim 0.06$, is captioned correctly, while much more work is
1142: required to understand the individual nanotube response to IR
1143: radiation generating the phenomenological factor $\Delta$ used in
1144: this analysis.
1145:
1146: \begin{figure} %[h]
1147: \centering \resizebox{0.4\textwidth}{!}{\includegraphics{fig14}}
1148: \caption{The result of the affine theoretical model,
1149: Eq.~(\ref{avL}); the dashed line shows the linear approximation at
1150: small pre-strain. Nanotube contraction factor is chosen to be
1151: $\Delta = 0.8$, as suggested by the crossover strain value
1152: $\varepsilon^*\sim 0.1$. } \label{theo}
1153: \end{figure}
1154:
1155:
1156: \section{Conclusions}
1157:
1158: To summarize, this work describes the rich photo-actuation
1159: phenomena of carbon nanotubes embedded in crosslinked rubbery
1160: matrices. The composite materials show the ability to change their
1161: actuation direction, from expansive to contractive response, as
1162: greater imposed strain is applied to the sample. We use differing
1163: host polymers and confirm their relatively neutral role in the
1164: actuation mechanism.
1165:
1166: Theoretical models have been put forward to describe the
1167: orientational order imposed on the nanotubes by a uniaxial strain
1168: and the resulting actuation. Treating the nanotubes as rigid rods
1169: that rotate affinely in a deforming matrix is a very simplistic
1170: view, but it gives predictions that agree with experiment
1171: qualitatively and often quantitatively. We believe that the
1172: (certainly wrong) idea of the whole tube acting as a rigid rod is
1173: not actually necessary -- in effect, in our model, the ``rigid
1174: rods'' are nanotube segments below persistence length. In that
1175: case, as in main-chain semiflexible nematic polymers, the model is
1176: non-controversial and the agreement with experiment not
1177: coincidental. The tube orientational distribution appears to
1178: account well for the key macroscopic features of the observed
1179: photo-actuation.
1180:
1181: The strength of photo-mechanical response, at a given radiation
1182: intensity, is of the order of tens of kPa. Translated into the
1183: stroke, this corresponds to actuation strains of +2 (expansion) to
1184: -10\% (contraction) depending on the nanotube concentration,
1185: alignment (controlled by pre-applied strain) and the host matrix.
1186: As expected, the response increases at higher nanotube loading --
1187: however, only up to a limit. Beyond this limit ($\sim 2\%$ in
1188: PDMS), the macroscopic actuation is inhibited by inter-tube
1189: interactions and possible charge accumulation. The similar
1190: (thermal actuation) behavior is also observed when the samples are
1191: heated by the same amount, but this has a much lower amplitude.
1192:
1193: Understanding the nature of the actuator mechanisms in this system
1194: certainly warrants further theoretical and experimental
1195: investigation. Many questions remain completely unclear, in
1196: particular, what the effect would be if different types of
1197: nanotube were used i.e. smaller multi-wall diameters, single-wall
1198: tubes, various chirality etc. With actuating materials already
1199: used in such widespread applications, from micromanipulators to
1200: vibration control, the discovery of a structure that can respond
1201: to stimulus in both directions may open new possibilities and
1202: could mean an important new step toward finding applications for
1203: nanotube based materials above and beyond improvements in existing
1204: carbon fibre technologies.
1205:
1206: \begin{acknowledgments}
1207: We thank H. K\"orner, R. Vaia, A. Craig and P. Cicuta for useful
1208: discussions, and K. Channon for help with nanotube absorbance
1209: data. This work was carried out with the support of the
1210: Engineering and Physical Sciences Research Council and a CASE
1211: award from Makevale Ltd.
1212: \end{acknowledgments}
1213:
1214: \begin{thebibliography}{10}
1215:
1216: \bibitem{Huber1997}
1217: J.~E. Huber, N.~A. Fleck, and M.~F. Ashby,
1218: \newblock Proc. Roy. Soc. London Ser. A. {\bf 453}, 2185 (1997).
1219:
1220: \bibitem{refAlloys}
1221: K.~Bhattacharya,
1222: \newblock {\em Microstructure of Martensite},
1223: \newblock Oxford University Press, Oxford, 2004.
1224:
1225: \bibitem{Langer01}
1226: A.~Lendlein, A.~Scmidt, and R.~Langer,
1227: \newblock Proc. Nat. Acad. Sci. {\bf 98}, 842 (2001).
1228:
1229: \bibitem{Eugenebook}
1230: M.~Warner and E.~M. Terentjev,
1231: \newblock {\em Liquid Crystal Elastomers},
1232: \newblock Oxford University Press, Oxford, 2003.
