scieee Open visual document viewer

Dynamics of matter wave solitons

Polo Gómez, Juan

Abstract

[ANGLÈS] We study the implementation of a matter wave bright soliton interferometer formed by a Gaussian potential barrier placed at the center of a harmonic trap potential. After numerically evaluating the transmission coefficient of the potential barrier as a function of the ratio between the kinetic energy of an incident matter wave bright soliton and the height of the barrier, we focus on the case for which the transmitted and the reflected split solitons have the same number of atoms. This beam splitter behavior is fully characterized in terms of the phase and the energies of the two outcome solitons. The recombination process through the barrier is also investigated in terms of the phase difference between the two arms of the interferometer.

Full text

MSc in Pho onics PHOTONICSBCN Uni e si a Poli ècnica de Ca alunya (UPC) Uni e si a Au ònoma de Ba celona (UAB) Uni e si a de Ba celona (UB) Ins i u de Ciències Fo òniques (ICFO) h p://www.pho onicsbcn.eu Mas e in Pho onics MASTER THESIS WORK TOWARDS INTERFEROMETRY WITH MATTER WAVE BRIGHT SOLITONS Juan Polo Gomez Supe ised by D . Ve ònica Ahu inge B e o (UAB) P esen ed on da e 4 h o Sep embe , 2012 Regis e ed a Towa ds in e e ome y wi h ma e wa e b igh soli ons Juan Polo Gomez G up d’ `Op ica, Dep . de F´ısica, Uni e si a Au `onoma de Ba celona, E-08193 Bella e a, Spain E-mail: zes zes zes gmail.com Abs ac . We s udy he implemen a ion o a ma e wa e b igh soli on in e e ome e o med by a Gaussian po en ial ba ie placed a he cen e o a ha monic ap po en ial. A e nume ically e alua ing he ansmission coe icien o he po en ial ba ie as a unc ion o he a io be ween he kine ic ene gy o an inciden ma e wa e b igh soli on and he heigh o he ba ie , we ocus on he case o which he ansmi ed and he e lec ed spli soli ons ha e he same numbe o a oms. This beam spli e beha io is ully cha ac e ized in e ms o he phase and he ene gies o he wo ou come soli ons. The ecombina ion p ocess h ough he ba ie is also in es iga ed in e ms o he phase di e ence be ween he wo a ms o he in e e ome e . Keywo ds: ma e wa e soli ons, in e ac ion wi h po en ial ba ie s, a om in e e ome e . 1. In oduc ion In e e ome y has been o many yea s an impo an scien i ic ool wi h a my iad o applica ions, anging om p ecision measu emen s o undamen al s udies. A omic in e e ome e s ha exploi he wa e cha ac e o a oms allow o e y high sensi i i ies due o he associa ed small wa eleng h. Mo eo e , he inhe en a omic p ope ies like mass o pola izabili y o e new possibili ies o high-p ecision measu emen s o , o ins ance, undamen al cons an s, in e nal o ces o o a ions [1–4]. Speci ically, he use o ul acold a oms, ha can be p ecisely manipula ed a he quan um le el and can be held o long imes in aps [5], p o ides e y high p ecision and sensi i i y [6]. Ne e heless, in linea in e e ome y, he minimum unce ain y in de e mining he phase di e ence is gi en by he s anda d quan um limi [7], i.e., he expec ed o a classical measu emen . I has been shown ha his limi can be su passed, app oaching he Heisenbe g limi , using nonclassical inpu s a es [8–12] o en angled s a es wi hin he in e e ome e in he egime o la ge a om numbe wi h con olled in e ac ions [7]. One po en ial p oblem