J . Chem. SOC., Furaday Trans. 1, 1986, 82, 213-232 Influence of Cocations on the Location and Coordination Geometry of Cu2" in M+A Zeolites (M+ = Na, K, T1, Rb, Cs, NH, and CH,NH,) An Electron Spin Resonance and Electron Spin-Echo Modulation Spectroscopic Study Mysore Narayana and Larry Kevan" Department of Chemistry, University of Houston, Houston, Texas 77004, U.S. A . The changes in the locations and coordination geometries of Cu2+ in various cationic forms of A zeolite has been examined by electron spin resonance and electron spin-echo modulation techniques. It is observed that Cu?+ interacts with six deuterons (three water molecules) in NaA in the x-cages, while it interacts with only two deuterons (one water molecule) in the w a g e in the other monovalent ion-exchanged A zeolites.The changes in the coordination of Cu2+ as a function of hydration conditions have been monitored and an anomalous irreversible change in the immediate environ- ment of Cu2+ in KA and TlA is observed. After dehydration at T > 353 K in KA and T1A rehydration does not restore the spectra seen in fresh samples. Instead the spectra in such rehydrated samples are identical to those in NaA. It is also found that on ammonia-exchange and partial dehydration Cu2+ exhibits markedly different behaviour in NaA when compared with exchange in NaA or in NH,A zeolites. The difference in the initial Cu2+ coordination in NaA as against the other A zeolites is explained in terms of a-cage crowding by the monovalent cations. ~ ~ ~~~ Because of the efficiency exhibited by copper-exchanged zeolites in several catalyst- promoted reactions such as oxidation,l cracking2 and is~merization,~ numerous studies have been made to characterize the active species in these zeolites.A prerequisite for such characterization of Cu2+ is a detailed knowledge of its location and immediate environment. Several electron spin resonance (e.s.r.) reports have been published about the behaviour of Cu2+ in ~eolites.~-* Except for an indication that Cu2+ ions are distributed over several sites in the zeolite, e.s.r. has not been successful in unambiguously determining details of the immediate environment in these polycrystalline samples. In recent years the pulsed magnetic resonance technique, electron spin-echo modulation (e.s.e.m.) spectroscopy has been successfully exploited9-13 to garner critical information about short-range order around paramagnetic species which cannot be obtained from continuous-wave e.s.r. measurements.In particular. interactions of cupric ion with various adsorbates in activated NaA zeolite have been e1ucidated.l4 From Mossbauer studies Dickson and Rees15 concluded that in partially hydrated Fez+-exchanged A zeolites the metal ion is in tetrahedral coordination. We have been able to show by analysis of 133Cs modulation16 and deuterium mod~lationl~ for Cuz+ in several A zeolites that Cu2+ exists in an unusual tetrahedral coordination even in fully hydrated zeolite samples. Pafomov et aI.l* studied the e.s.r. spectra of Cu2+ in A zeolites with different cocations but were unable to draw any conclusions about the site location or coordination geometry of the majority of the observed Cu2+ species.In this paper we present results of e.s.r. and e.s.e.m. studies on Cu2+ under various 213214 Coordination Geometry of Cu2+ in M+A Zeolites hydration conditions with different cocations in A zeolites. It is clearly demonstrated that the bulkier cations force severe changes in the coordination sphere of Cuz+ in these A zeolites. Experiment a1 Linde 4A zeolite was washed with 0.1 mol dmW3 sodium carbonate solutions and ion-exchanged with 0.05 mol dm-, solutions of T1+, Cs+, Rb+, NH,t and CH,NH: with either NO; or C1- as the counterion to obtain Tllo.5 Na,.,-A, Cs,Na,-A, Rb,,,Na,-A and (NH,),,-A. The numbers are with reference to the total number of exchangeable cations in the primitive unit cell of A zeolite and are obtained by atomic absorption analysis.The actual number of methylammonium cations has not been analysed, but from the literature it is known that at best only ca. 77% of the sodium ions can be replaced by CH3NH$.l9 Linde 3A was washed with 0.1 mol dm-, potassium carbonate solutions to obtain K,,-A. These zeolites will be referred to as NaA, TlA, CsA, RbA, NH,A and CH,NH,A. Ca. 0.3% of Cu2+ was exchanged into these zeolites at room temperature using lop4 mol dm-, cupric nitrate solutions at a pH of 5.2-5.5. Several samples of NaA were exchanged with ammoniacal copper nitrate solution ( mol dm-, pH 1 1) which yielded pale blue samples, indicating some complexation of Cu2+ with NH,. This colour disappeared on leaving the zeolite-in air (relative humidity 60%) for 1 day.Such ammoniacally exchanged samples will be referred to as Cu(NH,),-NaA. For the various e.s.r. and e.s.e.m. experiments samples were subjected to the following treatments. (1) For studying the coordination changes by e.s.r., zeolite samples were loaded in 3 mm 0.d. Suprasil quartz tubes and evacuated in a vacuum line to a residual pressure of 2 x lop5 Torrt for various time intervals at 293 (room temperature), 323,373,673 and 773 K, and sealed at that temperature. ( 2 ) Samples heated at any of the above-mentioned temperatures were exposed to H,O or D,O vapour (ca. 25 Torr at room temperature) for 4 7 h before sealing. For samples treated at 673 and 773 K oxidation with 400 Torr of oxygen for 3 h was necessary to regenerate all of the Cuz+, as part of it was reduced on dehydration at such temperatures. Considerable difficulty was experienced in preparing rehydrated Cu-RbA samples.When RbCl was used for ion exchange some of the samples lost crystallinity on dehydration at 773 K.,O Also, the RbA zeolite turned out to be an excellent scavenger of Mn2+ in the RbCl solutions, which interferes considerably with both e.s.r. and e.s.e.m. experiments. None of the other cationic forms (Na, K, T1, Cs) showed either loss of crystallinity or specific scavenging ability for Mn2+. (3) For e.s.e.m. measurements samples were soaked in D,O and dried in a vacuum desiccator; this was repeated three times to ensure complete deuteration. Then the samples were sealed in 3 mm 0.d.quartz tubes. Some samples were deuterated by exposing