1233:
1234: \bibitem{Ahir2005}
1235: S.~V. Ahir and E.~M. Terentjev,
1236: \newblock Nature Mater. {\bf 4}, 491 (2005).
1237:
1238: \bibitem{Dai2005}
1239: N.~W.~S. Kam, M.~O'Connel, J.~A. Wisdom, and H.~J. Dai,
1240: \newblock Proc. Nat. Acad. Sci. {\bf 102}, 11600 (2005).
1241:
1242: \bibitem{Ajayan1997}
1243: P.~M. Ajayan and T.~W. Ebbesen,
1244: \newblock Rep. Prog. Phys. {\bf 60}, 1025 (1997).
1245:
1246: \bibitem{Forro2001}
1247: L.~Forro and C.~Schonenberger,
1248: \newblock {\em Physical properties of multi-wall nanotubes}, vol.~80 of {\em
1249: Topics in Applied Physics},
1250: \newblock 2001.
1251:
1252: \bibitem{PhysicalNanotubesBook}
1253: R.~Saito, G.~Dresselhaus, and M.~S. Dresselhaus,
1254: \newblock {\em Physical properties of carbon nanotubes},
1255: \newblock Imperial College Press, London, 1998.
1256:
1257: \bibitem{Breuer2004}
1258: O.~Breuer and U.~Sundararaj,
1259: \newblock Polym. Comp. {\bf 25}, 630 (2004).
1260:
1261: \bibitem{Andrews2004}
1262: R.~Andrews and M.~C. Weisenberger,
1263: \newblock Curr. Opin. Solid State Mater. Sci. {\bf 8}, 31 (2004).
1264:
1265: \bibitem{Harris2004}
1266: P.~J.~F. Harris,
1267: \newblock Int. Mater. Rev. {\bf 49}, 31 (2004).
1268:
1269: \bibitem{Thostenson2001}
1270: E.~T. Thostenson, Z.~F. Ren, and T.~W. Chou,
1271: \newblock Comp. Sci. Technol. {\bf 61}, 1899 (2001).
1272:
1273: \bibitem{Nalwa2005}
1274: S.~V. Ahir and E.~M. Terentjev,
1275: \newblock {in: {\em Polymeric Nanostructures and Their Applications}},
1276: \newblock American Scientific Publishers,
1277: \newblock ed H.S. Nalwa (2006)
1278:
1279: \bibitem{Cumings2000}
1280: J.~Cumings and A.~Zettl,
1281: \newblock Science {\bf 289}, 602 (2000).
1282:
1283: \bibitem{Poncharal1999}
1284: P.~Poncharal, Z.~L. Wang, D.~Ugarte, and W.~A. de~Heer,
1285: \newblock Science {\bf 283}, 1513 (1999).
1286:
1287: \bibitem{Willams2002}
1288: P.~A. Williams, S.~J. Papadakis, A.~M. Patel, M. Falvo, S.
1289: Washburn and R. Superfine,
1290: \newblock Phys. Rev. Lett. {\bf 89}, 255502 (2002).
1291:
1292: \bibitem{Fennimore2003}
1293: A.~Fennimore, T.~D. Yuzvinsky, W.~Q. Han, M.~S. Fuhrer, J. Cumings
1294: and A. Zettl,
1295: \newblock Nature {\bf 424}, 408 (2003).
1296:
1297: \bibitem{Zhang1999ElasticLight}
1298: Y.~Zhang and S.~Iijima,
1299: \newblock Phys. Rev. Lett. {\bf 82}, 3472 (1999).
1300:
1301: \bibitem{Landi2002}
1302: B.~J. Landi, R.~P. Raffaelle, M.~J. Heben, J.~L. Alleman, W.
1303: VanDerveer and T. Gennett,
1304: \newblock Nano Lett. {\bf 2}, 1329 (2002).
1305:
1306: \bibitem{Koerner2004}
1307: H.~Koerner, G.~Price, N.~A. Pearce, M.~Alexander, and R.~A. Vaia,
1308: \newblock Nature Mater. {\bf 3}, 115 (2004).
1309:
1310: \bibitem{Tahhan2003}
1311: M.~Tahhan, V.~T. Truong, G.~M. Spinks, and G.~G. Wallace,
1312: \newblock Smart Mater. Struct. {\bf 12}, 626 (2003).
1313:
1314: \bibitem{Naciri2003}
1315: J.~Naciri, A. Srinivasan, H. Jeon, N. Nikolov, P. Keller and B.~R.
1316: Ratna,
1317: \newblock Macromolecules {\bf 36}, 8499 (2003).