ha a omic in e e ome e s ace is based on he ac ha he wa e unc ion dispe ses du ing i s e olu ion. A possible solu ion is o use ma e wa e b igh soli ons in Bose-Eins ein condensa es (BEC). Ma e wa e b igh soli ons a e Towa ds in e e ome y wi h ma e wa e b igh soli ons 3 soli a y wa es wi h a ac i e in e a omic in e ac ions, ha p o e o be a he obus when colliding wi h each o he o when in e ac ing in an ex e nal po en ial and p esen a e y impo an cha ac e is ic: hey can p opaga e wi hou dispe sion [13,14]. Wi hin he mean ield app oxima ion, hei dynamics is well desc ibed by he G oss Pi ae skii equa ion (GPE), al hough e ec i e pa icle models also exis and ha e been used o cha ac e ize he cen e o mass ajec o ies o he soli ons [15–17]. No e ha he use o soli ons o in e e ome y [18–26] opens new pe spec i es and allows o add ess undamen al ques ions due o he dual na u e (pa icle and wa e) o soli ons in BEC. Wi hin he quasi 1-D limi , collisions o ma e wa e b igh soli ons wi h ba ie s ha e been la gely s udied in o de o cohe en ly spli he soli on [25] and he ecombina ion o iden ical soli ons h ough a po en ial ba ie has also been add essed [19,20]. In his wo k, we will in es iga e he implemen a ion o a ull ma e wa e b igh soli on in e e ome e o med by a ha monic po en ial ap and a Gaussian ba ie a i s cen e in which he inciden ma e wa e soli on spli s and he wo ou pu spli ones ecombine a e some ime. This se up is cu en ly unde expe imen al s udy in he g oup o P o . R. Hule [27]. In sec ion 2, we in oduce he physical sys em ha we a e conside ing. In sec ion 3, we analyze in de ail he ole o he ba ie as a beam spli e cha ac e izing he wo ou pu soli ons as a unc ion o he sys em pa ame e s. Sec ion 4 is de o ed o he ecombina ion p ocess and in sec ion 5 we p esen he conclusions. 2. Physical sys em We conside a Bose-Eins ein condensa e o 7Li a T= 0 whose dynamics wi hin he mean ield app oach is desc ibed by he 3D ime dependen GPE [28]: i¯h∂ ∂ ϕ( , ) = −¯h2∇2 2m+V( ) + g3D|ϕ( , )|2!ϕ( , ) ; (1) whe e he i s e m o he igh hand side o Eq. (1) co esponds o he kine ic con ibu ion, he second e m is he ex e nal po en ial ha eads V( ) = m[ω2 (x2+ y2) + ω2 zz2]/2, and he las e m is he nonlinea e m, whe e g3D= 4π¯h2asN/m, and N,asand ma e he a om numbe , s-wa e sca e ing leng h and a omic mass, espec i ely. The wa e unc ion is no malized o 1, and he sca e ing leng h is nega i e, as<0, which means ha he in e ac ions a e a ac i e. Fo su icien ly igh adial con inemen (ω ωz), bu allowing ha in e ac ions emain 3D, as(¯h/mω )1/2, i is possible o sepa a e he adial and he axial dynamics using he ansa z ϕ( ) = Ψ(z)(mω /π¯h)1/2exp (−mω [x2+y2]/2¯h), ob aining he educed 1D GPE [26]: i¯h∂ ∂ Ψ(z, ) = −¯h2 2m ∂2 ∂z2+V(z) + g1D|Ψ(z, )|2!Ψ(z, ),(2) whe e g1D= 2N¯hω as[29] and he ex e nal po en ial educes o V(z) = 1 2mω2 zz2. By using C ank-Nicolson me hod in imagina y ime e olu ion we calcula e he g ound s a e o he sys em o s a e o he a pa ame e alues [13]: ω = 2π×710 Hz, ωz= 2π×78 Hz, as=−0.21 nm, m= 1.165 ×10−26 kg and N= 15000 a oms, ob aining a ma e Towa ds in e e ome y wi h ma e wa e b igh soli ons 4 wa e b igh soli on. Once a s able soli on is c ea ed, we displace he cen e o he ha monic po en ial a dis ance z0and in oduce a Gaussian po en ial ba ie cen e ed a z= 0 on which he soli on will collide. The ex e nal po en ial eads now: V(z) = 1 2mω2 z(z−z0)2+Vbez2 2σ2,(3) whe e Vband σa e he s eng h and he wid h o he ba ie , espec i ely. Wi h his con igu a ion we implemen an in e e ome e o ma e wa e b igh soli ons ha can be sepa a ed in o h ee s eps: a) he soli on is spli in wo by colliding wi h he ba ie ; b) he wo spli soli ons e ol e in he ex e nal ha monic po en ial educing hei kine ic ene gy and inc easing he po en ial one un il hey s op and go back in opposi e di ec ions; and c) a ecollision o he soli ons wi h he ba ie occu s a he cen e o he ap. Fig. 1 shows he p ocess schema ically. In o de o cha ac e ize he dynamics o he spli ing o he ini ial soli on and he Figu e 1. Schema ics o he in e e ome e sequence: (a) ini ial si ua ion in which he soli on is displaced wi h espec o he cen e o he ha monic po en ial ap by a dis ance z0; (b) he wo spli soli ons ob ained a e he collision wi h he ba ie sepa a e om each o he (c) he wo soli ons e u n o he posi ion o he ba ie and ecollide. e olu ion o he sys em a e wa ds, we calcula e he expec ed alue o he posi ion and momen um o he soli on be o e i s collision wi h he ba ie and o he wo soli ons ob ained a e he spli ing: hzii=K−1 iZCi dz|Ψ(z, )|2z, (4) hpii=−i¯hK−1 iZCi dzΨ∗(z, )∂ ∂zΨ(z, ),(5) whe e Ki=ZCi dz|Ψ(z, )|2and Cideno es he egion con aining he soli ons: being i=I, R, T o Inciden , Re lec ed and T ansmi ed egions, espec i ely. We also de ine he posi ion and momen um unce ain y: ∆zi= (hz2 ii−hzii2)1/2,(6) ∆pi= (hp2 ii−hpii2)1/2,(7) whe e: hz2 ii=K−1 iZCi dz|Ψ(z, )|2z2,(8) Towa ds in e e ome y wi h ma e wa e b igh soli ons 5 hp2 ii=−¯h2K−1 iZCi dzΨ∗(z, )∂2 ∂z2Ψ(z, ).(9) The numbe o pa icles o each soli on is gi en by Ni=NKiand he ene gy pe pa icle eads: ei=Ei N=ZCi dz   ¯h2 2m ∂ ∂zΨ(z, ) 2 +V(z)|Ψ(z, )|2+g1D|Ψ(z, )|4 ,(10) whe e he i s e m o he igh hand side co esponds o he kine ic ene gy Ek, he second one is he ex e nal po en ial ene gy, Eho, and he las one is he in e nal ene gy due o he a om-a om in e ac ion, Ein . No e also ha by di ec in eg a ion o he GPE unde he s a iona y ansa z Ψ(z, ) = Ψ(z) exp (−iµ /¯h), whe e µis he chemical po en ial, we ob ain µ=Ek+Eho + 2Ein [28]. 3. Spli ing p ocess In his sec ion we analyze he main aspec s ha a ec he dynamics o he spli ing p ocess o a ma e wa e b igh soli on unde he collision wi h a Gaussian po en ial ba ie . 3.1. T ansmission By an ins an aneous displacemen o he apping po en ial, he cen e o he soli on is placed a −z0and he soli on acqui es a po en ial ene gy ha , as p edic ed by pa icle models [30,31], is ully con e ed in o kine ic ene gy once he soli on eaches he cen e o he po en ial. The e o e, he kine ic ene gy o he soli on when in e ac ing wi h he ba ie is Eho =1 2mω2 z(z−z0)2. We ha e checked ha his esul is in ull ag eemen wi h Eq. (10). Fig. 2(a) shows he nume ically e alua ed ansmission coe icien as a unc ion o he a io be ween he kine ic ene gy o he soli on and he heigh o he po en ial ba ie o a ixed po en ial ba ie heigh o Vb= 17.14¯hωz(squa es), and o a ixed kine ic ene gy o Ek= 10.83¯hωz(ci cles); in bo h cases σ= 0.5µm. The di e en beha io s ob ained close o T= 1 can be unde s ood by aking in o accoun ha he changes in he s eng h o he po en ial o in he kine ic ene gy o he inciden soli on do no p oduce exac ly he same e ec due o he di e en a ea ha he soli on has o pene a e in o de o pass h ough he ba ie . Ne e heless, a common ea u e is he high slope in he egion whe e he ansmission coe icien is 0.5, i.e. whe e he ini ial soli on spli s in o wo soli ons wi h he same numbe o a oms, which makes his poin e y sensi i e o he ini ial condi ions. Fig. 2(b) shows he ansmission as a unc ion o he wid h o he ba ie (ci cles) o a ixed a io Ek/Vb= 0.63. As expec ed, he inc ease o he wid h o he ba ie co esponds o a dec ease in ansmission. In Fig. 2, we also plo he analy ical ansmission coe icien o a ec angula ba ie (do ed-dashed line), which Towa ds in e e ome y wi h ma e wa e b igh soli ons 6 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 T ansmission T Ek/Vb (a) 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 0 0.2 0.4 0.6 0.8 1 1.2 T ansmission T σ(µm) (b) Figu e 2. T ansmission as a unc ion o (a) he a io be ween Ek/Vband (b) he wid h o he ba ie ob ained by ixing Ek= 17.14¯hωz(ci cles) o Vb= 10.83¯hωz(squa es). The analy ic esul o a ec angula ba ie (do ed-dashed line) and o he WKB app oxima ion (dashed line) in he linea case a e also plo ed. The pa ame e s o he sys em a e N= 15000, ωz= 2π×78 Hz, ω = 2π×710 Hz, as=−0.21 nm. In (a), σ= 0.5µm and in (b) Ek/Vb= 0.63. eads [32]: TuE<Vb=1 1 + V2 bsinh2(k1(2σ)) 4E(Vb−E) TuE>Vb=1 1 + V2 bsin2(k1(2σ)) 4E(E−Vb) ; (11) whe e k1=q2m|Vb−Ek|/¯h2; and he Wen zel−K ame s−B illouin (WKB) app oxima ion (dashed line). The WKB app oxima ion es ima es he ansmission coe icien , o small alues o Tand o po en ials ha do no change ab up ly, aking in o accoun he shape o he po en ial and conside ing he ac ha he ampli ude and he phase o he wa e unc ion a y slowly. The de ailed analy ical de i a ion o he ansmission coe icien can be ound in [32]: TWKB ≃e1 ¯hRb adz√2m(V(z)−Ek)+1 4e −1 ¯hRb adz√2m(V(z)−Ek)−2 ; (12) whe e aand ba e he z-axis poin s whe e he kine ic ene gy and he ba ie coincide, which o he Gaussian shape ba ie a e b=−a=q2σ2ln (Vb/Ek). No e ha hese analy ical ansmission unc ions (Eq. (11) and (12)) a e only alid in he linea egime. The e o e, in o de o be alid o ou nume ical simula ion Ekhas o be la ge enough o he wid h o he ba ie small enough o minimize he in e ac ion ime o he soli on wi h he ba ie . Ne e heless, ou speci ic se up limi s hese wo pa ame e s: he minimum wid h o he ba ie is ixed by he di ac ion limi , since he Gaussian po en ial is expe imen ally c ea ed by using a ocalized lase beam wi h a beam wais equal o σand he kine ic ene gy is limi ed by he ini e po en ial ap displacemen , z0. Mo eo e , he e is an addi ional aspec ha makes he compa ison wi h he analy ical unc ions di icul : he soli on is a wa e packe desc ibed by a mac oscopic wa e unc ion, whe eas he analy ical cases conside plane wa es. The e o e, while plane wa es ha e a Towa ds in e e ome y wi h ma e wa e b igh soli ons 7 well de ined momen um, he soli on has a e y well de ined posi ion and he momen um unce ain y is e y la ge, ul illing he unce ain y ela ion ∆xI∆pI≥¯h/2 close o he sa u a ion alue. Fo ins ance, o an inciden soli on wi h N= 15000 he unce ain ies using Eq. (6) and (7) a e ∆xI= 0.583 nm and ∆pI/¯h= 0.898 1/nm, gi ing ∆xI∆pI/¯h= 0.524. In spi e o he discussed limi a ions, he ec angula po en ial ba ie i s a he well wi h he nume ical esul s (see Fig. 2). Howe e he slope o he ansmission unc ion o he ec angula ba ie nea T= 0.5, egion o in e es o in e e ome ic applica ions, is lowe han he one ob ained nume ically wi h he Gaussian ba ie . The WKB app oxima ion ag ees wi h he nume ical ansmission coe icien when T→0. Howe e his egion is ou o he egime o in e e ome ic applica ions whe e he aim is o spli he soli on in wo iden ical soli ons (T= 0.5). Fig. 3 shows he po en ial s eng h Vbas a unc ion o Ek o which T= 0.5 and o wo di e en alues o he numbe o pa icles N. We can obse e ha he dependence is app oxima ely linea and independen o N o la ge kine ic ene gies. Ne e heless, o small kine ic ene gies he nonlinea i y s a s o play an impo an ole and b eaks he linea dependence (see inse o Fig. 3). In he ollowing we conside a soli on wi h N= 15000. This alue has been chosen in o de o ob ain wo soli ons in he spli ing p ocess i.e, he nonlinea i y is su icien o compensa e he dispe sion a e he spli ing, whe e he numbe o a oms o each o he ou come soli ons is N/2. 0 10 20 30 40 50 60 70 80 0 10 20 30 40 50 60 70 Vb/¯hωz Ek/¯hωz 0.3 0.45 0.6 0.75 0.9 0 20 40 60 Ek/Vb Ek/¯hωz Figu e 3. Po en ial ba ie heigh as a unc ion o he kine ic ene gy o he inciden soli on o a ixed ansmission coe icien T= 0.5, o an inciden soli on o N= 15000 (squa es) and N= 7500 (ci cles). The inse shows he a io Ek/Vbas a unc ion o he kine ic ene gy. The es o he pa ame e s o he sys em a e ωz= 2π×78 Hz, ω = 2π×710 Hz, as=−0.21 nm and σ= 0.5µm. 3.2. Ene gy and Phase The sensi i i y o an in e e ome e has a s ong dependence on how he phase di e - ence is measu ed and o how long he spli pa hs can accumula e phase wi hou losing cohe ence [6]. The e o e, in sec ion 4, we will analyze he phase e olu ion o he inciden Towa ds in e e ome y wi h ma e wa e b igh soli ons 8 soli on and o he wo spli soli ons p oduced by he in e ac ion wi h he ba ie . Ne e heless, in his sec ion we s udy i s he phase di e ence be ween he ansmi ed and e lec ed soli ons, in oduced by he in e ac ion wi h he ba ie du ing he spli - ing p ocess, by se ing he ha monic ex e nal po en ial equal o ze o. Fig. 4(a) shows 0.5 0.75 1 1.25 1.5 1.75 2 2.25 20 30 40 50 60 70 80 ∆φ( ad) Ek/¯hωz (a) 58 60 62 Ek/¯hωz (b) -17 -15 -13 3 3.5 4 4.5 5 5.5 6 Ein /¯hωz (ms) Figu e 4. (a) Phase di e ence be ween he ansmi ed and he e lec ed soli on as a unc ion o he kine ic ene gy o he inciden soli on o T= 0.5. (b) Kine ic and in e nal ene gy o a o al e lec ed soli on wi h ini ial kine ic ene gy Ek= 58.85¯hωz. The solid (dashed) line shows he ime a e aged ene gy be o e (a e ) he collision wi h he ba ie o bo h Ekand Ein . The in e al o ime in which he collision wi h he ba ie occu s is ma ked wi h he black egion. The pa ame e s used a e: N= 15000, ω = 2π×710 Hz, ωz= 0 Hz, σ= 0.5µm and as=−0.21 nm. he linea dependence o he phase di e ence be ween he ansmi ed and he e lec ed soli on, in oduced by he e lec ion p ocess, wi h he kine ic ene gy o he inciden soli on. We conside N= 15000, po en ial wid h σ= 0.5µm and a heigh such ha he T= 0.5 condi ion is ul illed, i.e., he wo esul ing soli ons ha e he same numbe o a oms. By using exp ession (10), one can also calcula e he e olu ion o he di e en con ibu ions o he ene gy du ing he spli ing p ocess. We ob ain ha he e lec ed soli on has less kine ic ene gy han he ansmi ed one, and mo eo e , in he e lec ion p ocess some kine ic ene gy is ans e ed o in e nal ene gy. The e o e, he ansmi ed and e lec ed soli ons a e no iden ical ene ge ically. As an example, an inciden soli on wi h an ini ial ene gy Ek= 20.08¯hωzand Ein =−14.84¯hωzspli s in o wo soli ons wi h he same numbe o pa icles when in e ac ing wi h a ba ie wi h Vb= 20.57¯hω and σ= 0.5µm. The kine ic (in e nal) ene gy o he e lec ed and he ansmi ed soli on is EkT= 5.20¯hωz(Ein T=−1.73¯hωz) and EkR= 3.57¯hωz(Ein R=−1.80¯hωz), espec- i ely. Al hough he o al ene gy is conse ed, EkR+Ein R+EkT+Ein T=Ek+Ein , i is clea ha he e lec ed soli on has less ene gy han he ansmi ed one. An ex eme case occu s o he o al e lec ion o a soli on o N= 15000 whe e he ans e o kine ic o in e nal ene gy is clea ly obse ed (Fig. 4(b)): he solid (dashed) line shows he ime a e aged ene gy be o e (a e ) he collision wi h he ba ie o bo h Ekand Ein . I is wo h no icing ha he e lec ion p ocess can also induce b ea hing exci a ions in which some ac ion o he ene gy is exchanged pe iodically be ween Ekand Ein as shown in Towa ds in e e ome y wi h ma e wa e b igh soli ons 9 Fig. 4(b), o he o al e lec ion case (T= 0). 4. Recombina ion p ocess In his sec ion we add ess he las s ep o he p oposed implemen a ion o a ma e wa e b igh soli on in e e ome e , i.e., he ee e olu ion o he spli soli ons and hei ecombina ion a he cen e o he ha monic po en ial (Fig. 1(c)). We conside he case desc ibed in sec ion 3.2 in which one soli on collides wi h a Gaussian ba ie and spli s in o wo soli ons wi h he same Nand a ela i e phase induced by he po en ial ba ie . Du ing hei e olu ion in he ha monic po en ial, he phases o he spli soli ons e ol e di e en ly due o he di e en eloci y acqui ed du ing he spli ing p ocess (see sec ion 3.2) as shown in Fig. 5. The inciden soli on inc eases i s phase as i p opaga es owa ds he ba ie . A e he collision wi h he ba ie a = 4.34 ms, wo soli ons appea wi h a phase di e ence (see inse Fig.5) as desc ibed in sec ion 3.2. A e hei e olu ion in he ha monic po en ial pe o ming a dipole oscilla ion, he wo spli soli ons collide a he posi ion o he ba ie a = 10.40 ms. The ou come o his collision depends on he ela i e phase be ween he wo soli ons ha ecombine. In gene al, one ob ains wo soli ons wi h di e en numbe o pa icles a e he ecombina ion. The e o e, he accumula ed ela i e phase in he wo a ms o he in e e ome e can be de e mined by measu ing he ela i e pa icle numbe o he wo ou going soli ons [25]. Fig. 6 shows di e en ou pu s o he s udied a omic in e e ome e co esponding o 42 44 46 48 50 52 54 56 58 60 62 4 5 6 7 8 9 10 φ( ad) (ms) 45 46 47 48 4.2 4.4 4.6 φ( ad) (ms) Figu e 5. Time e olu ion o he phase o he inciden soli on (black line) be o e he in e ac ion wi h he ba ie and o he e lec ed (blue line) and ansmi ed ( ed line) soli on c ea ed in he spli ing p ocess du ing i s dipole oscilla ion in he ha monic ap. The pa ame e s used a e N= 15000, ωz= 2π×78 Hz, ω = 2π×710 Hz, as=−0.21 nm, Ek= 10.74¯hωz,σ= 0.5µm, he po en ial used o spli ing in o wo soli ons wi h he same numbe o pa icles is Vb= 17.70¯hωz. The ime in e al in which he collision wi h he ba ie occu s is ma ked wi h he black egion. The inse shows an enla ged iew o he spli ing p ocess. di e en phase di e ences be ween he ecombined soli ons. The phase di e ence in his