them to D,O after evacuation in a vacuum line at different temperatures. (4) KA and T1A zeolites were also prepared by using the respective acetate solutions from NaA, as some anomalous results were observed in the dehydration-rehydration cycle. These results were reproducible, irrespective of the type of exchange solution used indicating different behaviour of these two zeolites compared with the others. This point will be elaborated in a later section. E.s.r. measurements at room temperature or at 77 K were made with a modified Varian E-4 spectrometer. E.s.e.m. spectra at 4.2 K were recorded with a home-built spectrometer considerably modified from the earlier reported version.21 Some of these modifications have been already published elsewhere.22 A Tracor Northern signal averager (TN17 10 or TN 1550) was modified to start the scan of the pulse intervals in the auto mode and also to drive a 0-180" phase shifter at the input of the high-power microwave amplifier.Phase-shifting the microwave pulse on every alternate scan by 180" cancelled out the base- line drift and also reduced the dead time of the spectrometer.22b The other major t 1 Torr = 101 325/760 Pa.M . Narayana and L. Kevan 215 additions were a double balanced mixer and an Ortec 726 timing amplifier on the receiver side, both of which considerably improved the signal-to-noise ratio. Finally the digitized data were transferred from the signal averager to a Tektronix 4052 microcomputer system, where they were processed for plotting in the time domain and could also be fast- Fourier-transformed using a Tektronix program. Theory and Analysis The theory and analysis of electron spin echoes and the nuclear modulations on the echo decay envelopes is well do~umented.~-ll When a paramagnetic system is subjected to resonant microwave pulses in suitable sequence, microwave echoes are generated due to reformation of macroscopic magnetization. When the time intervals between the microwave pulses are swept these echoes decay due to various relaxation processes.The weak dipolar, isotropic hyperfine and quadrupolar interactions experienced by the un- paired spin with the surrounding magnetic nuclei often show up as periodic variations in the echo decay envelope.These modulations are characteristic of the Larmor frequencies of the interacting nuclei and thus can be quantitatively analysed to give information about the radial distribution of the interacting nuclei, their number and the contact hyperfine coupling when present. l3 Deuterium nuclei have a slower precession frequency, and the modulation from them can be more accurately analysed than that from protons. Thus, in most of the samples used in this study H,O has been exchanged with D20. In fact, when adsorbate molecule geometry is the object of the study one can selectively deuterate a molecule like CH,OH to determine accurately the orientation of the with respect to the unpaired spin. Recently generalized expressions in the zero quadrupole interaction approximation were obtained by Dikanov et al.23 for two- and three-pulse electron spin-echo modulation.These expressions can be briefly summarized as follows for three-pulse echo experiments. For an unpaired spin with S = 1/2 interacting with a nucleus of spin I = 1/2 and for a nucleus of spin I where a and P are the electron up and down states. The final expression for the stimulation echo amplitude is J q T , T ) = 1/2[vy(r, T)+ Vf(Z, T ) ] . In these expressions K = ( C O ~ B/w, COD)^ C O ~ , ~ = [(CO~ f A/2)’+ B2/4]”’ A = D(3 cos20- 1)+2na B = 3 0 sin0 C O S ~ D = ggn PPn/fir3* If the unpaired spin is interacting with N nuclei, the overall modulation is given by21 6 Coordination Geometry of Cu2+ in M+A Zeolites To consider the interactions with N equivalent nuclei with uncorrela use the methodology suggested by Mim~.~y lo s < w, T))# = n < I / I k ( Z , T)n.k=l In these expressions Y is the electron-nucleus vector making an angl ed orientations we ! 0 to the direction of the external magnetic field Ho and a is the isotropic hyperfine coupling. The basic approximations made in obtaining and using these expressions are ( I ) the local dipolar interaction is smaller than the nuclear Zeeman terms, (2) the nuclear quadrupolar interaction is smaller than the dipolar term, (3) the unpaired electron spin is sufficiently localized to fit a point-dipole description and (4) the nuclei are considered to be uncorrelated as far as the interacting electron spin is concerned, so that spherical averaging can be used. In most cases these are all excellent approximations, except when the interacting nucleus has a strong nuclear quadrupole moment.In such a case even with fairly regular symmetry at the site of the nucleus one has the quadrupolar interaction almost of the same order as the dipolar term, and the Hamiltonian will have to be exactly diagonalized to analyse the modulation rigorously. For example, in a recent two- dimensional n.m.r. study Samoson and LippmaaZ4 have shown that even in hydrated NaA zeolite there is sufficient anisotropy in the electric field gradient at the site of aluminium nuclei to give a quadrupole coupling of 1.1 MHz with an asymmetry parameter of 0.75. The average dipolar interaction in most cases is a few MHz, and thus one can no longer ignore the quadrupolar interaction in such a zeolite sample.However, in principle one can measure strong quadrupole interactions directly when present from the fast Fourier transform of the time-domain electron spin+cho modulation. This has been suggested in a study of the chlorophyll-a cation in glassy systems by Dikanov et aZ.25 In the analysis of the time-domain data the decay inherent in the modulation spectra is fitted to a polynomial and divided out before comparing with the theoretically simulated spectra as a function of: n, the number of equivalent interacting nuclei, Y, their distance to the unpaired electron spin and a, the isotropic hyperfine coupling constant. For good-quality data the errors in determination of the parameters are : n to the nearest integer for n 5 10, Y to kO.01 nm when < ca.0.45 nm and a to f 10%. If a satisfactory fit of at least three different spectra with T as the variable for three-pulse echoes and T as the variable for two-pulse echoes is not obtained for the same set of parameters n, Y and a, then a two-shell model is used in which two groups of inequivalent nuclei are considered to be interacting with the paramagnetic species. Results Cu-NaA Several reports are already available on the e.s.r. spectrum of Cu2+ in NaA zeolite.E However, only recently have the specific characteristic coordination geometries under various hydration conditions been identified.27c In freshly prepared as well as in rehydrated samples there is only one type of Cu2+ present under low loading conditions. Sometimes, if the rehydration time is shortened, one can simultaneously see Cu2+ species coordinated to zero and three water molecules.In fig. 1 the e.s.r. of spectrum of one such rehydrated sample is shown, along with that of a partially dehydrated sample. In the latter a conspicuous feature of the spectrum is a third Cu2+ species with ‘reversed’ g values8 (g,, < gl). The three types of Cu2+ thus far described will be referred to as follows, in which the roman numeral subscript denotes the number of coordinated waters: CUIII for a copper ion coordinating to three zeolite oxygens and three water molecules, CuII for a trigonal bipyramidal copper in the six-membered ring windowsM. Narayana and L. Kevan 217 Y 91' c urn) Fig. 1. E.s.r. spectra at 77 K of (a) partially dehydrated and (h) partially rehydrated Cu-NaA.The sharp line at g = 2.004 is a colour centre in the e.s.r. Dewar; (a) was recorded on a sample dried at 383 K in air for 24 h, while (b) was recorded on a sample dehydrated and oxidized at 673 K before exposing it to water vapour for 2 h at room temperature. coordinated to three zeolite oxygen atoms and two water molecules, and Cu, for the copper species in the fully dehydrated zeolite coordinating to only three six-membered-ring oxygen atoms. CU,,, is partly converted into CUII on evacuating Cu-NaA samples for a prolonged time at ambient temperature or by drying in air at 383 K for 2 h. The most interesting feature of this CUII is that it cannot be easily dehydrated further to form Cu, or rehydrated to form CUIII.Prolonged exposure of the partially dehydrated sample to water vapour does not restore CUI,, as the sole copper species in the zeolite; this can only be achieved by evacuation at T 2 383 K and then rehydration at room temperature. CU-KA, Cu-TlA Unlike in NaA, freshly prepared Cu-KA samples exhibit a completely different dominant copper species. Using e.s.e.m. analysis of the deuterated samples this has been identified by us1' as Cu2+ coordinating to only one water molecule. This will be referred to as Cu,. In the fresh samples of Cu-KA a small amount of CUIII is always present, but it disappears if the sample is evacuated at 323-353 K and then rehydrated. Upon such a treatment Cu, reforms as the sole Cu2+ species in the zeolite. Note that after 2 h of evacuation at room temperature Cu, completely disappears.It is converted into CUII, the g and A values of which are identical to those of CuII in NaA. However, if the samples are evacuated at 7' 2 383 K and then rehydrated, there is no trace of Cu,, a species similar to CU,,, being the only copper complex observable by e.s.r. spectroscopy. This is218 Coordination Geometry of Cu2+ in M+A Zeolites 1 77 K I u I * w S p , ~ 200 G , H Fig. 2. Comparison of e.s.r. spectra at 293 and 77 K of (a) fresh Cu-NaA, (6) fresh Cu-KA and ( c ) Cu-KA rehydrated at room temperature after dehydration at 383 K. Note the formation of Cu,,, instead of Cu, in rehydrated Cu-KA and that only fresh Cu-KA shows different e.s.r. spectra at 77 and 293 K. demonstrated in fig. 2, where the 293 and 77 K e.s.r.spectra of Cu-NaA are compared with those of Cu-KA before and after the activation-rehydration cycle. A similar behaviour of Cu2+ is observed in TlA, the only exception being that in the rehydrated samples Cu, is present in traces while in KA it is no longer present. Fig. 3 shows the e.s.r. spectra of Cu-TlA before and after the activation-rehydration cycle. The anomalous disappearance of Cu, on activation is further elaborated in the next section. In Cu-TlA the reversed g-value spectrum also appears on partial dehydration and has features identical to those in Cu-NaA and in Cu-KA. Cu-RbA and Cu-CsA In both these zeolites the dominant copper complex is CuI.16 In both it disappears on evacuation at room temperature and Cu,, is formed. However, dehydration at any temperature followed by exposure to water vapour completely restores Cu, as the major species.As mentioned in the Experimental section, some of the samples of RbA were irreversibly transformed into amorphous material on dehydration at 773 K. No such problem was encountered with CsA samples. Note that in neither of these was CU,,, observed either in freshly made samples or in rehydrated samples. The e.s.r. parameters of Cu2+ in freshly prepared, variously dehydrated and rehydrated samples of several of these A zeolites are given in table 1 . Note that Cu-RbA has theM . Narayana and L. Kevan 219 1 - 200 G , H Fig. 3. Comparison of e.s.r. spectra at 77 K of (a) freshly prepared Cu-T1A and (b) Cu-T1A rehydrated at room temperature after dehydration at 483 K.For increased clarity the spectra in the gl region are also recorded at higher gain. Note the decrease in the intensity of Cu, in the rehydrated sample associated with a corresponding increase in Cu,,,. highest gI1 value and also the lowest hyperfine coupling among the Cu, species seen in the other A zeolites. It was rather difficult to determine the gl value for Cu, in all these zeolites because of severe overlap of the e.s.r. features of all three species in the gl region of the e.s.r. spectra. In fact, in ref. (17) the gl signal of CUII, was mistakenly identified as that of Cu,. It is unclear whether those spectra are axially symmetric; a rhombic symmetry assignment may also be made. Cu-NH,A and CU-CH,NH,A In NH,A and CH,NH,A the major copper species is CU, as in KA and T1A.If partial dehydration is performed by evacuation between room temperature and 353 K, or by drying in air at 383 K, Cu, is restored as the dominant species on rehydration at room temperature. Partial dehydration produces a ‘reversed’ g-value spectrum (gl > g,,), but with a gl different from that of the reversed g-value spectrum denoted as CuII, which is formed in NaA, KA, T1A and RbA. In the case of NH,A an additional hyperfine splitting is seen in fig. 4 in the g 2.1 region, which is probably due to interaction with two nitrogen nuclei (14N, Z = 1). The appearance of this additional splitting was found t o be completely reversible in Cu-NH,A subjected to several dehydration-rehydration cycles with T > 353 K. On dehydration at higher temperature the e.s.r.signal intensity falls off rapidly and is not restored on rehydration. In CH,NH,A another ‘reversed’ g-value spectrum with yet a different gl is obtained on evacuation at T > 353 K or by drying in air at 383 K. In fact in CH,NH,A two different copper species with ‘reversed’ g values are observed as shown in fig. 5. However, no additional splittings are seen in either of those two species. As in NH,A, the Cu2+ signal intensities in CH,NH,A fall off rapidly when the dehydration temperatures are higher than 353 K, and the Cu2+ signals are not restored on rehydration.220 Coordination Geometry of Cu2+ in M+A Zeolites Table 1. E.s.r. parameter9 of Cu2+ at 77 K in several A zeolites under various hydration conditions zeolite NaA KA T1A RbA NH,A CH,NH,A Cu(NH,),-NaA major species hydration statusb gll g, A1 minor species probable Cu2+ configurations" 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 4 1 2 4' 1 2 4' 2.352 2.352 2.388 2.349 2.479 2.004 2.391 2.345 2.469 2.004 2.394 2.342 2.502 2.004 2.391 2.498 2.474 2.004 2.471 2.49 2.004 2.491 2.309 2.004 1.998 2.067 2.067 2.070 1.069 2.129 2.289 2.065 2.067 2.129 2.293 2.067 2.065 2.130 2.290 2.067 2.134 2.130 2.243 2.130 2.134 2.71 2.130 2.060 2.23 1 2.289 159 158 131 133 100 84 133 168 103 83 130 171 90 85 130 90 102 132 102 95 149 95 178 126 93 - 2.004 - - 2.345 2.349 - - 2.352 2.346 2.469 2.350 - - - ~ 2.022 2.3 17 2.022 2.024 1.959 2.020 - - 2.350 - 2.289 __ - 2.070 2.070 - - - 2.060 2.130 2.070 ~~ - - ~~ 2.270 2.060 2.280 2.289 2.23 2.292 - - 2.070 - 84 - - 166 154 - - 163 165 103 158 - -~ - - 102 161 102 134 105 131 - ~.15 a Errors in g values are k0.003 and in A values are 1 x lo-* crn-l; units of A are 1 x lo-, ern-'. Hydration status numbers mean the following: (1) freshly prepared samples; (2) partially dehydrated, either by evacuation at room temperature or by drying in air at 383 K; (3) fully dehydrated at 673 K and oxidized; (4) rehydrated after (3); (4') rehydrated after (2). The subscripts for the configuation of Cu2+ have following meaning: Cu,,, means coordinated to three waters, CUI, to two waters, Cu, to one water, and Cu, not coordinated to any water. Cu,,. represents the several new reversed g-value spectra which are not identical to CU,,. CUII- refers to coordination to two hydroxyls in place of two waters.Cu(NH,),NaA When Cuz+ is ammoniacally exchanged into NaA zeolite none of the copper species, CU~II, Cu,, or Cu,, seen in the freshly prepared samples of other A zeolites is observed. Instead a new Cu, species with gll z 2.31, gl = 2.06 and a large copper hyperfine splitting of 178 x lop4 cm-l is the major copper species, as shown in fig. 6. While no explicit splitting due to interaction with nitrogen nuclei is seen, all the hyperfine lines are considerably broader than those of CUIII, CUII or Cu,, which suggests some unresolved superhyperfine coupling. Also, the low value of gll is indicative of ammonia or an ammonium ion being in the first coordination sphere of C U ~ + . ~ ~ Any dehydration treatment of these ammoniacally treated samples was irreversible.In fig. 6 the e.s.r. spectra of freshly prepared Cu(NH,),-NaA and those in partially dehydrated and rehydrated samples are shown. A ' reversed ' g-value spectrum dominates on evacuation at room temperature or higher, but it is different from the spectrum assigned to CUII as well as from the reversed g-value spectra seen in partially dehydratedM . Narayana and L. Kevan 22 1 Fig. 4. E.s.r. spectra at 293 and 77 K of Cu-NH,A: (a) 29.3 K spectrum of fresh sample and (h) 77 K spectrum of fresh sample. (The e.s.r. spectra in Cu-NH,A rehydrated at T < 343 K are identical to this.) (c) 77 K spectrum of Cu-NH,A air dried at 383 K for 48 h. If this sample is exposed to air or water vapour (h) is restored completely. Fig. 5. E.s.r. spectra (a) at 293 and 77 K of fresh Cu-CH,NH,A and (h) at 77 K of Cu-CH,NH,A dried in air at 383 K for 48 h.The spectra seen in (a) are completely restored if sample (b) is exposed to water vapour for 3 h at room temperature.222 Coordination Geometry of Cu2+ in MSA Zeolites s,c c urn 1 Fig. 6. E.s.r. spectra of Cu(NH,), in NaA at 77 K: (a) fresh sample, (b) after the sample has been evacuated at 293 K for 6 h, (c) after drying the fresh sample at 383 K for 24 h, ( d ) when samples (b) or (c) are exposed to air or water vapour for 3 h at room temperature. Note the formation of Cu,,, in (a) while almost no Cu,,, is observable in (a). Cu-NH,A and Cu-CH,NH,A. On rehydration another reversed g-value spectrum is formed which looks like the Cu,, species, although its A , , value is a little larger.In fact this reversed g-value species is similar to that observed in C U - C ~ X ~ ~ ~ and C U - C ~ X ~ ' ~ in which hydroxyls have replaced water as the two apical ligands2* The other species in the rehydrated sample is CuIII, which was not seen in the fresh samples of Cu(NH,),-NaA. Fig. 7 shows three-pulse e.s.e.m. spectra of fresh and rehydrated ( T = 383 K) Cu-KA samples prepared in D,O which clearly reflect the differences observed in their e.s.r spectra. In fig. 8 the three-pulse e.s.e.m. spectrum of Cu-RbA exchanged with D,O is shown. This is almost identical to the spectra observed in fresh samples of Cu-KA17 or Cu-TlA exchanged in D,O. The analysis of the e.s.e.m. data indicates that Cu2+ interacts with only two deuterons at 0.26 nm with a small isotropic coupling of 0.2 MHz.In contrast, analysis of the e.s.e.m. spectra in the rehydrated ( T = 383 K) samples of Cu-KA or Cu-TlA in fig. 9 indicates that Cu2+ interacts with six deuterons at 0.28 nm with an isotropic coupling of 0.2 MHz, which is identical to the results in either fresh or rehydrated samples of Cu-NaA. Thus both