Chlorine N Q R on Derivatives of Chloral

Transcription

1 Chlorine N Q R on Derivatives Chloral Masao Hashimoto, Norbert Weiden, and Alarich Weiss nstitut für Physikalische Chemie, Physikalische Chemie, Technische Hochschule Darmstadt Z. Naturforsch. 35 a, (980); received July 28, 980 The 35C-NQR spectrum several derivatives chloral, Cl3CCHO, was studied in the ranee 77 K ^ T Tm. By use the spin echo double resonance technique a relative assignment the resonances to different ClaC-groups within one crystallized compound was possible. The solid compounds studied are: High temperature phase chloral hydrate, Cl3CCH(OH) 2 ; parachloral, (Cl 3 CCHO) 3, (a- and ß-isomer); the two phases chloral hemihydrate C3CCHO /2 H 2 0 ; chloralide (); chloralhemithiohydrate, C3CCHO /2 H 2 S. The structure the molecules in the solid state and the fade out the NQR resonances are discussed. ntroduction A simple compound earring a trichloromethyl group, CC 3, is trichloroacetaldehyde (chloral), CC 3 CH0. t was prepared f o r the first time and extensively studied by Liebig [ ]. The diemistry chloral was recently reviewed b y Luknitskii [ 2 ]. n course studies chloral hydrate, C C 3 C H ( 0 H ) 2 [3, 4 ] and several trichloromethyl derivatives [ 5, 6 ] the relative intramolecular configuration different CC3 groups within one molecule or within one crystallographically asymmetric unit was interest. t has been shown that Spin Echo Double Resonance, S E D O R [ 7 ], fers an experimental way to contribute to the problem [ 8, 9 ]. Here S E D O R is applied besides wide line Nuclear Quadrupole Resonance ( N Q R ) f o r the study the relative configuration the following compounds: high temperature phase chloral hydrate, C C 3 C H ( 0 H ) 2 ( ), the two solid phases () and ( ) chloral hemihydrate, CC 3 CH0 /2 HoO, the ester chloralide, CCl 3 CHOCOCHOCCl 3 (which shows a solid state phase transformation above r o o m temperature), the two isomers trimeric chloral, (CC 3 CH0) 3, known as a-parachloral and ^-parachloral, and chloral hemithiohydrate, CC 3 CH0 / 2 H 2 S. The application NQR is particular advantage in the study solid trichloromethyl derivatives where in numerous examples solid state transformations have been observed. By means Cl N Q R it is possible to follow up the transformation process Reprint requests to Pr. Dr. A. Weiss, nstitut für Physikalische Chemie, Technische Hochschule Darmstadt, Petersenstraße 20, D-600 Darmstadt. and the onset hindered rotation the CC3 groups which results in a " f a d e - o u t " the N Q R lines. Experimental Preparation the Compounds The high temperature phase chloral hydrate, C C 3 C H ( O H ) 2 ( ), and trichloroethylidene trichlorolactic ester (chloralide), CCl3CHOCOCHOCCl3, were obtained as described previously [ 3, 4, 6 ]. Chloral hemithiohydrate, CCl 3 CHO / 2 H 2 S, was synthesized f r o m chloral hydrate and H 2 S according to Chattaway and Kellett [ 0 ]. a-parachloral and ^-parachloral, ( C C 3 C H 0 ) 3, were prepared f r o m chloral hydrate and sulphuric acid by the method given by Chattaway and Kellett [ ]. Chloral hemihydrate, CCl 3 CHO / 2 HoO, was synthesized f r o m CCl 3 CHO and C C 3 C H ( 0 H ) 2 in the following w a y : One mole C C 3 C H ( 0 H ) 2 (solid) was mixed with one mole CCl 3 CHO ( l i q u i d ). The mixture was warmed up to ca K to dissolve the solid part thoroughly. T o the liquid thereafter a trace sulphuric acid was added and by annealing at ca. 295 K the liquid solidified completely within one hour. T h e c o m p o u n d can be recrystallized f r o m HCC3. The results chemical analysis are (calc/ found): C(5.360%/5.0%); H(.289%/.5%). The existence chloral hemihydrate has been proposed by van Rossen [ 2 ] and Shill [ 3 ]. mmediately after solidification Cl 3 CCHO /2 H 2 0 showed at 77 K a very weak, six line 35 C-NQR spectrum. Within a few days the signals became relatively strong. The same spectrum was observed ,/ 80 / $ 0.00/0. - Please order a reprint rather than making your own copy.

