ELECTRICAL FIELD EFFECTS ON CRYSTALLINE PERFECTION OF MBA-NP CRYSTALS BY MAPPING OF BRAGG-SURFACE DIFFRACTION S.L. MORELHÃO*, L.H. AVANCI**, M.A. HAYASHI**, L.P. CARDOSO** *Dept f Applied Physics, University f Sã Paul, Sã Paul, Brazil, mrelha@if.usp.br **Institute f Physics, Univesity f Campinas, Campinas, Brazil ABSTRACT The interactin f weak electrical field with the mlecular diple f the MBA-NP crystals is investigated. Such interactin affects the misrientatin and size f the perfect diffracting regins f the crystal, which is mnitred by mapping the Bragg-surface diffractin cnditin. This diffractin technique has als cnfirmed that the diffracting regins are large enugh t allw primary extinctin and the nn-unifrm msaicity alng the crystal surface. INTRODUCTION. High quality nn-centrsymmetric rganic single crystals, such as 2-(amethylbenzylamin)-5-nitrpyridine (MBA-NP), have exceptinally large secnd rder plarisabilities. The pssibilities f expliting this fundamental prperty are perceived t be extensive because f the almst endless variatins in chemical structure that can be prduced thrugh mdern rganic synthesis. In additin t the exceptinal nnlinearities, all space grups are f the type that als prduces piezelectricity. In terms f cntrlling material nnlinear prperties, characterizatin f the piezelectric tensr has been imprtant [1]. X-ray diffractin techniques are ften used t determine small changes in the crystal lattice under an applied electric field. Recently, a methd based n the X- ray multiple beam diffractin phenmenn was develped fr determining mre than ne piezelectric cefficients frm a single (I) scan (azimuthal scan) [2]. A strng limitatin fr applying the methd in MBA-NP crystals is the unexpected features ften bserved in the scans, such as peak bradening, splitting and shifting in magnitude with can nt be crrelated t any piezelectric prperty f the material. In rder t prperly address this issue, the surce f these features, we have used an X-ray diffractin technique that cnsist in mapping the slid angle arund a three-beam diffractin cnditin [3]. THEORY The three beam diffractin, as well as any multiple beam diffractin, is different frm cnventinal tw beam diffractin because it defines a specific directin, nt a cne f directin, t be fulfilled by the wavevectr f the incident beam. A special three-beam diffractin, called Bragg-surface diffractin (BSD) [4], was chsen t be investigated in this wrk. It presents a strng signal and a shrt extinctin distance fr the secndary surface reflectin [5]. The incident wavevectr directin fr the BSD is illustrated in Fig. 1 by the vectr TO, frm the T-pint t the rigin O f the reciprcal space f a perfect diffracting regin. And, the psitin f the T-pint is given by the intersectin f three spheres (dispersin surfaces) centered at the rigin O and, at the extreme f the F1 01 and H02 diffractin vectrs. The 01 and 02 reflectins are the symmetric Bragg reflectin and the surface reflectin, respectively. The H21 diffractin vectr stands fr the Mat. Res. Sc. Symp. Prc. Vl. 561 1999 Materiais Research Sciety 87
cupling reflectin, which is respnsible fr the energy transfer between the diffracted beams. This energy transfer prcess changes the intensities, and by mnitring the Bragg diffracted beam, the beam frm the 01 reflectin, the ccurrence f BSD can be measured. The perfect regins must be large enugh, f the rder f the extinctin distance fr the surface reflectin, t allw primary extinctin regime f energy transfer. Only under this regime, the T-pint exists fr each perfect regin [2]. Otherwise, the whle crystal defines the T-pint as the cnvlutin between the msaic spreads f the 02 (surface) and 21 (cupling) reflectins. That is the secndary extinctin regime, the dminant regime fr ideally imperfect crystals. In a crystal made up f large perfect regins, the BSD cnditin is given by an aggregate f T-pints, Fig. 2. When the slid angle arund this aggregate is mapped by cmbining c (incidence angle) and (I) (azimuthal rtatin) scans, the spatial distributin f Fig. 1: Bragg-surface diffractin cnditin fr a perfect diffractin regin. The incedent wavevectr directin is given by the TO vectr. perfect regin misrientatin, i.e. the