the e.s.r. and e.s.e.m. spectra clearly show some irreversible rearrangement of Cu2+ surroundings in KA and TlA on dehydration at T > 353 K.M. Narayana and L. Kevan 223 Fig. 7. Comparison of three-pulse e.s.e.m. spectra of fresh Cu-KA (---) and Cu-KA rehydrated at room temperature after dehydration at 383 K (-). Note the difference in the depths of deuterium modulation of the two samples; this indicates an increase in the number of interacting deuterons in the rehydrated sample. The spectra were recorded at H, = 3080 G with a first interpulse time z = 0.25 ps.I I I I I I I I I 1 I 2 3 4 5 TIP Fig. 8. (---) Experimental and (-) calculated three-pulse e.s.e.m. spectra of fresh Cu-RbA. The spectrum was recorded at H, = 3040 G with a first interpulse time, 7 = 0.26 ps. The decay function used was G(T) = exp (2.5 -0.3 T+ 0.026 T2). n = 2, Y = 0.26 nm and a = 0.15 MHz. 8 F A R 1224 Coordination Geometry of Cu2+ in M+A Zeolites I I I I I I I I I 1 I 2 3 4 5 TIPS Fig. 9. (---) Experimental and (-) calculated three-pulse e.s.e.m. spectra of Cu-KA rehydrated at room temperature after dehydration at 383 K. This spectrum was recorded at H, = 3 180 G with a first interpulse time, z = 0.25 ps.The decay function used was G(T) = exp (2.7 - 0.11 T). n = 6, Y = 0.28 nm and a = 0.2 MHz. Fig. 10 shows frequency-domain spectra obtained by fast Fourier transformation (f.f.t.) of the time-domain e.s.e.m. spectra of fresh and rehydrated (383 K) Cu-KA. Both these spectra have one strong peak near 2 MHz and another weak peak at ca. 4 MHz. The free-precession frequency of deuterons in a magnetic field of ca. 3180 G is ca. 2.1 MHz, which accounts for the main peak in the f.f.t. spectra. The origin of the weak peak at 4 MHz is unclear at this time. All the simulations done with such large hyperfine couplings yield very poor fits to the experimental time-domain spectra. Samples of Cu(ND,),-NaA/D,O were prepared with ND, and/or with D,O. We observed identical deuterium modulation for both preparations.The three-pulse e.s.e.m. spectrum and its f.f.t. of these samples are shown in fig. 11. Good fits to the time-domain spectra could not be obtained with one or two shells of equivalent interacting deuterons. The f.f.t. clearly shows that Cu2+ is interacting with at least two types of deuterons, with one type having an isotropic coupling of ca. 0.5 MHz or greater.13 This failure to obtain a good fit to the time-domain data with one or two shells of equivalent deuterons suggests that either the point-dipole approximation has failed owing to large delocalization of the unpaired spin into the ligands or that all the ligands of Cu2+ are completely inequivalent. In fig. 12 two-pulse e.s.e.m. spectra of freshly prepared Cu-NaA/H,O and fully dehydrated Cu-NaA are compared.The considerable damping of the modulation in the dehydrated sample is probably due to stronger quadrupolar interactions caused by stronger electric field gradients at the sites of nuclei. Part of the reason for such an increase in the electric field gradient could be the severe distortion of the six-membered ring caused by Cu2+ movement off the trigonal axis as a result of Jahn-Teller dist~rtion.,~ Fig. 13 compares the f.f.t. spectra obtained from the two-pulse e.s.e.m. spectra ofCu-NaA/H,O and Cu-RbA/H,O. Both the spectra have three well defined peaks above the noise level, at ca. 3.5, 7 and 13 MHz. The first two peaks correspond well toM . Narayana and L. Kevun 225 5 10 15 20 frequency/ M Hz Fig. 10.Comparison of the frequency-domain spectra obtained by fast Fourier transformation of the time-domain three-pulse e.s.e.m. spectra of (a) fresh Cu-KA and (b) Cu-KA rehydrated at room temperature after dehydration at 383 K. The spectra were recorded at H,, = 3180 G with a first interpulse time, z = 0.29 pus. 10 8 2 frequency/MHz I I I I I 0 I 2 3 4 5 TIPS Fig. 11. (a) Three-pulse e.s.e.m. spectra of Cu(NH,),-NaA prepared in D,O and (b) its f.f.t. The spectrum was recorded at H,, = 3150 G with a first interpulse time, z = 0.25 ps. Note the clean resolution of two peaks, which is indicative of two types of deuterons, one with 0.5 MHz isotropic hyperfine coupling and the other with no isotropic coupling. 8-2226 Coordination Geometry of Cu2+ in M+A Zeolites 8 1, Fig.12. Comparison of two-pulse e.s.e.m. spectra of (a) fresh Cu-NaA and (b) fully dehydrated Cu-NaA showing 27Al modulation. Note the drastic decrease in the 27Al modulation amplitudes, presumably due to increase in the nuclear quadrupole interactions. the expected positions of the free-precession 27Al frequency and its second harmonic. The third corresponds well to the expected position of protons. The proton peak is considerably stronger for the Cu-NaA sample. This difference in the proton peak intensities in Cu-NaA as against Cu-RbA is authentic, as verified by recording e.s.e.m. two-pulse spectra at a variety of magnetic-field positions and using several sets of new samples. Such a difference is also consistent with the picture obtained by analysis of deuterium modulations in the samples of Cu-NaA/D,O and Cu-RbA/D,O, the former showing an interaction of Cu2+ with six deuterons while the latter shows only two.Analysis of the aluminium modulations such as shown in fig. 12(a) in the zero quadrupolar interaction approximation indicates a Cu2+ interaction with three aluminium nuclei at a distance of 0.36-0.38 nm for CUIII as well as for Cu,. However, these distances are probably shorter owing to modulation damping by quadrupolar 31 E.s.e.m. spectra of the dehydrated samples could not be analysed owing to a lack of sufficient modulation depths. Discussion We have already established27c via analyses of e.s.e.m spectra that in fresh and rehydrated samples of Cu-NaA, Cu2+ is coordinated to three water molecules and is located at site S2* in the a-cages; this Cu2+ species is denoted as CUIII.The cation site locations are shown in fig. 14. On partial dehydration part of the intensity of this complex is lost and a new spectrum with reversed g values (g,, < gl) appears which is denoted CUII. It is wellM . Narayana and L. Kevan 1-3.19 2 1-3.38 17.05 113.60 5 10 15 20 0 frequency/MHz 227 Fig. 13. Comparison of f.f.t. spectra of two-pulse e.s.e.m. spectra in (a) fresh Cu-RbA and (b) fresh Cu-NaA. The spectrum of Cu-RbA was recorded at Ho = 3030 G and that of Cu-NaA at H , = 3165 G. Note the decrease in the intensity of the proton peak. Fig. 14. Site nomenclature in A zeolite. Site S2* projects into the a-cage above a six-membered ring window while S2' projects into the Q-cage.S2 is in the plane of the 0, oxygens of the six ring. S5 is in the plane of the eight-membered ring window of the a-cage, while S3 is an ill-defined site in front of the four-membered rings in the a-cage. known8tZ6 that for Cu2+ to exhibit such reversed g values the )3z2-r2) state should be the ground state for the unpaired spin which occurs in the following geometries: (a) compressed tetrahedral, (b) compressed tetragonally or rhombically distorted octahedral, (c) cis distorted octahedral, ( d ) compressed square pyramidal and (e) trigonal bipyramidal. E.s.e.m. analysis of the Cu2+ species in partially dehydrated Cu-NaA indicates that228 Coordination Geometry of Cu2+ in M+A Zeolites Cu2+ interacts with four deuterons (consistent with two water molecules) thus indicating the trigonal-bipyramidal configuration to be most probable.It should be emphasized that on exposure of water vapour to such a partially dchydrated sample Cu,, is not converted into CUIII. This can be achieved only by complete dehydration at higher temperatures followed by rehydration. The specific reasons for the stability of CUII are not known at this stage. On complete dehydration deuterium modulation can no longer be used as a tool to monitor the location of Cu2+, and the aluminium modulations are also considerably damped, as shown in fig. 12, presumably through increased nuclear quadrupolar interactions. An indirect conclusion from such a damping is that Cu2+ causes distortions in the six-membered ring on dehydration, partly owing to the Jahn-Teller instability, which forces Cu2+ off the trigonal axis and removes the ground-state degeneracy.Analysis of the aluminium modulation in the zero quadrupole interaction approximation for the fresh or rehydrated Cu-NaA samples shows that Cu2+ interacts with three aluminium nuclei at 0.38 nm. Such a result places Cu2+ cu. 0.2 nm above the plane of the 0, oxygens in the six-membered ring windows. However, the actual distance is probably shorter, since quadrupole interactions for 27Al cannot be completely While expressions incorporating quadrupole interaction using a first-order perturbation approach are available, we have shown3’ that such an approach incorporates approximations in the primary dipolar interaction which appear to largely defeat the purpose of the apparently more rigorous analysis.In Cu-KA and Cu-TlA fresh samples the major species is conspicuously different from the Cu,,, species found in Cu-NaA. By analysis of deuterium modulations this species was shown to be interacting with one water and can be denoted Cu,. Another important difference between Cu-NaA and Cu-KA or Cu-T1A is that the reversible formation of Cu, in the dehydration-rehydration cycle is dependent on the dehydration temperature. In Cu-NaA, CUII, is reformed on rehydration following dehydration at any temperature between 373 and 773 K. In Cu-KA and Cu-TlA, Cu, is irretrievably lost if the dehydration is carried out at T > 353 K. Instead, rehydration in both these zeolites results in the formation of CUIII as the major species. In both Cu-KA and Cu-TlA room-temperature evacuation is sufficient to completely destroy Cu,.A strong ‘reversed’ g-value spectrum identical to that in partially dehydrated Cu-NaA; hence Cu,, appears. However, unlike the case in Cu-NaA, on exposure to water vapour CUII disappears in these partially dehydrated Cu-KA and Cu-T1A samples and Cu, is restored as the major species. It is clearly seen in fig. 7 that Cu2+ interacts with more deuterons in a rehydrated ( T > 353 K) than in a fresh sample; the e.s.e.m. results for the rehydrated sample in fig. 9 indicate six interacting deuterons, which translate into three water molecules as expected for CuIr1. Note that such a drastic change takes place only for samples rehydrated after dehydration above 353 K. Since very little Cu,,, is seen in fresh samples of Cu-KA, it is clearly not the lack of water molecules that prevents CUI~I from being the dominant species in the fresh samples or in samples dehydrated at temperatures below 353 K and then rehydrated at room temperature.When bulkier cations are present in the a-cage, such as K+, NH;, Cs+, Rb+ and T1+, they crowd the a-cage considerably compared to the case of NaA. For example, in hydrated KA the eight K+ associated with the six-membered windows project ca. 0.15 nm into the a-cage above the plane of the 0, oxygens, while the corresponding displacement of Na+ from such a plane in NaA is only ca. 0.05 nm. Thus in freshly prepared or rehydrated samples of NaA, a Cu2+ at an S2* site does not experience much electrostatic repulsion from the two adjacent Na+ which are at S2 sites.However, in Cu-KA Cu has to compete with K+ for the S2* positions and will experience more electrostatic repulsion from the two adjacent K+, which are at S2* sites. This repulsion causes Cu2+ in S2* sites to be coordinated to only one water molecule in fresh CU-KA.~~ This kind of repulsion in the a-cages becomes more severe for Cu2+ in NH,A,,, CsA, RbA and CH,NH,A. Consequently in all these zeolites Cu2+M . Narayana and L. Kecan 229 is forced to occupy S2* positions, thus explaining the occurrence of Cur as the dominant or only species. On dehydration at high temperatures it has been established that some of the K+ ions move into the P-cage.,,9 35 No crystal-structure data are available for rehydrated KA, but it is possible that these K+ ions cannot move back into the a-cages.Thus if one or more of the K+ ions at S2* sites were to be irreversibly pushed into the P-cages in the dehydration-rehydration cycles the electrostatic repulsion for Cu2+ in the a-cages is reduced, possibly promoting the formation of CUIII at S2* sites. That Cu, reversibly returns as the dominant species when the dehydration-rehydration cycle is carried out at milder temperatures could indicate an energy barrier for K+ to go through the six-ring windows, since the ionic radius of