4886
R. KENThlURMANN AND HENRY'TAUBE
VOl.
'is
tion of certain 2,2'-bipyridinelg and ethylene- tion of d-cis-[CoenzFZ]+ is analogous to the bediamine20 metal complexes were likewise observed havior which Mathieuzl reports for 1 4 s - [Coenzto be acid catalyzed. Finally i t is of interest to Clz]+. He suggests that since the rate of raceminote that the rate of alcoholysis of benzyl fluoride zation of d-cis- [CoenzH20C1l2+is not dependent 011 is acid catalyzed but that of other benzyl halides the rate of replacement of the chloro group, it is not;'O this is analogous to the behavior of the must involve a dissociation of the coordinated haloamminecobalt(II1) complex ions. water to yield a symmetrical intermediate. ReThe rate of loss of optical activity of an aqueous centlyZ2i t was shown that the loss of optical acsolution of d-cis-[Coen2F2]+ a t a pH of 1 shows tivity is primarily due to isomerization to the symclearly that this proceeds in two distinct steps metrical trans- [ C 0 e n ~ H ~ 0 C l which ] ~ + can in turn (Fig. 1). There is first a change in optical rotation regenerate the racemic cis isomer. Tentatively followed then by a slower loss of optical activity. it can also be concluded that the racemization of Within rather large limits of experimental error, d-cis- [ C O ~ ~ ~ H ~likewise OF]~+ proceeds by an exthe initial mutarotation proceeds a t approximately change of coordinated water with the solvent. the same rate as does the hydrolysis under these Before more definite statements can be made, i t same conditions (Table IV). This suggests that must be determined whether or not the rate of d-cis- [CoenzFz]+isconverted to d-cis- [Coen~Hn0F]~+water exchange is sufficiently rapid to permit such without extensive loss in optical rotation. The a mechanism. One final point of interest is that fluoroaquo product in turn undergoes racemization the rates of racemization for a series of complex at a rate independent of the rate of release of the ions of the type [Coen2HzOX]decrease with changes second fluoride. No detailed kinetic study was of X in the order C1-> F-> NOz-NCS->> made of the hydrolysis of the second fluoride ex- NH3. If all that is involved in racemization is the cept to observe that i t was very slow compared to water exchange, then it is surprising that the chlorothe rate of loss of optical activity. aquo complex racemizes approximately 200 times faster than does the nitroaquo whereas the rate of TABLE IV acid hydrolysis of cis- [CoenzClZ]+ is only twice that RATESOF ACID HYDROLYSIS, MUTAROTATION A N D RACE- of cis-[CoenzNOd21]+. This can be understood if MIZATION OF d-cis-[CoentFn] + GIVENIN k(min.-'). the replacement of water in the nitroaquo complex Temp., Acid MutaroRacemizatakes place a larger percentage of the time without [HNOP] OC. hydrolysis tation tion loss of optical activity as compared t o the chlorono acid 25 1.8 X 1 0 ... ... aquo compound. This is in agreement with the 0.001 25 3 . 6 x 10-4 4 . 9 x io-' -2 x 10-4 25 3 X 10-2" 6 . 2 X 10-8 3 . 9 X 10-4 0.1 r-bonding hypothesis' which predicts that the con0.1 35 1 x 101 . 3 x 10-2 4.7 x 10-4 tribution of n-bonding electrons from chloride to a The method of removal of complex on an ion exchange resin prior to titration o f fluoride was employed. Be- cobalt in the pentacoordinated intermediate, [enzenhances , the formation of a trigonal cause of difficulties described in experimental these data are C O = C ~ ] ~ + only semi-quantitative. bipyramid structure whereas the tetragonal pyramid structure is favored for [enzCo-NO2I2+which This change in optical rotation of an acid solu- does not have this type of n-bonding. ( I O ) J. H. Baxendale and P. George, Trans. Faraday S o c . , 46, 736 (1950) : F. Basolo, J. C. Hayes and H. hl. Neumann, THISJOURNAL, 75, 5102 (19.53). (20) J. Bjerrum, K. G. Paulsen and I. Paulsen, "Symposium on Coordination Chemistry," Danish Chemical Society, 51 (1953).
[CONTRIBUTION FROM THE
DEPARTMENT OF
(21) J. P. Mathieu, Bull. SOL. chim., [ 5 ] 4, 687 (1937). (22) R. G. Pearson, R . E. Meeker a n d F. Basolo, THISJOURNAL, 78, 2673 (1956).
EVANSTON, ILL.
CHEMISTRY, UNIVERSITY O F CHICAGO, AND UNIVERSITY O F CONNECTICUT]
The Mechanism of the Formation and Rearrangement of Nitritocobalt(II1) Ammines' BY R. KENTMURMANN~ AND HENRY TAUBE3 RECEIVED MAY2, 1956 The formation of a group of nitritocobalt(II1) ammines from the aauo complexes proceeds in weakly acidic solutions without breaking the cobalt-oxygen bond. The formation of the nitro derivative from nitritopentammine-cobalt (111) ion proceeds in water solution by an intramolecular process involving either the formation of a pentacoordinated intermediate followed by instantaneous reaction with the group released or via a heptacoordinated activated state.
I.
Introduction
The rate of substitution of a ligand for another on cobalt(II1) and chromium(II1) coordination compounds is often quite slow. Certain ions, however, substitute more rapidly and a correlation ( 1 ) This work was supported in part by the O N R under contract N6-ori-02026. (2) University of Connecticut, Storrs, Connecticut. (3) George Herbert Jones Laboratory, University of Chicago, Chicago, Illinois.
between the nucleophilic character of these reactants and the rate and mechanism of the reaction has been made.4 A consideration of the literature values of the rates of substitution reactions reveals a group which seem to proceed more rapidly than one would anticipate on this basis, Comparison of the following rates indicates that the fast group consists of the reaction between an aquo-coordination compound and a labile-oxygen containing (4) D. D. Brown a n d C. K. Ingold, J . Chem. SOL.,2680 (1953).
FORMATION AND REARRANGEMENT OF NITRITOCOBALT(III) AMMMINES
Oct. 5 , 1956
ion. Thus there is a plausible explanation for the abnormal speeds in the possibility of the cobaltoxygen bond remaining intact. k25oa
-
+ +'+ +' + +
[CO(NH3)aSO4] + HzO + [CO(NHa)s(OHz)] sod[Co(NHs)sCl] Hz0 [Co(NHa)dOHz)]+3 C1-
[Cr(Hz0)6] .".
7.2
x
10-5
1.2
x
10-4
4.7
x
10-4
1.8
x
10-3
4.7
x
10-4
+'
+
8 X
+
7
x
10-3
+'+
pH5.4 ~ i s - [ C o ( e n ) ~ ( H ~ O ) ( N 0 ~ ) ]NO*- _?c cis- [ Co( en),( NOZ)Z] HzO 8 X [Co(NHa)s(OHz)] +3 HNOz [ C O ( N H ~ ) ~ O N O ] + 'HsO' 2.5 x lo-' [Co(NH3)aHC03] Ha0+ + [Co(NH3)sOHzI+3 f coz 1.9 x 10-1 [Cr(H20),]+a polymolybdate + [Cr(polymolybdate)e] HzO rapid H+ [ C O ( N H ~ ) ~ O N O ]and [Cr(NH3)50NO] --f [M(NHa)50Hz] HONO rapid
+ + + +'+ +
+
+
+'
+' +'+
+
PH 4
cis-[Co(en)z(NOz)(H~O)]+e HNOZ ~is-[Co(en)~(NO~)(ON0)1 i. HaO+ ~is-[Co(NH3)4(HzO)z] +3 Mood[Co(NHa)~(MoOa)l+ 2Hz0 t i s - [ Co( NH&( OH&] +a CrO4- + 2H20 [Cr(NH3)4Cr04] + a kl in mim-1, ka in liters mole-' min.-1.
+
+
+
+
__f
+
2.1
+
x
state or in solution. It has been p o s t ~ l a t e d that ~,~ this is an intramolecular reaction on the basis of the solid state reaction and the rate in solution being essentially independent of the nitrite ion concentration. This paper reports the results obtained using 01* as a tracer, to elucidate the mechanism of the formation of nitrito compounds of Co(II1) and Cr(111) ammines and of the nitrito-nitro conversion. 11. Experimental
+ Hz6
[Cr(H,6)6] HzO cis-[Co(en)~(NOz)Cl] +l HzO -+ cis-[Co(en)z(NO~)(H~O)] +z C1-
4887
10-1
rapid rapid
Isotope studies on the acid decomposition of [Co(NH3),C03] + have shown that it is the carbonoxygen bond which is broken while the oxygen originally attached to the cobalt remains attached in aquo The acid decomposition of [Co(NH3)4C03]+lt however, results in one of the two cobalt oxygen bonds being broken.6 It has been reported' that the formation of [Co(NH3)60NO] f 2 from the aquo salt proceeds in slightly acid medium according to the following rate law rate = k( [CO(NHB)~OH] +z)(HNO~)z
The interpretation is that N203 (NO+N02-) attacks an unshared pair of oxygen electrons forming the nitrito compound and releasing H+NOz-. It was suggested that if this were the mechanism, then the cobalt-oxygen bond would not be broken. The rearrangement of nitrito compounds to the nitro form proceeds either in the anhydrous solid (6) J. P. Hunt, A. C.Rutenberg and H. Taube, THISJOURNAL, 74, 268 (1952). (6) F. Posey and H. Taube, ibid., 76, 4099 (1953). (7) R. G.Pearson, P . M. Henry, J. G . Bergman and F. Basolo, ibid., 76, 5920 (1954).
A. Preparation of Compounds.-The starting materials [Co(NH3)a(OHz)](C104)3-H~0,9cis- [Co(en),( IGH3)ClI( N0a)2,10 cis- [Co(en)~( NO,)Cl]( TS03),11 cis- [Co(NH&( O H Z ) Z ] C ~cis~ , ,[Co(en),( ~~ NO,), ]N03,13 [Cr(N H ~ ) E , C ~ ] C ~ Z , ' ~ were [Co( NH3)5(,r';Oz)j(C104)216 and cis- [Co(en)zCl~]Cl~~ prepared according to procedures in the literature. They were recrystallized twice and dried under vacuum. Aquo compounds enriched in 018were prepared in two ways: ( 1 ) equilibration of the normal aquo salt with acidified enriched water for 48 hr. a t 50", in the dark, followed by precipitation with sodium perchlorate or bromide; (2) triturating the pure chloro compound in enriched water with 0.2% less than the equivalent amount of Agr';Oa or AgClO4 a t 5" and removing the silver halide on a filter. The filtrates which gave no test for silver ion and only a faint test for halide ion, were evaporated to dryness under vacuum a t 20'. The crystals were washed with alcohol and dried or, if an oil was obtained, it was solidified by repeated extraction with absolute alcohol. The chromium compounds were protected from light to prevent decomposition. It is known that there is little cis-trans conversion in these reactions." The nitrito compounds were prepared by the addition of an excess of sodium nitrite to a saturated water solution of the aquo complex a t 5'. Perchloric acid was added until a PH of about 4 was reached and after 10 minutes the crystalline product collected on a filter. (In cases of cis-bis(ethylenediamine)-amminenitritocobalt( 111)and cis-tetramminedinitrito-cobalt( 111) ions i t was necessary to add an excess of sodium bromide to obtain a product.) The nitrite compounds were washed with 95% ethanol to remove the adhering sodium nitrite, recrystallized by solution in the minimum amount of water a t 5" and addition of a cold saturated solution of sodium perchlorate or bromide. The crystals were repeatedly washed with cold alcohol, then with acetone and dried under vacuum a t 10". Recrystallization was absolutely necessary in order to remove the last traces of NOZ- which, if present, seriously interfere with the determination of the O1*-content of the compounds. Since the nitrito compounds are unstable with respect to the nitro form, they were stored a t approximately 0' and in all cases were used within an hour of their preparation. Anal. Calcd. for [ C O ( N H ~ ) ~ O N O ] ( C ~ OCo, ~ ) ~15.14; : N, 21.62; H , 3.88. Found: Co, 15.21; N, 21.46; H , 3.69. Calcd. for cis-[Co(en)~(NH3)(ONO)]Brz: Co, 14.65; N,20.92: Br, 39.8. Found: Co, 14.59: N, 20.76: Br, 39.6. Calcd. for cis-[Co(en)z(N0z)(0NO)](C104):' Co; 15.90; C, 12.95; H , 4.35; N, 22.70. Found: Co, 15.80; C, 13.08; H , 4.39; N, 22.58. Calcd. for [Cr(NH3)5ONOIBrz: N, 24.52; Br, 46.6. Found: N, 24.29; Br, 46.3. Calcd. for [Co(en)~(ONO)p]C10~:Co, 15.92; N, 22.70. Found: Co, 15.78; N,22.79. Isotopically labeled nitrito compounds of the type (Rr
*
*
CO-0-N-0)
were prepared from the enriched aquo compound by treatment with an aged, acidic, enriched water solution of enriched NaNO2. Separation and purification were the same as for the unenriched samples except that enriched water was used for recrystallization. Complete ex( 8 ) B. Adell, Svcnsk. Kcm. Tids., 6 6 , 318 (1944). (9) F.Ephraim. Bcr., 66, 1936 (1923). (10) S. M.Jorgensen, J . prakt. Chem., [2]41, 454 (1890). (11) A. Werner, Licb. A n n . , 886, 248 (1912). (12) S. M.Jorgensen, 2. anorg. Chcm., 8 , 294 (1892). (13) A. Werner and E. Humphrey, Bey., 84, 1821 (1901).
(14) 0.T.Christensen, J . prakt. Chcm., 8 3 , 54 (1881). (15) F.Ephraim, Ber., 66, 1532, 1540 (1923). (16) A. Werner, ibid., 44, 875 (1911). (17) A. Werner, Bcr., 44, 2445 (1911).
4888
R. KEXTT~IURMANX AND HENRY TAUBE
change of the nitrite ion with water was assumed and fractionation effects were small enough t o be neglected. Sodium nitrite, enriched in 0l8,was prepared by mixing known amounts of KaNOe of normal isotopic composition with a large amount of enriched HzOand acidifying the solution to a p H of 4.5 with 70y0 perchloric acid. After 3 hours at room temperature the exchange was complete and SacHPOl (anhydrous) was added to a p H of about 7 . The water mas removed under vacuum. The presence of sodium phosphate does not interfere with the preparation of the nitrito complexes. All common chemicals mere of reagent grade and were purified by crystallization where necessary. B. Isotopic Determination.-In order to obtain a gaseous sample representative of the 018-content of the nitrito complexes, which could be introduced into the mass spectrometer, thermal decomposition of the complexes was carried out. Unfortunately a mixture of oxides of nitrogen was obtained which could not conveniently be used. The 0l9-content of N20can be measured conveniently and accurately in the mass spectrometer, and has the advantages of being relatively inert and condensable allowing its separation from COS, N2 and 02. Thus the nitrito complexes reacted to give NOn- which was converted t o N20which had the same oxygen isotopic composition as the average of the t x o oxygens of the nitrito compound. It has been shown t h a t NaO exchanges with the solvent very slowly18a and the observation t h a t KO?-exchanges at high pH's has been described .18b The latter exchange has never been noted even in the presence of a large excess of azide ion. It would appear t h a t the inorganic complex is very successful in preventing this exchange. Three methods were used to obtain NOZ- from nitrito compounds. Method A.-The action of 0.1 N S a O H at about 40" for 1 hour yielded NO*- and either the corresponding hydroxy complex or at higher temperatures degradation products including C O ( O H ) ~and hTH3. This reaction takes place with about 90% fission of the (20-0bond and ahout 10% S-0 breaking. The relative contribution of the two reactions :i not easily controlled and thus this method gave poor precision. Method B.