msaic distributin, is visualized. Fr msaic crystals where the secndary extinctin dminates, the BSD prfiles are gaussians in the c scan, as well as in the 4> scan [3]. Mrever, it is imprtant t emphasize that the size f the perfect regin, at the surface diffracted beam directin, is the regin dimensin defining the regime in which the BSD des ccur. EXPERIMENT The mlecular diple f the MBA-NP mlecules is frmed between the nitrgens in the amin and nitr grups, and the resultant mment f diple p is alng the b lattice vectr f the unit cell frmed by tw mlecules [6]. Lattice parameters f the unit cell are a=5.408à, b=6.371a, c=17.968à, a=y=90, and p=94.60. The (001) plane is a natural cleavage plane f crystal. The analyzed sample was cut in dimensins f abut 10x10x3mm 3, with the largest face as the (001) plane, and the lateral faces as the () and (010) planes. The electrical field was applied in the crystal thrugh cnductive spnges, placed between metal plates and the sample lateral faces. The measurements were carried ut at the synchrtrn radiatin surce (SRS), high-reslutin diffractin statin 16.3 [7], Daresbury Labratry, Warringtn, U.K.. The wiggler beam was mnchrmatized t 8.334 kev by tw Si (111) channel cut crystals, and limited by 0.1x0.1 mm 2 crss slit screen. It prvides an incident beam with hrizntal and vertical divergences smaller than 5 arcsec. The linear Aggregate f T-pints Fig. 2: An aggregate f T-pints defines the BSD cnditin fr a msaic crystal with large diffracting regins. 88
plarizatin f the synchrtrn beam was perpendicular t the incidence plane f the 01 reflectin. BSD mapping were perfrmed with step size f 9 arcsec in bth c and 6 axes, and cunting time f 0.5 secnds. A Si (111) channel-cut crystal was used as analyzer fr reciprcal space mapping (RSM) measurements. The high reslutin f the diffractmeter at statin 16.3 allws the cupled scan f the.) and 20 (detectr arm) axes. These measurements give the intensity distributin as functin f the angles 20 and where -(a-20/2. Step size f 18 arcsec was used in the c axis, and the detectr axis cupled, in rder t prvide the same step size in 30 420 '6" g RESULTS Fig. 3: BSD mapping perfrmed at tw adjacent arcas near the center f the sample surface. The The BSD chsen fr ur measurements has angular axes are in arcsec. the 00 6 (Bragg) and T O 3 (surface) reflectins. Their structure factrs are 9% and 31% f the ne fr the 1 1 3 reflectin, respectively. This BSD is bserved at arund w 0=14.431 and 6 0=1.513, if the [0 0] directin is taken as the reference directin (4=0). The first interesting result is the changes in the BSD mapping as a functin f the illuminated area in the sample. In Fig. 3, tw maps perfrmed with the incident beam hitting different areas near the center f the (001) face are presented. The map in Fig. 5a is als frm a different area. The backgrund intensity due t the Bragg diffractin was remved, by first measuring its prfile in a c scan away frm the BSD psitin, and then subtracting it frm the maps. In rder t analyze the msaicity f the crystal with a standard X-ray diffractin technique, RSM f the Bragg reflectin was als carried ut at the statin 16.3. The results frm RSM perfrmed at tw psitin are shwn in Fig. 4, ne with the a axis in the incidence plane (6=90 ) and the ther with the b axis in this plane (4)=0). Due t the incidence gemetry and the relatively deep penetratin f the X-rays in the sample, the diffracting vlumes at each f these measurements are nt the same, even if the illuminated area was maintained during rtatin. Althugh, the tw maps are different, in bth the mapped reflectin is much wider alng the - 80-60- 40-20- 6=0 50 420 O -500-500 500-5 500-5 Fig. 4: Reciprcai space mapping f the 006 Bragg reflectin perfrmed at tw different psitin. 4 and 420 are in arcsec. 89
axis, and the peak bradening alng the 20 axis is the same, f the rder f 50 arcsec. The effects f the electric field nt the BSD mapping were verified by the fllwing prcedure. After fixing the electric cntacts and aligning the sample (H, parallel t the 4) axis), ne mapping was carried ut. Then, the vltage supplier was turned n at V, and anther mapping was perfrmed exactly at the same area. With the electrical field parallel t the plar axis f the crystal, axis b, nt even a small change has been bserved. On the ther hand, when the field was perpendicular t the plar axis, alng the [] directin (a axis), drastic changes can be seen. In Fig. 5, it is shwn the BSD mapping befre and after the electric field be applied. The msaic distributin befre (Fig. 5a) presents a majr peak spread ver arcsec, with a minr preferential alignment, fr Aw>0, alng the diagnal f the mapped angular range. Under the field (Fig. 5b), the peak seems t split up int several peaks, distributed ver mre than 500 arcsec, and the preferential alignment at 45 with the 4) axis is much clear. The rughness in this map results frm the backgrund remving prcess abve mentined. The w scans f the backgrund intensities are shwn in Fig. 6. The first distributin (in Fig. 5a) is nt recvered by tuming ff the vltage r even changing its signal. Due t ur simple set up fr applying the field, it culd nt be re-applied in the [010] directin withut cmprmise the psitin where the beam hits the sample. DISCUSSION The aspect f the BSD mapping in Fig. 3 demnstrates that the BSD cnditin fr the MBA- NP crystal is defined by an aggregate f T-pints. Therefre, it falls in a crystal mdel made up f perfect regins large enugh (in the surface parallel directin) t underg primary extinctin. The absence f a lw intensity gaussian prfile arund the BSD cnditin des eliminate the ccurrence f secndary extinctin (i.e. rescattering amng the regins), and als the existence f sme significant amunt f small regins. Theretically, the primary extinctin distance fr the surface reflectin is null, because it is an extremely asymmetric reflectin with directin csine zer. In practice, even diffracting regins much smaller than 1 gm shuld define a T-pint. But in lw absrptin crystals (like rganic crystals), such small regins wuld give rise t secndary extinctin. Laminar diffracting regins, thin in thickness and large in the in-plane directin, prvide an apprpriated crystal mdel fr the MBA-NP. In this case, the BSD mapping gives the msaic distributin mainly frm stacked regins than frm adjacent nes. The bserved changes in the maps due t crystal translatin, suggest that the in-plane size f the regins shuld be smaller than the incident beam size at the sample surface (-0.15x0.60mm 2), r smaller than gm. The RSM measurements allw just t visualize the prjectin, int the incidence plane, f the misrientatin f the H, diffractin vectrs fr the diffracting regins, regardless their inplane size. The reuslts in Fig. 4 are very characteristic f crystals with msaic distributin, in a magnitude f the distributin bserved in the BSD mapping, arcsec. The spatial nnsymmetric msaic distributin are respnsible fr the changes seen in the RSM perfrmed at (1)=0 and 4)=90. The thickness f the diffracting regins is perhaps the mst imprtant infrmatin that we withdraw frm RSM. Fr instance, the peak width in the 20 directin f 50 arcsec des imply in a thickness f abut 0.6 gm. Finally, let us discuss the electric field effects int the msaic distributin. Each perfect regin has a ttal diple mment, which is the number N f unit cells in the regin times the resultant diple p f the unit cell. In ther wrds, each regin is a dmain with ttal diple mment Np. Under an electric field E, the ptential energy f each dmain is -NE.p, which is 90
prprtinal t their vlume. It is minimum when the field is parallel t the plar axis f the crystal, and maximum when they are anti-,t a) 20 parallel in an unstable equilibrium psitin. In bth cnfiguratins the trque ver the 10 regin is zer. Therefre, the msaic distributin shuld nt be affected by the -t field alng the plar axis, as bserved. On the - ther hand, when E and p are perpendicular, 200 the trque is maximum ver the regins, and they can still realize sme wrk but nt t than NEp. This ptential energy can b) á rtate as well as break a large regin in ti 5- smaller nes, all depends n the regins 200 bundary cnstrains and n their internai 0 bnding energies. -200 In Fig. 5a, the minr preferential -400-200 200 At alignment falis ver the c and 6 psitins in which the H02 diffractin vectrs are tuching Fig. 5: BSD mapping perfrmed at