K+ and the opening radius or a six-ring window are comparable (K+ ionic radius = 0.133 nm and the average six-ring opening radius is 0.1 1-0.13 nm). A corresponding analogy cannot be extended to explain the irreversible decrease of Cu, in T1A zeolite.Riley et aZ.36 studied the structure of Tl,,Na,A. Both in the hydrated and dehydrated forms they placed 7 TI+ ions projecting ca. 0.15 nm into the a-cages above the 0, plane of the six rings and one Tl+ recessed into the a-cage ca. 0.17 nm below the 0, plane. The position of Na+ was not given. Thus no T1+ ions were found to move into the P-cages on dehydration. Thus a better explanation may be that CU~II is formed in S2' on r e h y d r a t i ~ n . ~ ~ We have already shown that in Cu-CsA Cu, is the dominant species in fresh samples.16 In Cu-CsA the e.s.e.m. of 135Cs modulations have been reinterpreted from ref. (16) to indicate that Cur is most probably in the a-~age.,~ Since the modulation frequencies of deuterium and caesium are of the same order, it was not possible to obtain unambiguously the number of water molecules from an e.s.e.m.analysis of deuterium interactions in this zeolite. This difficulty was not a problem in Cu-RbA zeolite. The magnetic isotopes *"b and s7Rb have strong quadrupole moments and their Larmor frequencies are considerably different from that of deuterium. No modulations assignable to these Rb nuclei were observed in hydrated or dehydrated samples. Fig. 8 is typical of the e.s.e.m. results we have obtained for deuterated Cu-RbA, which unambiguously indictates that Cu2+ in this zeolite interacts only with two deuterons both in fresh as well as in rehydrated samples, irrespective of the temperature of dehydration. In Cu-NH,A and Cu-CH,NH,A Cu, is the dominant species, but we did not carry out any deuterium e.s.e.m.studies because of the ease with which the NH, protons exchange with the deuterons; one cannot easily distinguish the interactions of water deuterons from NH, deuterons. This was also the situation in e.s.e.m. studies of Cu(NH,),-NaA. The e.s.r. spectrum in partially dehydrated samples of Cu-NH,A clearly indicates that some NH,+ ions are quite close to Cu2+, as reflected by the nitrogen superhyperfine splitting. However, the two-pulse aluminium modulations recorded in these two samples were identical to these in fresh Cu-RbA and in fresh Cu-KA, thus indicating the same S2' location for Cu2+ in all these zeolites. It is not clear why the reversed ' g-value species observed in partially dehydrated Cu-NH,A and Cu-CH,NH,A are different from one another and different from the Cu,, observed in other A zeolites.One possibility could be that in the trigonal-bipyramidal configuration the ligand in the a-cage in these two zeolites is not a water molecule but the respective cation, NH; or CH,NHi. In Cu(NH,),-NaA we could not arrive at a satisfactory model for the Cu2+ deuterium interactions of Cu,. As shown in fig. 9 (b), the frequency-domain spectra clearly indicate that there are at least two types of deuterons in the Cu2+ coordination sphere. The e.s.r. g-values indicate that this Cu2+ is probably in square-planar or distorted-octahedral coordination with some NH, groups in the first coordination sphere, since the gll value of Cu2+ decreases when Cu-N bonds are present compared with g values of Cu2+ with only oxygens as ligands.26 On increasing the copper concentration in the exchange230 Coordination Geometry of Cu2+ in M+A Zeolites solution we were not able to see any clean resolution of nitrogen superhyperfine splitting, but the spectral features were considerably broader than in other A zeolites.On partial dehydration this copper species rapidly loses intensity and the sample changes from pale blue to white, indicating loss of NH, ligands. The irreversible nature of this loss as observed by formation of CUIII on rehydration indicates that NH, is involved in the coordination sphere of this Cu, species. An interesting feature in the rehydrated Cu(NH,),-NaA samples was the appearance of a well resolved ' reversed' g-value spectrum which is not similar to other reversedg-value species seen in any of the A zeolites but is almost identical to that seen in Cu-CaX2ia and Cu-CdX.2ib In both these X zeolites we were able to identify Cu2+ to be in trigonal- bipyramidal coordination in six-ring windows coordinating axially to two hydroxyls, one each in the a- and /&cages, respectively. Lee et from their crystallographic studies came to the conclusion that in partially dehydrated Cu(NH,),-NaA the major copper species is trigonal-bipyramidally coordinated in the six-ring windows with two hydroxyls as the axial ligands. However, they were not able to see CUII,, which e.s.r.and e.s.e.m. show to be present in these samples on rehydration. We have not attempted an analysis of deuterium e.s.e.m.spectra in this sample because we cannot distinguish deuterium between ND, and D,O ligands in Cu(NH,),-NaA. A few words are in order regarding fig. 13, where the fast Fourier transform of the time-domain data for the aluminium modulations is shown. The nuclear Zeeman frequency for 27Al in a field of ca. 3150 G is 3.88 MHz, and the main peak in the f.f.t. spectra is close to this frequency. Thus the nuclear quadrupole interaction in these zeolites is not strong enough to shift significantly the ENDOR energy levels and cause new frequencies, as has been observed in the chlorophyll cation.25 However, the quadrupole interaction is not negligible24 compared with the dipolar interaction because there is a sharp decrease in the intensity of the second harmonic of the 27A1 peak in fig.13. Conclusions E.s.r. and e.s.e,m. studies have been carried out for fresh, various vacuum treated, rehydrated and deuterated samples of Cu-NaA, Cu-KA, Cu-TlA, Cu-CsA, Cu-RbA, Cu-NH,A, Cu-CH,NH,A and Cu(NH,),-NaA zeolites. Three major copper species are identified and characterized from their deuterium and aluminium modulations. In fresh and rehydrated Cu-NaA only one type of Cu2+ exists, CUIII, coordinating to three water molecules at site S2* in the a-cage. On partial dehydration two of the water molecules in the a-cage are lost, Cu2+ moves into the plane of the 0, oxygen atoms in the six ring and interacts with another water in the P-cage. This complex is not reconverted into Cu,,, on exposure to water vapour. This CUII is coordinated to two water molecules and can be distinguished from the trigonal bipyramidal complex coordinated to two hydroxyls characterized by e.s.e.m.analysis in Cu-CaX and Cu-CciX.2i Unlike in NaA, the major copper species in all the other A zeolites studied has a very high gll value. This is assigned to a tetrahedrally coordinated Cu2+ in the a-cages at site S2* and is substantiated by e.s.e.m. analysis of Cu-Cs interactions in CU-CSA.,~ Deuterium modulation analysis for this species in Cu-KA, Cu-TlA and Cu-RbA clearly shows that Cu2+ interacts with only two deuterons, corresponding to one water molecule. On partial dehydration Cu2+ moves from S2* to S2 and interacts with another water molecule in the a-cage forming trigonal-bipyramidal coordination identical to that in Cu-NaA. However, unlike in NaA, on exposure to water vapour CUII is readily converted into Cu, in Cu-KA, Cu-TlA, Cu-RbA and Cu-CsA.This preferential formation of Cu, in these four zeolites at S2* in the a-cage is attributed to the electrostatic repulsion by the bulky cations K+, Tl+, Cs+ and Rb+ in the a-cages. In Cu-KA and Cu-TlA Cu, is irreversibly lost when dehydration is carried out atM . Narayana and L. Kevan 23 1 temperatures > 353 K. Rehydration after such a treatment results in the formation of Cu,,, in both these zeolites, as evidenced by changes in the e.s.r. and e.s.e.m. spectra. In Cu-KA this irreversible change could be due to Cu2+ moving into the P-cages on dehydration and then being rehydrated there. In Cu-CsA and Cu-RbA the loss of Cu, on dehydration is reversible on rehydration, irrespective of the rehydration temperature.In Cu-NH,A and Cu-CH,NH,A the major species is also Cu, in fresh samples. The reversed g-value spectra obtained on partial dehydration in these zeolites are different from that of Cur, observed in the alkali-metal ion zeolites. In Cu-NH,A a clear indication of Cu-N interaction is seen from hyperfine structure in the e.s.r. spectra on partial dehydration. Thus the difference in the reversed g-value spectra in Cu-NH,A and Cu-CH,NH,A as against the other A zeolites could be due to replacement of the a-cage water in the trigonal-bipyramidal configuration by a NH,+ or CH,NH: ion. Cu, is restored as the major species for dehydration below 353 K in both these zeolites. Dehydration above 353 K results both in drastic reduction of the Cu2+ signal intensities and in loss of crystal structure.In NaA zeolite in which copper is ammoniacally exchanged, a new copper complex not seen in the other A zeolites is observed. While we could not quantitatively analyse the deuterium modulation in samples hydrated with D,O, this complex is most likely square planar or a distorted octahedral complex in the a-cages. On mild dehydration and rehydration this complex is lost and Cu,,, and a new reversed g value Cu2+ appear. The latter is identical to the trigonal bipyramidal complex with two apical hydroxyl ligands seen in Cu-CaX and Cu-CdX zeolites. This work was supported by the U.S. National Science Foundation, The Robert A. Welch Foundation and the Energy Laboratory of the University of Houston.References 1 I. Mochita, S. Hoyata, A. Kato and T. Seiyama, J . Catal., 1969, 15, 314. 2 K. Tsutsumi, S. Fuji and H. Takahashi, J . Catal., 1972, 24, 146, 3 C. Dimitrov and H. F. Leach, J. Catal. 1969, 14, 336. 4 I. R. Leith and H. F. Leach, Proc. R . Soc. London, Ser. A , 1972, 330, 247. 5 E. F. Vansant and J. H. Lunsford, J. Phys. Chem., 1972, 76, 2860. 6 R. A. Schoonheydt, P. Peigneur and J. B. Uytterhoeven, J . Chem. Soc., Furaduy Trans. 1, 1978, 74, 7 D. R. Flentge, J. H. Lunsford, P. A. Jacobs and J. B. Uytterhoeven, J . Phys. Chem., 1975, 79, 356. 8 R. Herman, Znorg. Chem., 1979, 18, 995. 9 W. B. Mims, Phys. Rev. B, 1972, 5, 2609. 2550. 10 W. B. Mims, J. Peisach and J. L. Davis, J . Chem. Phys., 1977, 66, 5536. 11 L. Kevan in Time Domain Electron Spin Resonance, L. Kevan and R. N. Schwartz (Wiley-Interscience, New York, 1979) chap. 8. 12 T. Ichikawa, L. Kevan, M. K. Bowman, S. A. Dikanov and Yu. D. Tsvetkov, J . Chem. Phys., 1979, 71, 1167. 13 P. A. Narayana and L. Kevan, J . Magn. Reson., 1982,46, 84. 14 T. Ichikawa, and L. Kevan, J. Am. Chem. Soc., 1981, 103, 5355. 15 B. L. Dickson and L. V. C. Rees, J . Chem. Soc., Furaday Trans. I , 1974. 74. 2038. 16 M. Narayana and L. Kevan, J. Chem. Phys., 1981, 75, 3269. 17 M. Narayana and L. Kevan, J . Phys. C., 1983, 16, 361. 18 N. N. Pafomov, V. A. Silchenko and Yu. A. Bratashevskii, Teor. Eksp. Khim., 1978. 14, 269. 19 R. M. Barrer and W. M. Meier, Trans. Faraday Soc., 1958, 54, 1074. 20 J. J. Pluth and J. V. Smith, J. Am. Chem. Soc., 1983, 105, 2621. 21 T. Ichikawa, L. Kevan and P. A. Narayana, J. Phys. Chem., 1979. 83, 3378. 22 (a) P. A. Narayana and L. Kevan, Photochem. Photobiol., 1983, 37, 105; (b) P. A. Narayana and 23 S . A. Dikanov, A. A. Shubin and V. N. Parmon, J . Magn. Reson.. 1981,42,4748. 24 A. Samoson and E. Lippmaa, Chem. Phys. Lett.. 1983, 100, 205. 25 S. A. Dikanov, Yu. D. Tsvetkov, M. K. Bowman and A. V. Astashkin, Chcm. Phys. Lett., 1982, 90, 26 B. J. Hathaway and D. E. Billing, Coord. Chem. Rer., 1979, 5, 143. L. Kevan, Magn. Reson. Rev., 1983, 1, 234. 149.232 Coordination Geometry of Cu2+ in M+A Zeolites 27 (a) M. Narayana and L. Kevan, J. Chem. Phys., 1982, 78, 3573; ( h ) M. Narayana and L. Kevan, unpublished results; ( c ) L. Kevan and M. Narayana, Am. Chem. Soc. Sq’mp. Ser., 1983, 218, 283. 28 H. S. Lee, W. V. Cruz and K. Seff, J. Phys. Chem., 1982, 86, 3562. 29 H. A. Jahn and E. Teller, Proc. R. SOC. London, Ser. A , 1937, 161, 220. 30 A. A. Shubin and S. A. Dikanov, J. Mugn. Reson. 1983, 52, 1. 31 M. Romanelli, M. Narayana and L. Kevan, J. Chem. Phys., 1984, 80,4044. 32 M. Anderson and L. Kevan, J. Am. Chem. SOC., submitted for publication. 33 L. B. McKusker and K. Seff, J . Am. Chem. Soc., 1981, 103, 3441. 34 P. C . W. Leung, K. B. Kunz, K. Seff and I. E. Maxwell, J. Phys. Chem., 1975, 79, 2157. 35 J. J. Pluth and J. V. Smith, J . Phys. Chem., 1979, 83, 741. 36 P. E. Rdey, K. Seff and D. P. Shoemaker, J. Phys. Chem., 1972, 76, 2593. Paper 5/615; Receiwd 12th April, 1985