AMIDINIUMIoxs ; HINDERED ISTERNAL ROTATIOX
August, 1963
Ai\lTDIKIUM IONS. I. HINDERED IETERNAL ROTATION’ BY GEORGE S. HAMMOND AXD ROBERT C. XECMAN, JR.~ Contribution
$9939 f r o m the Gates and Crellin Laboratorzes of Chemistry, Californza Institute of Technology, Pasadena, California Received February 1, 1968
N.m.r. spectra of unsubstituted and symmetrically substituted aliphatic amidinium salts in dimethyl sulfoxide The magnetic resonance characteristics of protons bonded to N in these amidinium salts are discussed. or water showed that rotation about the C-N bonds of the amidiniurn groups is hindered.
I n the course of an investigation of the thermal decomposition of azobisamidinium s a l t ~an , ~independent n.1n.r. study of amidinium ions in solution was initiated. Results elucidating the stereochemistry of these ions in solution and the n.m.r. spectral properties of protons bonded to N in amidinium ions are discussed in this paper. The mechanisms of nitrogen-proton exchange in dilute and strong aqueous acid solutions are presented in part I1 of this series4and are compared with those of ammonium ions. Results and Discussion Hindered Rotation.-The proton magnetic resoiiance spectra of the amidinium salts 1-111 have been recorded a t a spectrometer frequency of 60 &lc.p.s. NHR
/, -
CH,-C:+
\‘ -
NHR I
I1
X-
Ia, R =H, X =C1b, R = H, X = NO,c, R=CH,, X=c1-
TABLE I X.M.R. SPECTRAL RESULTSFOR A 4 30°, 60 b1c.r.s.
(1) R. C. Neurnan, J r , G. 8. Harnmond, a n d T. J. Dougherty, J . Am. Chem. SOC,, 84, 1506 (1962). (2) National Institutes of Health Predoctoral Fellow (1960-1962). ( 3 ) G. S. Hamrnond and 11. C. Neuman, Jr., J . A m . Chem. SOC.,86, 1501 (1963). (4) R. C. Neuman, Jr. and G. S. Hammond. J . Phgs. C h c m , 67, 1659 (1963). ( 5 ) J. A. Pople, W. G. Schneider, a n d H. J. Bernstein, “High Resolution Nuclear Magnetic Resonance,” NcGraw-Hi11 Rook Co., Inc., New York, N. Y., 19.59, p. 272. (6) 1,. M. Jackman, “Nuclear Magnetic Resonance Spectroscopy,“ Pergamon Press, Inc., New Yorl:, N. Y., 1959, p. 73. (7) E. Grunwald, .4. Loowenstein, a n d S. Meiboom, J . Chem. Phys., 26, 382 (1956); 27, 630 (1957).
J(N-€I,
AY’/2
Compd.
Solvent
Ia
DMSOd
Ib
DMSO
IC
DMSO
IC
HyO
IC
14% HzS04
IC
60% DySOa
VN-H”
(N-HIh
(C.p.8.)
(0.p.S.)
530 560 530 560 554 609
...
-425 -453
...
10 11 70 11 13 16 ,
.
.. ..
.,
U N - G C I I ~ ~ N-CHI)
(0.P.S.)
.
. t
(C.p.5.)’
...
...
...
171.5 176.5 177.5 187.0 179.0 188.5 184.0 194.0
5.0 5.0 0 0 5.0 5.0 0 0
534 9 ... ... 5-1-7 9 ... ... ... I11 DMSO 608 15 ... a Chemical shift in 0.p.s. referenced to external tetramethylsilane (0 c.P.s.). Positive values of u represent downfield shifts. Yo corrections have been made for bulk susceptibility effects. Spin-coupling constant obtained Signal width a t half-height. from the observed splitting in the N-CH, signals (see below). Anhydrous methyl sulfoxide (dimethyl sulfoxide).
I1
DMSO
H(b)
I11
(Table I). Assignment of X-H and K-CH3 protons is based on integrated relative intensities of the signals and agrees with the generally observed order V C - C H ~ < V & - C H ~ < vG-11. The chemical shifts of the N--H protons correspond closely to those of and ammonium salts.’ Two nitrogen-proton signals are observed for the unsubstituted amidinium salts Ia, Ib, and I1 in anhydrous DMSO. Integrated signal areas are in the ratio 1: 1. These results are consistent with the view that rotation about the C-K bonds is slow in comparison with the difference between the absorption frequencies of protons in the magnetically noli-equivalent “inside” (H(a)) and “outside” (H(b)) positions in IF’. Restricted rotation in the amidinium systems is no doubt due to the partial double bond character associated with each CCN-N bond. Because of the symmetry of the
~ IONS ~ IN ~ SOLUTIOK; ~ ~ ~
I+ X-13
R-C
13
I
S-11
(a)
,./ t-f
\
12-c
/ \+
N- 3H (a)
N-13
I
I
H
H (b) IV
ion, a bond number8 of 1.5 can be assigned to each CCY-X bond. For maximum ?r overlap, the group should be planar. Hindered rotation about Cco-N bonds in amides has been extensively investigated and attributed to a similar restriction of rotation about the C-nT bond.g That a single N-H proton signal is observed for I11 in DMSO (Table I) is in agreement with the above hypothesis since the nitrogen protons are restricted to the magnetically equivalent “outside” positions. The amidinium group in l\i,nT’-dimethylacetamidinium chloride (IC)could conceivably assume three possible conformations, Va-c. Spectra of this compound in a series of solvents are given in Table I and indicate that the conformation Va is the only detectable form present in solution. I n each (8) L. Pauling, “ T h e Nature of the Chemical Bond,” 3rd Ed., Coinell University Press, New York, N. Y.,1960, p. 239. (9) See, for example, ref. 5 , p. 36;.
u
~
GEORGE S. HAMMOND AND ROBERT C. NEUMMAX, JR.
1656
Vol, 67
first glance to be the most sterically favorable, the three methyl groups are approximately as crowded as the three methyl groups in 1,2,3-trimethylbenzene in which steric strain is obvious.13 The results thus demonstrate the predominaat presence of the unsymmet’rical conformation Va in solution.14 The possibility of formation of unsymmetrical ionpairs between the amidinium cation and an anion had been considered as an alternate rationale for the magnetic lion-equivalence of K-H protons in Ia, Ib, and I1 in DMSO. However, the magnetic non-equivalence of the X-CHz groups of ICin water (where ion-pair formation is expected to be negligible) and the magneticequivalence of the t’mo K-H protons of IIi in DNSO make this explanation untenable.16 The barrier to rotation about the C-X partial double bonds of I a in DMSO has been estimated. The effect of temperature on the K-H signals is shown in Fig. 1. The kinetic dat,a for the coalescence of these signals are given in Table 11. The observed signal separation as a
95.6’C 70. 8’C
TABLE I1 PROTON RESOKANCE SIGNALSor A 4 C E T A M I D I N I C ? d C H L O R I D E WITH TEMPERATURE O F DMSO
VARIATION OF
THE
N-H
ina/T,
T,OX.
Fig. l.--K-H
signals of acetamidinium chloride in DMSO as a function ot temperature.
H
H I
N-CH, CHJ-C,+/
N-H I
CH, V‘1
F.”3
I N-CHj
CH,-C:
N-H CH,-C
N-CHj I H Vb
N-H
I
CH3 VC
solvent two separate N-CH, resonances of equal area are observed. I n DMSO and 14% HzSO4, in which two K-H resonances of equal area are observed, each N-CH, signal is split into a doublet ( J = 5 c.P.s.). The S-H resonance signals in water and DZS04 are not observed, due, respectively, to rapid proton exchange and deuterium replacement of nitrogen proton^.^ Correspondingly, the S-CH, signals are singlets in these solvents.10 The magnitude of J in DMSO and in 14% HzS04implies that the splitting of the S-CH, signals is due to the adjacent proton on the same nitrogen atom.ll The spectra could alternatively be interpreted as arising from equal amounts of the “inside-inside” and “outside-outside” conformations T’b and T’c ; however, models indicate that steric interactioiis betx-een the ‘linside” S-CHJ groups in Vb are very unfarorable. il statistical mixture of the three conformations would give the results only if the chemical shift of the “inside” S-CHs group of TTamere identical with that of the “inside” N-CH3 groups of T’b. This is considered very unlikely. Although conformation T’c may appear at (10) The diffeience in chemical ahifts foi X-CH8 protons and N-H protons in the difieient solpente is due to a solrent effect on the slnelding of these protons (11) N-€1, N-CH? coupling constants of 4 and 4 8 c p s . are observed for N-methJlacetamide a n d N-meth>lfoimamide, respectire15 ’7 The oorrespondinp N-H quaitet 1s not resolied bemuse of the broadness of the N-H signal (Lzde i n f J a ) (12) G. I‘iaenkel and C. Franconi, J . B m Chem Soc , 82, 4478 (1960).
OK.
7 -
6ue/8uma
l/2TT8V
TZ = 0.03 sea.
Tz = 0.01sec.
3.22 1.00 0 0 2 . 9 1 0.85 i 0 . 0 3 0.24 i 0.03 0 . 0 2 f 0 . 0 3 .28 f .03 .06 f .02 2.85 ,132 f . 0 3 2.79 .68 i .03 .37 f .02 .14 f .01 2.77 .62 f .03 .42 f .02 .17 f .01 he = observed signal separation, 8 v m = signal separation a t 37.6’. r = mean lifetime of protons in the two environments, T z = transverse relaxation time. 310.6 344.0 350.8 3%1 : 361.2
function of temperature has been corrected for overlap of the signals by the method of Gutowsky and H0lm.l’ Their treatment requires that the transverse relaxation tinies of the signals be iiivariant n7ith temperature. This was not true for the IC’-H signals of Ia. A qualitative inspection of Fig. 1 shows that a t about 20’ above the coalescence temperature the single signal is much broader t’lian the initial signals a t 37’. Further increase of temperature did not noticeably sharpen this signal. The initial signals (37’) give a value Tz 50.03 see., while the final signal (115’) gives T2 0.01 sec. Using the data in t,he last two columns of Table 11, activation parameters were calculated for each of these values of T zfrom the graphical plot,s shown in Fig. 2 and 3. They are: T , = 0.03 sec. (Fig. a),E, = 9 2 kcal. mole-I, h-, = lOj-10’ sec.-l; T z = 0.01 see. (Fig. 3), Ea = 25 f 8 kcal. mole-l, k, = 1011-1021set.-'. Since T 2 probably decreases in some regular manner with increase in temperature the true activation parameters lie somewhere between these extreme values. The dependence of Tz for the 3 - H signals on temperature is probably due to the effect of temperature on the (13) H. C . Brown, G. K. Barbaras, H. L. Berneis, W. H. Bonner, R . B. Johannesen, 31. Grayson, and K. L. Nelson, ibid., 7 6 , 1 (1953). (14) Small amounts of the other conformers may be present, b u t are not detected b y the n.m.r. method. Rapid interconversion between the conformers would require the spectra to show only one N-CHI signal and one N-H signal. (15) The effect of ion-pairing in DMSO on the magnetic shields of ring protons in aniliniuin ions has been discussed,16 b u t appears t o have no apparent consequences in these studies of amidinium ions. (16) G. Fraenkel, Abstracts of the Symposium on High Resolution N. m.r. Spectroscopy, July 2-4, 1982, Boulder, Colorado. (17) €1. S. Gutowsky a n d C. H. Holm, J . Chem. Phgs., 26, 1228 (1956).
August, 1963
1657
AMIDINIUM I O N S ; ~ ~ I S I ) F ; H E I ISTICI1N.41, ) 1iOT.il‘IOS
quadrupole broadening of protons bonded to
14K
1 .7
(uide
infra). Although the barrier to rotation in Ia is not knon-n with high precision, it is interesting that the value is iii the same range as those of amides (7-18 kcal./ mole) . y , 1 8 * 1 9 It might be cxpcctrd that barriers to rotation in amidinium ions would be consistently higher than those in simple amides due to the greater doublc, bond cliaractu associatcd with an aniidiiiium C-S bond. It is reasonable to assign a bond order of -1.3 to each C=K bond in symmetric amidinium ions (vide supra) whereas tlie corresponding bond order for the C-N bonds in amides should he smaller.9I 2 The existing data for amidiniiim ions is insufEcient tjo test this hypothesis. Absolute Spectral Assignments-A tentative absolute spectral assigrinient can be made by a comparison of the spectra of Io and 111 in 1131SO (Table I). The similarity htween the chemical shift of the “outside” proton i n 111 (608 c.P.s.) and the loivcr field ?;-H signal i n IC at 609 C.P.S.suggests that this latter resonance is due to the “outsid(>”protons; therefore, thc signal a t 534 C.P.S.is due to the “inside” proton. Ratioiialc for this comparison lies in the similarity of substitution of each nitrogen in ICand 111. It is reasoilable to assume that the relative screeiiiiig of “iriside” arid “outside” nitrogen protons is similar in each amidinium ion and is dctcrrnined by the position of the proton relative to the
1 (i
+,“
(18) M.T.Rogers i i r i d J. (”. Woodbrer, .I. I’hys. Chem., 66, 840 (1962). (19) 1%. Siinncrs, I,. 11. Piette, arid W. (?. Schneider, Can. J . Chem., 38, 681 (1!)00). ( 2 0 ) V. J. Rowalcwski arid 1). 0. de Iion-aleu.ski, .I. Chern. Z’hye., sa, 1272 (lW30); A r k h Kemi,16, 373 (1901). (21) lt. I t . lhsc:r, Can. .I. C l ~ e m .38, , 5.10 (1960). (22) J. 1%.liichards and W. E’. Jieaclr, J . Vrw. Chcm., 26, 628 (10(il).
““
\
31
1.5
k N \ e
-r 1.4
2 80
1/T
x
2 8s 10’.
2 90
2 95
Fig. 8.-Estimntion of activation energy for iotation in amidhiurn ion assuming 7’2 = 0.03 sec.
+
N-C-n’ linkage. Thus it can be tentatively assumed that the lower field S-H signals in Ia-c and I1 correspond to “outside” protons and the higher field X-13 signals to “inside” N-H protons. l’liis conclusion is further strengthened by an analysis of the N-CH, signal shapes of IC. In all solvciits, the lower field S-CH3 resonance pattern is much sharper than the high field N-CHJ resonancc. For example for ICin water, values of Avl for the signals are: x-C& (low field) 1.0 c.P.s.; S-CH3 (high field) 1.7 c.P.s.; and C-CH,. 1.6 C.P.S. The relative line widths may be attributed to unequal (unresolved) spin coupling bctween the C-CH3 protons aiid the conformationally different S-CHs protons. Thus the strongest coupling is between the C-CHs protons and the high field S-CH3 1,4-coupling group. In amides20and other systems21.22 between methyl group protons has established the relationship J,,,,l, > J c f v . This leads to the conclusion that the high field S-CII3 group is trans to the C-CH3 group or in the “inside” position. Data to be given i n the second paper of this series4 show that the high field member of tlie two S-CII, resonances and the low field member of the two N-H resonaiices of ICare due to groups attached to the same nitrogen. Thus, the assignment of the high field N-CE-I, resonance to the “inside” proton requires the low field K-H resonance to be assigned to tlir “outside” proton. This is in agreement witli the previously discussed assignments based on cornparison of IC and 111. Analogous assignments
’k
:im t -
1 .5
I .I1
@i
+ ‘c
6
0
\ 31
-
ts
0.5
0.0
__ 76
+
2.80
+
2.85
.L ___-’
I _
2.90
2 ,!)5
1 / T X 108.
Pig. 3.--Estimation of activation cwergy rotation in aniidiniurn ion assuming T2 = 0.01 SI’(;.
RC(
t-
using the abovc arguments have been made for X,X-dimetliylaniidrs.w2” Spectral Characteristics of Amidinium Nitrogen Protons.-The shapes of the N-H resonance signals are also of interest. This discussion will be limited to the results obtaiiied in D;\lSO solutions. In general, the spectral propcrtirs of N-H protons are strongly dependelit on ~vhctlier or not they are involved in cxcliangc proccsscs.?j Care was taken to assure that (23) Itefcrencc 6 , I). 72.
1658
GEORGE S.HAMMOXD AND ROBERT c. S E U R I S N , JR.
the DRISO used in these studies was anhydrous and the observable S-H, &CH3 spin-spin coupling for ICin DMSO (Table I) indicates that proton exchange is not occurring a t a significant rate in this solvent. Further proof for the anhydrous nature of this solvent lies in the observation that addition of small amounts of water to DMSO solutions of Ia causes the two separate N-H signals to coalesce into a broad singlet. Under such non-exchanging conditions various types of S-H signals have been observed. Ammonium salts2* customarily show the triplet pattern arising from coupling of the protons with 14X ( I = 1) as do anhydrous ammonia26 and certain amides a t elevated temperatures.26 On the other hand, certain amides at room temperature show only broad singlets with values of A v ~ , of , 10-75 C . P . S . ~ ~ Tiersz7has examined a series of n'-acylamino acids and guanidinoacetic acid under noli-exchanging conditions aiid finds values of Am,* in the range of 6-10 c,p.s. All of these results are explained in terms of the effect of the symmetry of the electrical field on the coupling of the I4IT quadrupole moment with proton magnetic moment.23 26-26 Coupling of this electric field with the 1 4 5 nuclear niomeiit results in efficient relaxation if the field is highly asymmetric ; relaxation is relatively inefficient if the field is highly symmetrical. The former case gives rise to a relatively sharp singlet and the latter to the triplet pattern. Broad N-H singlets are taken as indicative of electric fields of intermediate symmetry about the nitrogen atoms. Iiispection of values of for the P\'-H protoiis of ainidinium ions in DRISO (Table I) and comparison with the narrow h'-R signals observed by Tiers27 for a-amino acids and the broad S-H signals for amidesz6 leads to the conclusion that amidinium nitrogen atoms are surrounded by relatively asymmetric fields. Robe r t P has shown that the broad singlets observed for N-H protons of certain amides undergo transitions to triplets a t elevated temperatures and concluded that increased molecular motion effectively symmetrizes the field surrounding the 1 5 atoms. The transition temperatures were dependent 011 structures of the amides but in general were above 175'. We conclude that the increased width of the coalesced X-H signal for I a in DMSO a t 115' is a result of the same process and that a t still higher temperatures further broadening mould occur and eventually a triplet pattern would arise. Dihedral Angles and N-H, C-H Proton Coupling.Azo-bis-(N,N'-dimethylem)-isobutyramidiniumnitrate (111) deserves special mention. Spin-spin coupling between the nitrogen protons and protoiis on the adjacent N-methylene groups in this molecule has not been observed in anhydrous DMSO or sulfuric acid solutions. Rapid nitrogen-proton exchange does not satisfactorily explain these results since it has been demonstrated for other amidinium salts that exchange of nitrogen protons in sulfuric acid solutioiis of appropriate composition* and in anhydrous DMSO (uide supra) is extremely slow. We postulate that the absencezgof spin-spin coupling is due to the magnitude of the dihedral angle (probably (24) See foi example AI. T Emerson, E. Grunnald, a n d R. A. Kromhout, J Chem. Phus , 33, 547 (1960). ( 2 5 ) R. A. Ogg, zbzd., 22, 560 (1954). (26) J. D. Roberts, J . Am. Chem. SOC, 78, 4495 (1956). ( 2 7 ) G V. D. Tiers a n d F. A . Bovey, J . Phys. Chem., 63,302 (1959). (28) RPference 5, pp. 102, 227. (29) Coupling of less t h a n 1 c.p.s. probably mould not be resolved under our spectral Conditions.
Vol. 67
close to 60') between the methylene protons and adjacent nitrogen protons in this rigid ring system. Karplus30 and C01iroy~~ have discussed the effect of dihedral angle on vicinal proton coupling constants for saturated carbon systems. Maximum csuplirig is observed a t $ = 0' and b = 180' and minimum coupling occurs a t $ E 80'. For these satmuratedcarbon systems, J 1 2 a t 4 = 60' is predicted to be about 307, of the value found in open-chain systems such as ethyl and isopropyl groups. Since the S-H, K-CH3 1,Zcoupling constant for IChas been found to be 5 c.p,s.,.\.rrhileJ12forS-methylamides is -4-5 c.P.s., a rough approxiniatioii of the expected magnitude of coupling iii I11 is JIZ = 1-1.5 c.p.s. It is not necessary to discuss the lack of resemblance of an amidiniuni group to a saturated carbon system but it is perhaps significant t o observe that 110 resolvable coupling is observed in I11 which possesses a geometry similar to that expected to minimize coupliiig in thc saturated carbon systems. Experimental Materials. 32-Dimethyl sulfoxide (DMSO'i, Crown-Zellerbach, was distilled a t reduced pressure, shaken with Rloleculilr Sieve, Linde Co., Type 4.4, filtered through Celite in a drybos under nitrogen, and redistilled. The middle fraction, b.p. 45' at, 0.2 mm., was collected for use. Sulfuric acid, DuPont reagent grade, was used without further purification. Sulfuric acid-d2 was prepared33 from Baker and Adanwon "Sulfan B" (sulfur trioxide) and deuterium oxide (99.5yo),General 1)ynamics Corporation. Acetamidinium chloride (Ia) was prepared according to the method of Pinner,S4 recrystallized from absolute ethyl alcohol, and dried over phosphorus pentoxide in vacuo; m.p. 165-168" (lit. m.p. 164-166",34a166-167"34b). Acetamidiniuni nitrate was prepared from an aqueous solution of the corresponding chloride (IR)by additsionwith stirring of an eqnivalent amount of aqueous silver nitrate. The resultant silver chloride precipitate was filtered OR and the arnidiniuni nitrate was precipitated from solution by addition of a large qua,ntity of acetone, recrystallized from absolute et,hyl alcohol, and dried over phosphorus pentoxide in vacuo; m.p. 159-190". Anal. Calcd. for C2H;Xn0.,: C, 19.53; H, 5.83; N, 34.70; 0 , 39.64. Fonnti: C, 18.44, 15.22; H, 5.83, 5.97; S , 33.07, 34.88. The micronnalyst (Spang) reported violent, esplouion during comhustion under oxygen. N ,N'-Dimethylacetamidiniumchloride (IC)was prepared according t o the method of Pinner, 86 recrystallized from atsolute ethy! alcohol, and dried over phosphorus pentoxide in vacuo; n1.p. 214.5-215.5° (lit.86 m.p. 218"). ICisveryhygroscopic. Hydrolysis of IC in 1 N sodium hydroxide solution g w e equivalent amonnts of methylamine and S-niethylacetaniide. This was shown by n.ni.r. analysis and confirms the symmetrical st,ructure. Azobisisobutyramidinium nitrate (11) was prepared from the corresponding chloride by a procednre identicd with t,hat given for the preparation of acetaniidiriiuni nitrate (Ibl. The chloride T ~ obtained from t.he Yerkes Research T,ahoratory, E. I. du Polit de Nemoure and Co., and had been recr,lrstallized from mater. .4nnl. Calcd. for CxH?oNaOS: C, 29.63; 11, 6.22; N, 34.55; 0,29.60. Found: C, 29.175; H, 6.30; N , 34.46; rn.p., 162.5" dec. Azobis-(N,N'-dimethylene'i-isobutyramidiniumNitrate (111). A 10-g. sample of unrecrystallized az~bisisobutyramidinium chloride, obtained from DuPont, was dissolved in 100 ml. of 98% ethylenedianiine, Matheson, Colenian and Bell, with stirring. Cpon stirring the solution for several minutes a t rnom temperature a white solid precipitated, was filtered OR,and recrystnIlized first from chloroform and then absolute methyl alcohol; m.p. -
(30) ill. Karplus, J . Chem. Phgs., 30, 11 (1959). (31) H. Conroy, Sduan. Org. Chem., 2, 308 (1960). (32) Melting points are uncorrected. Microanalyses were done by Elek Micro Analytical Laboratories, Los Angeles, Calif., and Spang Microanalytical Laboratory, l l n n Arbor, Michigan. (33) A. P. Best and C. L. Wilson. J . Chem. Sac., 239 (1946). (34) (a) A. W. Dox, "Organic Syntheses," Coll. Vol. I, John Wiley and Sons, Inc., New P o r k , X. Y., 1941, p. 5; (b) A . Pinner, Ber., 16,1643 (1883); 17, 171 (1884). (35) A. Pinner, "Die Imidoather und ihre Derivate," Berlin, 1892, p. 112.
S
122.5’ dec. rlriul. Calcd. for n~olr~is-~,N’-tlinictliylerieiPoTemperature Dependence Study-Acetamidinium Chloride in hutyramidine ( C I J L S e j : C, 57.56; H, S.%; S , 33.57. Foiird: DMSO.--The sarnplc used in this study wiis prepared by disC, 57.65; 11, 8.97; N , 33.26. The hydrochloride of C,,lipiSe solving acctamidiriium chloride fI:i), recrystallized from ethyl way prepared. .Inal. Chlctl. for CI*H?IS&12: C, 44.58; H , alcoliol :md dried in uucuo, in anhydrous DlISO. The solution 7.40; N, 26.00; CI, 21.!)4. Found: C, 44.48; H, 7.51; S , contnined approxirn:ttely 13% of ILL:tiid was seiiled in a Pyrex 23.88; CI, 2136. The nitrate x-as prepared by addition of an n.m.r. trihe. A varitthle temperature probe with :t Pyrex dcwar oquivalcnt :mount of concentrated nitric acid to an absolute encasing the insert, was w e d . @ The ternperature in the probe ethyl alcohol solution of thc :tmidinc, CI2HnKn,anti recrystitllizrd insert, was rneasured hy means of N. L.ol.)per-constnrit,an tliernioG : placed within the probe insert in conjunction with :t from water; m.1). 146.5’ dec. -4nuZ. C:tlcct. for C I ~ H J ~ S ~ ~ Ocouple C, 38.29; 13, 6.43; K, 29.77; 0, 25.51. Found: C , 38.27, Leeds-Kort,hrup precision potentiometer. Tlle tenipcrature 38.83; H, 6.29, 6.17: S,49.Y2, 29.66. Thernid decomposition ‘ipu determined both before and after the spectrum was taken. of this nitrate salt in I)hISO gives a stoichiometric quantity of The peak sep:ir:ition was determined by the audio side-l):ind nitrogen gas based on the stnictur:tl formula 111. The rnicro:L Hrwlett-Packard Model 200 2 1 1 % riudio t e c l i n i q i ~ c using ~~ malyst (Spang) reported a violent expiosion during combustion oscillator and hlodel 5 2 1 4 frequency counter. The values under oxygrn. ‘/pir€iv were calcul:~t,edhy tlie method of Gritowsky and Holmi7 using the par:imeters 6v, = 29 c.P.s., (a) T2 = 0.03 sec., and N.m.r. Spectrometers.- The nuclear magnetic resonance ( b ) Yr = 0.01 sec. The two v d u e ~of Y’z were obtained from the spectrometers ured in this study were a Varian Model 5’-4.100 rclatioiistiip 2,”I’2 = 2 , , & 1 , ~where A Y ~ is, ~the width a t half13(HR-60) spectrometer, equipped with a Super-stabilizer, height of tho nitrogen-proton rcmmnce signal. The value 7‘2 = operating :it v u = 60 11e.p.s.; and H Vttrian 12-60 spectrornc~ter 0.03 soc. W:LS c:tlcuhtotf from theavcrttgo valuct A v l , , = 10.5 3~ 0.5 operating :it ul1 = RO 1lc.p.s. The majority of the spcctr:r werc :tnd , the value 7’, = 0.01 RCC. WHS c.p.s. for the two sirqids a t : < i o taken with the A-60 bpcctrometcr used ivithout modification. r . c:~lcril:itc~i from the valuc Ad/‘ = 30 c.p.s. for the coalesced 1 he variatde tcrnper:tture experiment W:LS performed using thc sigrial :it I 15”. T h o rate csprcssion iiscd for calculating activaIIll-60 spectronieter with suitable niodifications as descrilwd tion par:tnietcrs is that given in eq. 1 . 1 7 below. N.m.r. Sample Tubes.--Stand:trd Irarian antilytical s:iruplc tubes (4.28 mm. i d . ) were used for all spectra takcri with tho A-60 spectrometer. Sample tnhes for lice with tlie HI1-60 spectrometer were made from Pyreu g h s s tubing (5-mni. o.d.). All tribes were tlioroughly cleaned and dricd hcforo use.
(80) C . l’miiciini antl Q. Fracnkel, R e i . Sct. Inatr., 31, 657 (18G0). (d7) .I. T. Arnold and hl. E. I’ackartl, J . Chem. I ’ h p . , 19, 1608 ( 1 l 6 1 ) .
Gales und C ~ c l l i r iLabornlor ies oJ Cheirr istry, CnliJorniu Instil ute of ‘I’echnoloy y, I’asadcnta, California 12eceiocd Icebruurg 1, 1963
Kinetics of exdiaiige rcactions of protoris att:whetl to riitrogcn in :tmidiniurn ions 1i:tvr twcn stritiied in dilute hydrocliloric~ acid :tnd in various water-sulfuric wid mixtures. In diluto aqueous : i d , ttic principal merlianism for cxvliange of protons with the solvent appears to involve reaction of the amidiniurn ions with hydroxitic ion. In concentrated sulfuric arid the cxrhange of protons betwecw c~onforrnationallydistinct, “inside” and “outside” positions involves an acWcsatalyzed mechanism :LS dow the cxch:ange with solvent. Tlie two nitrogen atoms in N,N’-di~ricthylacetarnidiniumion-exchange hydrogen ions with tlic solvent at different rates.
The 1i.m.r. spectra of protoiis bound to nitrogen of amidiriiuin ions were reported in part I of this series.l I t was shown that rotation about the C-T\; bonds is rest’rict~edand that. protons in “inside” and “out.side” positions are Inagiletically distiuguishablc. The results of studies of t~lickinetics of various exchange reactions of arnidinium ions in aqueous acid arc reported ill this paper. Tlie work was aided substanttiallyby considcration of the reports of very careful studies of the exchange rcactioiis of ainnioiiiuni Results and Discussion Dilute Aqueous Acid.-hcetamidiniuni chloridc, 1, was used ill this study. I’rclimiiiary 1i.in.r. expcri( 1 ) 11. C . Neutnan9 J r . , (;. S. Ilamniond. antl T. .J. I)ounllerty, J . Am. C h ~ m S. o c . . 8 4 , 1300 (1002). (2) National Institutes of Ncalth Prcdoctor:il I~c!llou. (19(i0-1862). ( 3 ) ContrilJution No. 2!)-40. ( , I ) C,. S. Ilarnmrind a n d It. C. Neunian, Jr., J . I’hyn. Clirm.. 67, lli35
(l!lIi3). ( 3 ) (a) .\I. 1’. ISniwson, 1,;. Criinwald, and It. A . I i i o m h w t . J . Chetn. I’hys., 33, 3.17 ( l Y i X 1 ) : (b) K . Ciirn\vulil, 1’. .J. I