1318:
1319: \bibitem{Courty2003}
1320: S.~Courty, J.~Mine, A.~R. Tajbakhsh, and E.~M. Terentjev,
1321: \newblock Europhys. Lett. {\bf 64}, 654 (2003).
1322:
1323: \bibitem{Garg1998}
1324: A.~Garg and S.~B. Sinnott,
1325: \newblock Chem. Phys. Lett. {\bf 295}, 273 (2002).
1326:
1327: \bibitem{Sinnott2002}
1328: S.~B. Sinnott,
1329: \newblock J. Nanosci. Nanotechnol. {\bf 2}, 113 (2002).
1330:
1331: \bibitem{Kupfer1991LCE}
1332: J.~K{\"u}pfer and H.~Finkelmann,
1333: \newblock Macromol. Chem.-Rapid Comm. {\bf 12}, 717 (1991).
1334:
1335: \bibitem{SISbook}
1336: Papers~in:,
1337: \newblock {\em Developments in Block Copolymers},
1338: \newblock Applied Science Publishers Ltd, Oxford,
1339: \newblock ed I. Goodman (1982).
1340:
1341: \bibitem{Deutsch1991}
1342: M.~Deutsch,
1343: \newblock Phys. Rev. A {\bf 44}, 8264 (1991).
1344:
1345: \bibitem{Gorga2004MechAlign}
1346: R.~E. Gorga and R.~E. Cohen,
1347: \newblock J. Polym. Sci. B-Polym. Phys. {\bf 42}, 2690 (2004).
1348:
1349: \bibitem{Charlier1993}
1350: J.~C. Charlier and J.~P. Michenaud,
1351: \newblock Phys. Rev. Lett. {\bf 70}, 1858 (1993).
1352:
1353: \bibitem{rabin}
1354: S.~Panyukov and Y.~Rabin,
1355: \newblock Phys. Rep. {\bf 269}, 1 (1996).
1356:
1357: \bibitem{Kelarakis2005}
1358: A.~Kelarakis, K.~W. Yoon, I. Sics, R.~H. Somani, B.~S. Hsiao and
1359: B. Chu,
1360: \newblock Polymer {\bf 46} (2005).
1361:
1362: \bibitem{Lynch02}
1363: M.~Lynch and D.~Patrick,
1364: \newblock Nano Lett. {\bf 97}, 1197 (2002).
1365:
1366: \bibitem{Dierking2004LC}
1367: I.~Dierking, G.~Scalia, P.~Morales, and D.~LeClere,
1368: \newblock Adv. Mater. {\bf 16}, 865 (2004).
1369:
1370: \bibitem{ferronematic}
1371: S. Burylov and Y. Raikher,
1372: \newblock Phys. Rev. E {\bf 50}, 358 (2004).
1373:
1374: \bibitem{Somoza2001}
1375: A.~M. Somoza, C.~Sagui, and C.~Roland,
1376: \newblock Phys. Rev. B {\bf 6308}, 081403 (2001).
1377:
1378: \bibitem{Islam2004}
1379: M.~F. Islam, A.~M. Alsayed, Z. Dogic, J. Zhang, T.~C. Lubensky and
1380: A.~G. Yodh,
1381: \newblock Phys. Rev. Lett. {\bf 92}, 088303 (2004).
1382:
1383: \bibitem{Song2005}
1384: W.~H. Song and A.~H. Windle,
1385: \newblock Macromolecules {\bf 38}, 6181 (2005).
1386:
1387: \bibitem{Verissimo2001}
1388: M.~Verissimo-Alves, R.~B. Capaz, B.~Koiller, E.~Artacho, and
1389: H.~Chacham,
1390: \newblock Phys. Rev. Lett. {\bf 86}, 3372 (2001).
1391:
1392: \bibitem{Perebeinos2005}
1393: V.~Perebeinos, J.~Tersoff, and P.~Avouris,
1394: \newblock Phys. Rev. Lett. {\bf 94}, 086802 (2005).
1395:
1396: \bibitem{ignite}
1397: P.~M. Ajayan and M.~Terrones and A. de la Guardia and V. Huc and
1398: N. Grobert and B.~Q. Wei and H. Lezec and G. Ramanath and
1399: T.~W. Ebbesen,
1400: \newblock Science {\bf 296}, 5568 (2002).
1401:
1402: \bibitem{YakobsonPRL1}
1403: B.~I. Yakobson, C.~J. Brabec, and J.~Bernholc,
1404: \newblock Phys. Rev. Lett. {\bf 76}, 2511 (1996).
1405:
1406: \end{thebibliography}
1407:
1408: \end{document}
1409: