Complex Patterns in a Simple System Author(s): John E. Pearson Reviewed work(s): Source: Science, New Series, Vol.

261, No. 5118 (Jul. 9, 1993), pp. 189-192 Published by: American Association for the Advancement of Science Stable URL: http://www.jstor.org/stable/2881810 . Accessed: 24/04/2012 18:38
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org.

American Association for the Advancement of Science is collaborating with JSTOR to digitize, preserve and extend access to Science.

http://www.jstor.org

mm REPORTS
6. S. R. Marder, C. B. Gorman, B. G. Tiemann, L.-T. Cheng, Proc. SPIE 1775, 19 (1993); C. B. Gorman and S. R. Marder, in preparation. 7. W. Drenth and E. H. Wiebenga, Acta Crystallogr. 8, 755 (1955). 8. R. H. Baughman, B. E. Kohler, I. J. Levy, C. W. Spangler, Synth. Methods 11, 37 (1985). 9. P. Groth, Acta Chem. Scand. B 41, 547 (1987). 10. F. Chentli-Bechikha, J. P. Declercq, G. Germain, M. V. Meerssche, Cryst. Struct. Comm. 6, 421 (1977). 11. L. G. S. Brooker et al., J. Am. Chem. Soc. 73, 5332 (1951). 12. R. Radeglia and S. Dahne, J. Mol. Struct. 5, 399 (1970). 13. S. R. Marder et al. J. Am. Chem. Soc. 115, 2525 (1993). 14. S. Schneider, Ber. Buns. Ges. 80, 218 (1976). 15. H. E. Schaffer, R. R. Chance, R. J. Silbey, K. Knoll, R. R. Schrock, J. Chem. Phys. 94, 4161 (1991) 16. F. Kajzar and J. Messier, Rev. Sci. Instrum. 58, 2081 (1987). 17. S. R. Marder, J. W. Perry, F. L. Klavetter, R. H. Grubbs, in Organic Materials for Nonlinear Optics, R. A. Hann and D. Bloor, Eds. (Royal Society of Chemistry, London, 1989), pp. 288-294. 18. F. Kajzar, in Nonlinear Optics of Organics and Semiconductors, T. Kobayashi, Ed. (SpringerVerlag, Berlin, 1989), pp. 108-119. 19. S. H. Stevenson, D. S. Donald, G. R. Meredith, in Nonlinear Optical Properties of Polymers, Mat. Res. Soc. Symp. Proc. 109 (Materials Research Society, Pittsburgh, 1988), pp. 103-108. 20. Y. Marcus, J. Soln. Chem. 20, 929 (1991). 21. B. M. Pierce, Proc. SPIE 1560, 148 (1991). 22. C. W. Dirk, L.-T.Cheng, M. G. Kuzyk, Int. J Quant. Chem. 43, 27 (1992). 23. Enhancement of y by the P2 term has been discussed previously: A. F. Garito, J. R. Heflin, K. Y Yong, 0. Zamani-khamiri, Proc SPIE 971, 2 (1988). 24. Increased y in donor-acceptor-conjugated organics relative to centrosymmetric analogs has been observed experimentally: C. W. Spangler, K. 0. Havelka, M. W. Becker, T. A. Kelleher, L.-T. Cheng, Proc. SPIE 1560, 139 (1991); L.-T. Cheng et al., J. Phys. Chem. 95, 10631 (1991); L.-T. Cheng etal., ibid., p. 10643. 25. The research was performed, in part, at the Center for Space Microelectronics Technology, Jet Propulsion Laboratory (JPL), California Institute of Technology, and was supported, in part, by the Strategic Defense Initiative Organization-Innovative Science and Technology Office, through an agreement with the National Aeronautics and Space Administration (NASA). Support at the Beckman Institute by the Air Force Office of Scientific Research (grant F49620-92-J-01 77) is also acknowledged. C.B.G. thanks the JPL director's office for a postdoctoral fellowship. G.B. thanks the National Research Council and NASA for a resident research associateship at JPL.

terns presentedin this reportwill yield to these now-standard technologies. The Gray-Scott model correspondsto the followingtwo reactions: U + 2V 3V (1) V P Both reactionsare irreversible,so P is an inert product.A nonequilibrium constraint is represented a feed termfor U. Both U by andV areremovedby the feed process.The resultingreaction-diffusion equationsin dimensionlessunits are:

au

= DUV2U

UV2 + F(1 - U)

aV a = D'V2V+ UV2-(F + k)V (2)

where k is the dimensionless rate constant of the second reactionand F is the dimensionlessfeed rate. The systemsize is 2.5 by 2.5, and the diffusion coefficientsare Du = 2 x 10' and Dv = 10-5. The boundary 29 January 1993; accepted 19 May 1993 conditionsare periodic.Beforethe numerical resultsare presented,considerthe behavior of the reaction kinetics which are described the ordinary by differential equations that resultupondropping diffusion the termsin Eq. 2. In the phasediagram shown in Fig. 1, a John E. Pearson trivial steady-statesolution U = 1,V = 0 Numerical simulations of a simple reaction-diffusion model reveal a surprising variety of exists and is linearlystable for all positive irregular spatiotemporal patterns. These patterns arise in response to finite-amplitude F and k. In the region bounded above by perturbations. Some of them resemble the steady irregular patterns recently observed in the solid line and below by the dotted thin gel reactor experiments. Others consist of spots that grow until they reach a critical line, the system has two stable steady size, at which time they divide in two. If in some region the spots become overcrowded, states. For fixed k, the nontrivial stable all of the spots in that region decay into the uniform background. uniform solution loses stability through saddle-nodebifurcation as F is increased through the upper solid line or by Hopf bifurcation to a periodic orbit as F is Pattems occurin nature scalesranging to finite-amplitude at perturbations. The re- decreasedthrough the dotted line. [Fora from the developingDrosophila embryoto sponse of this model to such perturbations discussionof bifurcationtheory, see chapthe large-scale structure the universe.At of was previously studiedin one spacedimen- ter 3 of (7).] In the case at hand, the the familiarmundanescales we see snow- sion by Vastanoet al. (4), who showedthat bifurcatingperiodicsolution is stable for k < 0.035 and unstable for k > 0.035. flakes,cloud streets, and sand ripples.We steady spatial patterns could form even see convectiverollpattemsin hydrodynamic when the diffusioncoefficientswere equal. There are no periodicorbitsfor parameter experiments.We see regularand almost The responseof the system in one space values outside the region enclosed by the regularpattems in the concentrationsof dimension is nontrivialand dependsboth solid line. Outsidethis regionthe systemis and chemically reacting and diffusingsystems on the controlparameters on the initial excitable. The trivial state is linearly sta(1). As a consequence of the enormous perturbation.It will be shown that the ble and globally attracting. Small perturrangeof scalesoverwhichpattemformation patterns that occurin two dimensions range bations decayexponentiallybut largerperoccurs,new pattemformation phenomenon from the well-knownregularhexagons to turbations result in a long excursion is potentiallyof greatscientificinterest. In irregularsteady patterns similar to those through phase space before the system this report,I describepatternsrecentlyob- recentlyobservedby Lee et al. (5) to cha- returnsto the trivial state. servedin numerical on experiments a simple otic spatio-temporal patterns.For the ratio The simulations forward are Eulerintegrareaction-diffusion model.These pattemsare of diffusioncoefficientsused, there are no tionsof the finite-difference resultequations unlike any that have been previouslyob- stableTuringpatterns. of ing fromdiscretization the diffusion operaservedin theoreticalor numerical studies. Most work in this field has focused on tor. The spatialmesh consistsof 256 by 256 The systemis a variantof the autocata- patternformationfrom a spatiallyuniform grid points. The time step used is 1. Spot lytic Selkov model of glycolysis(2) and is state that is near the transitionfromlinear checksmadewith meshesas largeas 1024by due to Gray and Scott (3). A variety of stability to linear instability. With this 1024andtimestepsas smallas 0.01 produced in spatio-temporal patternsform in response restriction, standardbifurcation-theoretic no qualitative difference the results. Initially, the entire system was placed in tools such as amplitude equations have been developedand usedwith considerable the trivial state (U = 1,V = 0). The 20 by Center for Nonlinear Studies, Los Alamos National Laboratory, Los Alamos, NM 87545. success (6). It is unclearwhether the pat- 20 mesh point area located symmetrically

ComplexPatterns in a Simple System

SCIENCE * VOL. 261 * 9 JULY 1993

189

about the center of the grid was then perturbedto (U = 1/2,V = 1/4). These with + 1% conditionswere then perturbed randomnoise in orderto breakthe square The systemwas then integrated symmetry. for 200,000 time steps and an image was saved. In all cases, the initial disturbance propagated outward from the central square,leaving pattems in its wake, until the entire grid was affectedby the initial was The propagation squareperturbation. wave-like, with the leading edge of the movingwith an approximately perturbation on constantvelocity.Depending the paramof etervalues,it tookon the order 10,000 to 20,000 time stepsforthe initialperturbation over the entiregrid.The propagato spread is tion velocity of the initial perturbation of thuson the order 1 x 10-4 spaceunitsper time unit. After the initial period during spread,the system which the perturbation statethatwaseither wentinto an asymptotic or time-independent time-dependent,devalues. pendingon the parameter 2 one Figures and 3 are phasediagrams; canviewFig.3 asa mapandFig. 2 as the key in to the map. The 12 patternsillustrated by Fig. 2 are designated Greekletters.The of color indicatesthe concentration U with U red representing = 1 and blue representto ing U = 0.2; yellow is intermediate red and blue. In Fig. 3, the Greek characters indicate the pattem foundat that point in

parameter space. There are two additional symbols in Fig. 3, R and B, indicating spatially uniform andblue states,respecred to tively. The red state corresponds (U = 1,V = 0) and the blue statedependson the exact parametervalues but corresponds roughlyto (U = 0.3,V = 0.25). and Patterna is time-dependent consists of fledglingspiralsthat are constantlycolliding and annihilating each other: full neverform.Patternf is time-depenspirals called dent and consistsof whatis generally

phase turbulence(8), which occursin the vicinity of a Hopf bifurcationto a stable periodic orbit. The medium is unable to synchronize the phase of the oscillators so varies as a function of position. In the presentcase, the small-amplitude periodic orbitthat bifurcates unstable.Pattem -yis is time-dependent. It consists primarilyof stripesbut there are smalllocalizedregions that oscillatewith a relatively high frequenThe active regionsdisappear, cy (-10-). but new ones alwaysappearelsewhere.In

0.3

0.2-

Fig. 2. The key to the map. The patternsshown in the figureare designated by Greekletters,which
are used in Fig. 3 to indicate the pattern found at a given point in parameter space. 0.08 R Fig. 3. The map. The Greek letters indicate the location in parameter space where the patterns in Fig. 2 were found; B and R indicate that the system evolved to uniform blue R RR and red states, respectively.

0.1-

j.. 0.0 -0.00

. . . 0.02 0.04 k 0.06 0.08

0.06

B
BK c

B:

Fig. 1. Phase diagramof the reactionkinetics. Outside the region bounded by the solid line, there is a single spatiallyuniform state (called the trivial state) (U = 1, V = 0) that is stable for all (F,k). Insidethe regionbounded by the solid line, there are three spatially uniformsteady states. Above the dotted line and below the solid line,the system is bistable:Thereare two stable steady states in this region.As F linearly is decreased throughthe dotted line, the nontrivial stable steady state loses stability through The periodicorbit Hopf bifurcation. bifurcating is stable for k < 0.035 and unstable for k > 0.035. No periodic orbits exist for parameter values outside the region bounded by the solid
190

ICKI

- 0.04 B-'

B O B .
/

X R X RR

~~~
0.02
.

~B
B .

.
a
S E

, R sR R R R
R

R
0.00 0.03 0.04

R RR a R

0.05

0.06

0.07

SCIENCE * VOL. 261 * 9 JULY 1993

REPORTS Fig. 2 thereis an active regionnearthe top center of pattem y. Pattern 8 consists of staregular hexagonsexcept for apparently a ble defects. Pattern- is time-dependent: few of the stripesoscillatewithoutapparent decay, but the remainderof the pattern
remains time-independent. Pattem L is time-dependent and was observed for only a single parameter value. Patterns 0, K, and ,u resemble those observed by Lee et al. (5). When blue waves collide, they stop, as do those observed by Lee et al. In pattern ,u, long stripes grow in length. The growth is parallel to the stripes and takes place at the tips. If two distinct stripes that are both growing are pointed directly at each other, it is always observed that when the growing tips reach some critical separation distance, they alter their course so as not to collide. In pattems 0 and K, the perturbations grow radially outward with a velocity normal to the stripes. In these cases if two stripes collide, they simply stop, as do those observed by Lee et al. I have also observed, in one space dimension, fronts propagating toward each other that stop when they reach a critical separation. This is fundamentally new behavior for nonlinear waves that has recently been observed in other models as well (9). Patterns e, t, and X share similarities. They consist of blue spots on a red or yellow background. Pattern X is time-independent and patterns e and t are time-dependent. Note that spots occur only in regions of

parameter spacewhere the systemis excitsteadystateis the able and the sole uniform red state (U = 1,V = 0). Thus, the blue spots cannot persistfor extended time unless thereis a gradientpresent.Becausethe for gradientis required the existenceof the spots, they must have a maximumsize or there would be blue regionsthat were esSuch regionswould sentiallygradient-free. necessarily decayto the redstate. Note that these gradientsare self-sustaining and are not imposed extemally. After the initial perturbation, spots increasein number the until they fill the system. This processis visuallysimilarto cell division.Aftera spot has divided to form two spots, they move awayfromeach other. Duringthis period, each spot grows radially outward. The of growthis a consequence excitability.As the spots get furtherapart, they begin to to elongate in the directionperpendicular their motion. When a critical size is is achieved,the gradient no longersufficient to maintainthe centerin the blue state, so the center decaysto red, leaving two blue in spots. This processis illustrated Fig. 4. Figure4A was made just after the initial had to square perturbation decayed leavethe fourspots.In Fig. 4B, the spotshave moved awayfromeach other and are beginningto elongate. In Fig. 4C, the new spots are clearlyvisible. In Fig. 4D, the replication evoluprocessis complete.The subsequent tion depends on the control parameters. in state. Pattem4 PattemXremains a steady

but remains time-dependent with long-range spatialorderexcept for local regionsof activity. The activeregionsarenot stationary. At any one instant, they do not appear frompattem ; (Fig.2) different qualitatively but the location of the red disturbances to with time. Pattem? appears have changes ordereitherin time or space. no long-range Once the systemis filled with blue spots, they can die due to overcrowding.This occurswhenmanyspotsarecrowded togethover an extendedregion er and the gradient becomes too weak to supportthem. The spots in such a region will collapsenearly red simultaneously leavean irregular hole. to morespotson the boundary Therearealways of any hole, and aftera few thousandtime stepsno sign of the hole will remain.The will spotson its border have filledit. Figure 5 illustrates process. this expoPattem? is chaotic.The Liapunov the nent (which determines rate of separais tion of nearbytrajectories) positive. The time (the inverseof the Liapunov Liapunov equalto exponent)is 660 timesteps,roughly the time it takes for a spot to replicate,as shown in Fig. 4. This time periodis also about how long it takes for a molecule to diffuseacrossone of the spots. The time of average pattem ? is constantin space. All of the patternspresented here arose in response to finite-amplitude perturbacoefficients used tions. The ratioof diffusion was 2. It is now well known that Turing that pattem instabilities leadto spontaneous

Fig. 4 (left). Time evolution of spot multiplication. This figure was produced in a 256 by 256 simulation with physical dimensionsof 0.5 by 0.5 and a time step of 0.01. The times t at which the figures were taken
are as follows: (A) t = 0; (B) t = 350; (C) t = 510; and (D) t = 650.

Fig. 5 (right). Time evolution of pattern E. The images are 250 time units apart. In the corners (which map to the same point in physical space), one can see a yellow region in (A) to (C). Ithas decayed to red in (D). In (A) and (B), the center of the left border has a red region that is nearly filled in (D). 191

SCIENCE * VOL. 261 * 9 JULY 1993

in cannot occurin systems which formation all diffusioncoefficientsare equal. [For a of discussion theseissues,see comprehensive
Pearson and co-workers (10, 11); for a dis-

in cussionof Turinginstabilities the model at hand, see Vastanoet al. (12).] The only Turingpatternsthat can occurbifurcate off the nontrivial steadyuniform state (the blue in state). Mostof the patterns discussed this reportoccurfor parameter valuessuch that the nontrivialsteady state does not exist. With the ratioof diffusion coefficients used here, Turingpatterns occuronly in a narrow parameter regionin the vicinityof F = k = 0.0625, wherethe line of saddle-node bifurcations coalesceswith the line of Hopf biIn furcations. the vicinityof this point, the branchof small-amplitude is Turingpatterns unstable(12). With equaldiffusion no coefficients, patin ternsformedin which smallasymmetries the initial conditionswere amplified the by This observation probably can be dynamics. in understood terms of the followingfact: Nonlinearplane waves in two dimensions cannot be destabilized diffusionin the by case that all diffusion coefficients equal are (13). Duringthe initial stagesof the evolution, the cornersof the square perturbation are rounded off. The perturbationthen evolves as a radialwave, either inwardor outward on depending the parameter values. Such a wave cannot undergospontaneous unlessthe diffusion coefsymmetry breaking ficientsareunequal.However,I foundsymover a wide rangeof parammetrybreaking etervaluesfora ratioof diffusion coefficients of 2. Such a ratio is physicallyreasonable even for small moleculesin aqueoussolution. Given this diffusion ratioand the wide rangeof parameters which the replicatover ing spot patternsexist, it is likely that they will soon be observed experimentally. et RecentlyHasslacher al. have demonstratedthe plausibility subcellular of chemical patternsthroughlattice-gas simulations of the Selkov model (14). The patterns discussedin the presentarticlecan also be found in lattice-gas simulations of the Selkovmodelandin simulations carried out in threespacedimensions.Perhaps they are related to dynamicalprocessesin the cell such as centrosomereplication.
REFERENCESAND NOTES
1. G. Nicolis and 1. Prigogine, Self-Organization in Non-EquilibriumSystems (Wiley, New York, 1977). 2. E. E. Selkov, Eur. J. Biochem. 4, 79 (1968). 3. P. Gray and S. K. Scott, Chem. Eng. Sci. 38, 29 (1983); ibid. 39, 1087 (1984); J. Phys. Chem. 89, 22 (1985). 4. J. A. Vastano, J. E. Pearson, W. Horsthemke, H. L. Swinney, Phys. Lett. A 124, 6 (1987). ibid., p. 7; ibid., p. 320. 5. K. J. Lee, W. D. McCormick, Q. Ouyang, H. L. Swinney, Science 261, 192 (1993). 6. P. Hohenberg and M. Cross, Rev. Mod. Phys. 65, 3 (1993).

7. J. Guckenheimer and P. Holmes, Nonlinear Oscillations, Dynamical Systems, and Bifurcations of Vector Fields (Springer-Verlag, Berlin, 1983), chap. 3. 8. Y. Kuramoto, Chemical Oscillations, Waves, and Turbulence (Springer-Verlag, Berlin, 1984). 9. A. Kawczynski, W. Comstock, R. Field, Physica D 54, 220 (1992); A. Hagberg and E. Meron, University of Arizona preprint. 10. J. E. Pearson and W. Horsthemke, J. Chem. Phys. 90, 1588 (1989). 11. J. E. Pearson and W. J. Bruno, Chaos 2, 4 (1992); ibid., p. 513. 12. J. A. Vastano, J. E. Pearson, W. Horsthemke, H. L. Swinney, J. Chem. Phys. 88, 6175 (1988).

13. J. E. Pearson, Los Alamos Publ. LAUR 93-1758 (Los Alamos National Laboratory, Los Alamos, NM, 1993). 14. B. Hasslacher, R. Kapral, A. Lawniczak, Chaos 3, 1 (1993). 15. I am happy to acknowledge useful conversations with S. Ponce-Dawson, W. Horsthemke, K. Lee, L. Segel, H. Swinney, B. Reynolds, and J. Theiler. I also thank the Los Alamos Advanced Computing Laboratory for the use of the Connection Machine and A. Chapman, C. Hansen, and P. Hinker for their ever-cheerful assistance with the figures. 7 April 1993; accepted 13 May 1993

PatternFormation Interacting by ChemicalFronts

KyoungJ. Lee, W. D. McCormick, Ouyang, HarryL. Swinney* Qi
Experiments on a bistable chemical reaction in a continuously fed thin gel layer reveal a new type of spatiotemporal pattern, one in which fronts propagate at a constant speed until they reach a critical separation (typically 0.4 millimeter) and stop. The resulting asymptotic state is a highly irregular stationary pattern that contrasts with the regular patterns such as hexagons, squares, and stripes that have been observed in many nonequilibrium systems. The observed patterns are initiated by a finite amplitude perturbation rather than through spontaneous symmetry breaking.

In recent years, patternformationhas become a very active area of research,motivated in part by the realizationthat there are many common aspects of patterns formedby diversephysical,chemical, and biologicalsystemsand by cellularautomata and differential equationmodels. In experiments on a chemical system, we have discovered new type of patternthat differs a qualitativelyfrom the previously studied chemical waves [rotatingspirals (1)], stationary"Turing" patterns(2-4), and chaotic patterns(5). These new patternsform only in responseto large-amplitude perturbations-small-amplitude perturbations decay. A large perturbation evolves into an irregular pattern that is stationary(timeindependent)(Fig. 1). The patterns have a length scale determinedby the interaction of the chemical fronts, which propagate towardone anotherat constantspeeduntil they reach a critical distance and stop, as Fig. 2 illustrates. The growthof these front patternsis markedlydifferentfrom Turing The frontpatternsdeveloplocally patterns: andspread fill space,as in crystal to growth, whereasTuringpatternsemergespontanewhen the criticalvalue of ouslyeverywhere a control parameter exceeded. is The front patternsare highly irregular, in contrast with Turing patterns, which of emergeas a regular array stripesor hexagons (in two-dimensionalsystems) at the
Center for Nonlinear Dynamics and the Department of Physics, University of Texas at Austin, Austin, TX 78712. *To whom correspondence should be addressed.

transition from a uniform state (4). The interactionof frontsillustrated Fig. 2 also in contrasts with the behavior in excitable chemicalmedia, wherecollidingfrontsannihilateone another(1), and with solitons, where nonlinear waves pass through one another (6). Our experimentshave been conducted using an iodate-ferrocyanide-sulfite reaction, which is known to exhibit bistability and largeoscillationsin pH in stirredflow reactors(7). The other reactionsthat yield stationarychemical patternsare the wellstudied chlorite-iodide-malonic acid reaction (3-5) and a variantreaction (8) that uses chlorine dioxide instead of chlorite. We chose the iodate-ferrocyanide-sulfite reaction as a new candidate for studies of pattern formationbecause a pH indicator could be used to visualize patterns that might form. The followingexperiments illustratethe differences between our patternsand those previously observed in reaction-diffusion of systems.A diagram the gel disc reactoris shown in Fig. 3. Gel-filled reactorswere developed several years ago (9) to study reaction-diffusionsystems maintained in well-defined states far from equilibrium. These reactors are now widely used for studyingsustainedpatternsthat arisesolely fromthe interplay diffusion chemical of and kinetics-the gel preventsconvective motion. A thin polyacrylamide layer (0.2 gel mm thick, 22 mm in diameter) is fed diffusively a continuouslyrefreshed by reservoir of chemicals (10). There are two thin membranes between the polyacrylam-

192

SCIENCE * VOL. 261 * 9 JULY 1993