2 046 for M. Hashimoto et al. Chlorine NQR 0 Derivatives Chloral a specimen recrystallized from HCC 3. The formulated as polycrystalline substance obtained in that way was sealed in a glass ampoule and kept at room temperature. Within several weeks a different set 35C- N Q R signals appeared for this sample. The new signals became more intense day by day at the expense months the six a 35 C-NQR was line spectrum new spectrum after three twelve lines fairly g o o d signal to noise ratio observed. These designated and consisting two as phase crystalline (six lines) phases are and phase (twelve lines), respectively. The R spectrum the compound (KBr-disk) shows the presence ether bond, suggesting the following structure: H H an CloC-C-O-C-CCL OH NQR OH Spectroscopy The search f o r the 35 C-NQR lines was done with a superregenerative Zeeman modulated spectrometer ( D e c c a ). For the SEDOR experiments a two channel pulse spectrometer [ 9 ] was employed. The temperature measurements (thermocouple at the sample) are accurate to i K, the error in the frequency measurements is + 3 khz and only due to the line width. Theory and experiment S E D O R have been developed by Emshwiller, Hahn, and Kaplan [ 7 ]. For a SEDOR experiment the sample must contain at least two crystallographically inequivalent nuclei with spin /, which we call Aand B-nuclei. During the experiment the echo amplitude the A-nuclei is monitored by a 90 -T-80 pulse sequence. At the time r a third pulse with length n is applied to the sample and the frequency this rf-pulse is varied in steps while the whole pulse program is repeated. f the frequency the third pulse matches one the resonance frequencies the B-nuclei, that part the local magnetic field at the site the A-nuclei which originates f r o m the magnetic moments the B-nuclei, changes its sign: ^AB ~ ^AB > ÄAB = / / B / r A B - () This leads to a disturbance the phase memory the A-spins, and consequently the echo amplitude the A-nuclei is lowered. Following Emshwiller et the decrease the A-echo amplitude AE can al. be AE = yi y% TAAB <J (/ + ) r 2 h2 2 gjk k. (2) 7 A J 7B a r e the gyromagnetic ratios the A- and Bspins; T A A is the relaxation time the A-spins due to A-A interaction only, B the spin quantum number the B-system, gj k a geometrical factor and Tjk the distance between the nuclei A a n d B/v.. n our experiments the width the 9 0 pulse for the A-spin system was 5 psec in all cases. The time intervall r in the sequence r was around 000 psec at 77 K and p s e c for measurements close to r o o m temperature. For the length the n 8 0 pulse applied to the B-system psec were choosen in order to avoid broadening the B-spin resonance line through Fourier components the B-pulse. By heteronuclear double resonance it is possible to detect very weak N Q R signals the B-spin system via their influence on the A-signal, and the l / r 6 dependence fers a way to determine relative distances between A and B nuclei. For this purpose heteronuclear [ 7 ], homonuclear [ 9 ] and isotopic double resonance can be used. The last two techniques have been applied to study the relative intramolecular configuration chlorine nuclei in trichloroacetic acid and other compounds [ 9 ] and homonuclear double resonance was applied in this work, too. Results n Fig. 5 the 35 C-NQR frequencies are shown as a function temperature. A normal Bayer type behaviour [ 4 ] is found for all five compounds, and in Table the coefficients the polynominals v{t) =A + B/T + CT + DT2 (3) are given. This function describes the temperature dependence v( 3 5 Cl) in the range 77 K ^ T ^ Tm {Tm = melting point) adequately within the limits experimental error in v and T. Table 2 lists the resonance frequencies >'( 35 C) the compounds measured at 77 K, together with T m and the "fade-out" temperatures T f. The fade-out temperature is defined as that temperature at which the 35 C-NQR signals disappear in the noise (see [ 6 ] ). Some the signals fade out at temperatures far below the respective melting point, while others could be observed till the compound melted. No

3 047 M. Hashimoto et al. Chlorine NQR 0 Derivatives Chloral vi CD a-ctjccholj v!"ci) C3CCHO jhjo) - v u.s 0 v O: v U.s \ 39.5 Nq V 5 - V ' V \ \ v, * N T/K' 300 WO Fig.. Temperature dependence the 35 C-NQR frequencies in a-parachloral, a-(cl3ccho)3. vftp Fig. 2. Temperature dependence the 35 C-NQR frequencies in /?-parachlora, /?-(Cl3CCHO)3. anomaly was found in the v( 33 C) =j(t) curves except for chloralide, which compound shows a reversible phase transformation, chloralide ()^ chloralide () [6]. T/K T/K' Fig. 3. Temperature dependence the 35 C-NQR frequencies in chloral hemihydrate (phase ), C3CCHO /2 HaO(). Discussion All compounds studied here by 33 C-NQR are derivatives chloral, CCl3CHO, and by the chemical treatment the compounds the CC3 group was not affected. Each the compounds shows more than three 35 C-NQR lines at 77 K, an observation which proves that more than one CC3 group must exist within the asymmetric unit the crystallographic unit cell. By the SEDOR technique it was possible to decide which the 35 C-NQR lines (the number n lines at 77 K is 5 < n r^ 2 in the compounds studied) belongs to the same CC3 group. The relative intramolecular assignment found in this way is shown in Table 3, together with the frequency splitting within one CC3 group. a- and ß-parachloral From J H-NMR Novak and Walley [5] concluded, that a- and /?-parachloral in chlororm solution have Cs and C3 symmetry, respectively. The molecular structure a-parachloral is a six membered ring having chair conformation with one

4 048 M. Hashimoto et al. Chlorine NQR 0 Derivatives Chloral Fig. 4. Temperature dependence the 35 C-NQR frequencies in chloral hydrate (phase ), Cl3CCH(OH)2(T). axial and two equatorial CC3 groups. n Fig. 6 the structure the molecule a-parachloral is sketched. Thus a molecule a-parachloral has nine chlorine atoms, while the solid compound showed a five line 3j Cl-NQR spectrum in the temperature range investigated. At 77 K four lines were observed (see Table 2 and Figure ). This is due to an accidental coalescence the two lines )'3,?'4 to an unresolved doublet at The superposition two lines at 77 K is confirmed by the r( 35 Cl) =f{t) measurements (see Fig. ) and by SEDOR experiments at 77 K and K. The 35 C-NQR spectrum (SEDOR) indicates that i'j, r3, v5 (see Table 3 and Fig. ) belong to the two equatorial CC3 groups and that the two equatorial CC3 groups are crystallographically equivalent. The line group b, which consists two lines (v2 and v4) is assignable to the axial CC3 group. Two the three chlorine atoms in the axial group T/K Fig. 5. Temperature dependence the 35 C-NQR frequencies in chloral hemithiohydrate, C3CCHO /2 H2S. are also equivalent. So the molecule must have Cs symmetry in the crystalline state, the same symmetry as in the solution. Since the intensity the 35 C-NQR line v2 is two times weaker than that r4, v2 belongs to the chlorine atom located on the symmetry plane. On the other hand /?-parachloral loses its C3 symmetry by the solid state interactions. t becomes totally asymmetric (Cj) as evidenced by the 35 C-NQR spectrum: a) Nine lines are observed and b) SEDOR reveals three CC3 groups, each which is in itself asymmetric (Table 3). Chloral Hemihydrate Chloral hemihydrate, Cl3CCHO-/2 H20, appears with two crystalline phases. Nothing is known about

5 049 M. Hashimoto et al. Chlorine NQR 0 Derivatives Chloral Table. Coefficients the power series i-( 35 Cl) =f(t) = A + B/T + CT + DT 2. For numbering v see Table 2. Compounds No lines A B K r C- 0 3 K D-05 K2 Temperature range ATK a-parachloral J'l v o V J' J' /?-parachloral J'l J' J' J' J' J' J' J' ' Chloralhemihydrate >' phase J' J' »' J' <' Chloral hydrate phase J'l J' '' J' J' J' J' J' »' no rix J' Chloral hemithiohydrate J'l J' '' J' J' J' J' J' J' J' 'll J' their crystal structure; also the relations between C3CCH0-/2H20() and Cl3CCHO /2 H20 () are not clear yet. From the crystallographic point view the six 35 C-NQR lines chloral hemihydrate () can be accounted for by one the two following models: the unit cell contains () one molecule in the asymmetric unit with two crystallographically inequivalent CC3 groups or (2) two molecules in the asymmetric unit, each having Cs (or C2) symmetry *. Both models are in accordance with SEDOR experiments and no decision between () and (2) is possible on the basis 35 C-NQR. Model () seems to be the more likely one in accordance with an ether bridged molecular structure. n Fig. 7 the * f the molecule maintains the C3 symmetry in the crystalline state, there must be three independent molecules in the unit cell, to account for a nine line NQR spectrum. This is, however, crystallographically difficult.

6 050 M. Hashimoto et al. Chlorine NQR 0 Derivatives Chloral Table 2. The 35 C-NQR frequencies, r( 35 Cl), at 77 K, signal to noise ratios, S/N, fade out temperatures, Tf, and melting points, Tm, some chloral derivatives. Compounds»-/ (S/N) Tt/K Tm/K Compounds r/ (S/N) Tt/K Tm/K a-parachloral /5-para - chloral Chloral hemihydrate phase Chloral hemihydrate phase vi V2 J'3, 4 J'5 V t'2 V3 V4 J'5 J'6 v7 V8 vg vi V2 J'3 V4 J'5 V6 V V2 V3 V4 J'5 J'G J'7 VS J'9 J' [38.23] b [38.904] [39.60] [39.280] [39.327] [39.508] [38.60] [38.879] [38.886] [39.003] [39.036] [39.045] [39.07] [39.53] [39.355] [39.45] (53) (24) (all (70) the lines) (43) (4) 420 (4) (v4,v5,v6) (6) () (0) (the other (5) lines) (2) (8) (5) () (6) (9) (7) (D (20) (4) (4) (6) (3) (4) (4) (2) (5) (5) (7) Chloral hydrate c phase Chloral hemithiohydrate vu (6) [39.752] J' (5) [39.888] 'l (0) 255 [38.59] ' [38.538] (7) (N, ''s. V (4) [38.550] (8) [38.7]» (8) [38.956] '' (8) [38.99] J' (8) [39.4] J' (8) [39.94] J' (8) [39.306] >' (8) [39.465] 'H (6) [39.783] '' (7) [39.86] J'l (3) 345 V (7) (''3, J'6 J' (9) J' (2) " (9) J' (2) J'7, d (5) >' (7) '' (D 'll (3) '' (2) '' (7) V (7) ('', J'4 J' (7) J' (22) J' (7) '' (20) a S/N measured with lock in technique and a time constant r = 0 sec. b The frequencies in the paranthese, [ ], correspond to the deuterate. The data were taken from ref. [3]. The superposition two lines was confirmed by the frequency vs. temperature relationships. This compound has a phase transition at 344 K (Ref. [6]) NQR line groups are shown in form a stick diagram. Twelve 35 C-NQR lines are found for C3CCHO /2 HoO (). A comparison the assignment the frequencies to different CC3 groups in phase and phase leads to the assumption that the line group a in phase corresponds to the groups a and a' in phase and group b to b and b'. This proposal is based on () the large frequency splitting line group a in phase and ; and (2) the fact that both groups b show nearly the same frequencies in phase and phase. Thus on the basis the SEDOR experiments one can suggest that the twelve NQR-lines belong to two crystallographically nonequivalent molecules.

7 05 M. Hashimoto et al. Chlorine NQR 0 Derivatives Chloral Table 3. Line groups as determined by SEDOR and frequency splittings, Av, each line group at T K. Compounds Line Components Av/ Group a-parachloral * a J'l, J'2, J'5 0.6 b J'2, J'4 0.0 /?-parachloral a N, J'7, J' b J'2, J'3, J' c J'4, J'5, J'6 0.6 Chloral hemihydrate a J'l, J'4, VB.2 phase b J'2, J'3, J' Chloral hemihydrate a J'l, J'9, J'2.29 phase a' J'3, J'5, 'll 0.85 b J'2, J'8, J' b' J'4, J'6, J' Chloral hydrate a J'l, J'5, J'2.3 phase a' J'2, J'6, Vll.26 b J'3, V8, J' b' J'4, V, J' Chloralide a J'l, J'4, J'6.3 phase b J'2, J'3, J' Chloral a VL, V$, J'0.82 hemithiohydrate a' J'2, J'7, J'9.4 b J'3, J'6, J'2.05 b' V\, J'5, J'l! 0.92 * Complete assignment by SEDOR has been carried out at T = K. For the other compounds the assignment to the different line groups was done by SEDOR technique at 77 K. Chloral Hydrate () SEDOR divides the twelve 35 C-NQR lines chloralhydrate (phase ) into four line groups, which correspond to four crystallographically inequivalent CC3 groups. Therefore the unit cell this solid must contain four molecules within the asymmetric unit. n Fig. 7 the spectrum divided up into the different groups is shown. t is noted that the splitting pattern the line groups a and a' (and b and b' also) are strikingly similar to each other. Although four nonequivalent molecules in the solid evidently a-parachloral line group a ß-parachloral b ' 3 5 line group a ' '' chloral hemihydrate phase ) >ll< _ b c line group a ' ' 5 chloral hemihydrate (phase ) b 3, line group a > ' 2 chloral hydrate phase )! 3,.3 " b 2, b' i V line group a ' ' chloral hemithiohydrate a' 2 6 " b»» b' ' 'l '» line group a >., 0 chloralide phase H) a' 2 '' b 3 " ti ' l S... ' 5! 4 line group a ' ' 5 b ' ' V/ Fig. 7. Stick diagram the 35 C-NQR frequencies in several derivates chloral, CC3CHO. exist as mentioned above, there seems to be a pseudosymmetry which makes the molecules pairwise almost equivalent. Chloralhemithiohydrate Chloralhemithiohydrate is assumed to have a thioether structure [0] : H H C3C-C-S-C-CC3 O : C, : 0 Fig. 6. Structure the molecule a-parachloral, a (C3CCH0)3. OH OH Based on SEDOR, the twelve 33 C-NQR lines chloralhemithiohydrate are divided into four line groups shown in Table 3 and Figure 7. The splitting pattern line group a bears a striking resemblance to that line group a'. Similarly the splitting patterns line group b and b' are quite alike. These facts suggest that line-group a and b belong to one crystallographically independent molecule and the

8 052 M. Hashimoto et al. Chlorine NQR 0 Derivatives Chloral line group a' and b ' to another molecule. t is in- by deuteration experiments that the frequency split- teresting to compare the N Q R spectra chloral- ting Av =.43 at 77 K is not due to hydrogen hemithiohydrate and chloralhemihydrate. The two bonds C... H [3, 7 ]. molecules have a c o m m o n structure, the former The origin responsible f o r the large Av's, in the being a thioether and the latter an ether. From this compounds considered here is most probably structural similarity one may expect a resemblance tramolecular. n this connection it is interesting to between their N Q R spectra. n fact the two com- note that the largest splitting pounds show quite similar N Q R spectra (see Fig- C l 3 C C H O - l / 2 H 2 S, a compound in which one the ure neighbours the CC3 group is the bulky sulfur 7). t is obvious that the line groups a and a' the hemithiohydrate correspond to the line groups a chloralhemihydrate () hemihydrate (). line characterized by These four and chloral- groups their large frequency are observed was infor atom. Fade Out splittings. t is well known that the Cl-NQR signals CC3 Since this feature is c o m m o n to the two compounds, groups ten fade out at temperatures far below the the large frequency splitting is attributable to an in- melting point. This phenomenon has been ascribed tramolecular origin. Moreover, it will possibly be to the occurance reorientational motions the steric nature, because the hemithiohydrate shows group around its three fold axis [ 8 ]. n general Tf tends to rise with increasing magnitude the much larger splittings. potential barrier which hinders the reorientational Frequency motion [ ]. For the compounds studied here, Splitting The magnitude the frequency splittings aand ^-parachloral 0.88 at 77 K ) ( ( A v ) a = 0.6, = is the order the crystal field splitting observed for bonded to carbon {Av)ß [6]. 