the same the Ewald sphere. Then, it is the track f the sample psitin. a) Befre and b) after the electric secndary reflectin. Fr this BSD mapping, field be stablished alng the [] directin. Aw the diffracting vlume presents sme amunt and A6 are in arcsec. f small regins, as fr exemple, scratch in the surface r a deeper imperfectin. They d nt define a BSD peak, but they are able t rescatter the secndary beam when its directin is belw the surface, i.e. fr Ac>0. Under the E.1 p cnfiguratin, an in-plane, r azimuthal, rtatin f the regins wuld split up the distributin seen in Fig. 5a int anther distributin where the T-pints were aligned alng the 6 axis and, depending n the crystalline perfectin, their respective secndary reflectin track wuld be visible. In Fig. 5b tw tracks are clearly seen. Nte that they are nt cntinuus, and they are determined by well-defined Iw-intensity peaks. It shws that the msaic distributin, r aggregate f T-pints, in Fig. 5a split up in nly tw majr regins (r T-pints). The ther regins in the distributin are nt large enugh t generate T-pints, but they are rescattering the secndary beams f these tw majr regins, and s the tracks are seen. A few f these regins exist since the tracks are nt cntinuus and the secndary extinctin regime f energy transfer amng them did nt define a gaussian T-pint. Due t the length f the tracks we can nt determinie the peaks that are the T-pints, r even if they are in the mapped range. Hwever, by analyzing the w-scan f the backgrund intensity in Fig. 6b, we can figure ut the spatial misrientatin f these T-pints. The scan presents fur peaks, (1), (2), (3), and (4), and by just analyzing this scan, we wuld say that there are fur diffracting regins. The BSD mapping shws that the peaks (3) and (4), separated by A track = 70 arcsec, are just the tracks frm the tw majr regins, respnsible fr the peaks (1) and (2). The ut-f-plane and in-plane misalignments f these regins are Ar., = 150 arcsec and 46.,-= (Am r k)/tan(-45 ) = 80 arcsec, respectively. -Ac, The bundary cnstrains avid a pure in-plane rtatin, but the weak bunding between the (001) planes allws the bserved ut-f-plane rtatin f the regins. In the ut-f-plane rtatin, the ptential energy is nt reduced. Then, it shuld nt ccur, unless t vercme sme cnstrains and allws the in-plane rtatin. 91
CONCLUSIONS 20 In this wrk, we have demnstrated that fr certain directins, even a weak and static electric field can increase the msaic distributin f MBA-NP crystals. It des generate features that are able t jepardize the data analysis frm cnventinal X-ray diffractin techniques such as c scans, c/20 scans, r even scans. The BSD mapping was successfully applied fr crystalline perfectin characterizatin in this material. It has cnfirmed that the crystal presents several large regins, and als that BSD mapping allws the spatial visualizatin f the msaic distributin. ACKNOWLEDGMENT The authrs acknwledge Prf. Jhn N. Sherwd fr the MBA-NP crystal and Dr. Steve Cllins fr valuable help at SRS, statin 16.3. Financial supprt frm Brazilian agencies CNPq and FAPESP are als acknwledged. REFERENCES 15 6_10 5 4 6.3 g2 1 5-150 - -50 O 50 150 200 b) (4) (3) 4e) (arcsec),(2) -500-400 -300-200 - O 200 40) (arcsec) Fig. 6: (O-scans f the backgrund intensities remved frm the BSD mapping in Fig. 5. a) frm Fig. 5a and b) frm Fig. 5b. `.1,U1) 1. P.J. Halfpenny, J.N. Sherwd, and G.S. Simpsn in Orgnic Nn-Linear Optical Materiais: Crystal Grwth and Structural Characterizatin, Kikan Kagaku (Quartely Review f Chemistry Chemical Sciety fjapan) 15, 95-121 (1992). 2. L.H. Avanci, L.P. Cards, S.E. Girdwd, D. Pugh, J.N. Sherwd, K.J. Rberts, Physical Review Letters 81 (24), 5426 (1998). 3. S.L. Mrelhã, L.P. Cards, Jurnal f Applied Crystallgraphy 29, 446 (1996). 4. M.A. Hayashi, S.L. Mrelhã, L.H. Avanci, L.P. Cards, J.M. Sásaki, L.C. Kretly, S.L. Chang, Applied Physics Letters 71(18), 2614 (1997). 5. Surface reflectins are thse with diffracted beam in the surface parallel directin. 6. A. Carenc, J. Jerphangnn, A. Perignd, Jurnal f Chemical Physics 66, 3806 (1997). 7. S.P. Cllins, R.J. Cernik, C.C. Tang, N.W. Harris, M.C. Miller, G. Oszlanyi, Jurnal f Synchrtrn Radiatin 5, 1263 (1998). 92