35 C-NQR on chlorine n the line group a chloralhemihydrate ( ) the frequency vx is strongly depressed compared to the other lines, whereby a large frequency splitting Av =.2 results for this group. For sine group a in chloralhemihydrate() Av =.29 has been obtained (Table 3 ). Line groups a and a' chloralhydrate () exhibit splittings similar magnitude (Av =.3 and.26, respectively). A m o n g the compounds studied here, shows largest splitting the chloralhemithiohydrate almost 2 (group a ). The magnitude these splittings is too large to be explained simply by crystal field effects, because f o r CC3 groups they do not exceed [ 5 ]. t is tempting to assume C... H type hydrogen most the Tf are higher than K, except f o r the line group a C 3 C C H ( O H ) 2 ( ). rf,max = K was found f o r chloralide ( ), line group a. These values are much higher than those found f o r CC3 groups in trichloroacetates [5]. n both cases a CC3 group has the configuration COO as neighbours, but the electronic configurations the carbon attached to C3C differ f r o m each other. t is sp 3 (chloral derivatives) and sp 2 (trichloroacetates), respectively. Thus it can be seen that the reorientational motion CC3 groups attached to C ( s p 3 ) is much more hindered by a potential barrier higher than that C l 3 C - C ( s p 2 ). One the three CC3 groups /?-parachloral differs in Tf considerably f r o m the two other groups. This difference is induced by intermolecular forces because in solution the three CC3 groups are equivalent. A detailed discussion on this point, however, rests on the knowledge the crystal structure. bond formation in crystallized chloralhydrate () and A cknowledgemen chloralhemihydrate ( ). in Table. For chloralhydrate() it has been shown W e are grateful to the Deutsche Forschungsgemeinschaft and the Fonds der Chemischen ndustrie for support this work. M. Hashimoto is obliged to the Alexander v. Humboldt-Stiftung for a fellowship. [] J. Liebig, Ann. Pharmazia,, 82 (932). [2] F.. Luknitskii, Chem. Rev. 75, 259 (975). [3] D. Biedenkapp and A.Weiss, Z. Naturforsch. 22a, 24 (967). [4] M. Hashimoto and A. Weiss, J. Mol. Struct. 58, 243 (980). [5] M. Hashimoto, M. Watanabe, and H. Takada, J. Magn. Reson. 34, 553 (979). However, this assumption should be excluded, at least f o r C l 3 C C H ( O H ) 2 ( ) a n d Cl 3 CCHO / 2 H 2 0 ( ) because no significant change in v ( 3 5 C l ) was originated by deuteration, as shown t

9 053 M. Hashimoto et al. Chlorine NQR 0 Derivatives Chloral [6] M. Hashimoto, H. Paulus, and A. Weiss, Ber. Bunsenges., Physik. Chem., 84, 883 (980). [7] M. Emshwiller, E. L. Hahn, and D. Kaplan, Phys. Rev. 8, 44 (960). [8] N. Weiden and A. Weiss, J. Magn. Reson. 30, 403 (978). [9] N. Weiden and A. Weiss, Faraday Disc. Chem. Soc. Faraday Sympos. 3, 93 (979). [0] F. D. Chattaway and E. G. Kellett, J. Chem. Soc. 929, [] F. D. Chattaway and E. G. Kellett, J. Chem. Soc [2] C. van Rossen, Z. Phys. Chem. 62, 68 (908). [3] G. Shill, Svensk. Farm. Tid. 53, 65 (949). [4] H. Bayer, Z. Physik 30, 227 (95). [5] A. Novak and E. Whalley, Can. J. Chem. 36, 6 (958). [6] A. Weiss, Topics Current Chem. 30, (972). [7] H. Chihara and N. Nakamura, Bull. Chem. Soc. Japan 45, 3530 (972). [8] M. Buyle-Bodin, Ann. Phys. Paris 0, 533 (955). [9]. V. zmest'ev and V. S. Grechishkin, Zh. Strukt. Khim., 927 (970). [20] N. E. Ainbinder, B. F. Amivkhanov,. V. zmest'ev, A. N. Osipenkö, and G. B. Soifer, Sov. Physics-Solid State 3, 344 (97). [2] T. Kiichi, N. Nakamura, and H. Chihara, J. Magn. Reson. 6, 56 (972). [22] H. Chihara and N. Nakamura, Bull. Chem. Soc. Japan 45, 3530 (972).