-Reaction of the nitrito complex with methoxide ion (perhaps partially hydroxide ion because of the difficulty in obtaining alcohol in a n absolutely anhydrous state) in absolute methanol results in NO2- whose oxygens have the exact 0'8-composition of the original coordination compound. A small pellet of pure, dry, sodium was placed in the bottom of a Urey tube and the dry nitrito complex placed inside of the hollow bore stopcock. The vessel was evacuated and pure methanol (dried over BaO) was distilled over the sodium and frozen with liquid N2. On warming, methoxide ions were formed while the H2 was removed by evacuation. The Urey tube was tipped t o allow the complex t o drop into t h e reacting solution and after one hour at 50" the mixture was distilled t o dryness under vacuum. The NOz- formed was converted t o SzOby reaction with NaN3 under buffered conditions. Method C.--All known cobalt coordination compounds containing the nitrito group rearrange in the solid state to the nitro form. The dry nitrito compounds were heated a t 50" for 6 hr. to effect this conversion and the resulting nitro compound treated for 2 hr. with a n excess of 0.1 N IjaOH at 50'. Nitrite ion was released and either the hydroxy compound or decomposition products resulted. The mole fraction of 0 1 8 in the NOz- was identical with the average of those in the nitrito complex. Since Nor- exchanges with HzO only at low pH'sIsb the conversion to K20 should be carried out under neutral or alkaline conditions. The experiments with neutral hydrazine produced little NzO(mostly i Y 2 ) and some exchange occurred. Best results were obtained with buffered hTahT3 solution. The exchange of LTOz- with HzO is extremely sensitive to pH while the rate of reaction of N3- with h T 0 2 is less so. I n each case raising the H + concentration increases the rate of reaction. It was shown t h a t NzO could be formed without exchange if the pH was carefully controlled and if regions of high acidity were avoided during addition of reagents. This procedure was carefully followed, the order of addition, pH and the temperature being most im(18) (a) A. Bothner-By and L. Friedman, J. Chem. Phys., 2 0 , 4591 (1952); (b) M .Anbar and H. Taube, THISJOURNAL, 76, 6244 (1954).
lTOl.
7s
portant, but even with great care occasional samples showed anomalous exchange. Two milliliters of a solution containing the NO?- to be analyzed was placed into one side of a two-compartment reaction vessel and into t h e other 3 ml. of sodium phosphate buffer (pH 3.9).1g The vessel was immersed in a CO2(s)-methano1 mixture and about 0.1 g. of solid NaN3 placed on top of the frozen buffer solution. IVhile still frozen, the vessel was evacuated, closed off from the vacuum system, and the S O z- compartment warmed. Immediately on melting of a portion of the KO?- solution, the vessel was tipped so as to pour i t on the frozen NaN3buffer mixture. Rapid mixing was accomplished ana the solution warmed to 40" for 15 min. Isolation of the crude N?O was carried out by freezing the mixture with liquid 5 2 and removing the S2and O2by evacuation. The XZO was transferred to a tube containing NaOH pellets and i t was shaken five minutes to remove COSand H K s and tmnsferred to another vessel for isotopic analysis. I n the conversion of X02- to SSOthere is less than 1% exchange with the solvent. When method B was used to prepare SOa-, a modified procedure was necessary. T o the evacuated vessel containing the dried sample a 1.O M solution of KaN3 was introduced through the stopcock without the admission of air. Following this, the cold buffer solution was added slowly in the same manner with agitation. After 15 min. a t 40' the isolation and purification were carried out in the usual manner. The mole fraction of 0 1 8 present in enriched KaSOp was determined by conversion to N 2 0 and measurement of the mass ratio 46 (44 46). The experimental value agreed reasonably well with t h a t calculated from the relative amounts of normal NaNOI and enriched HzOused in its preparation. I n the case of [Co(NH3)6(0H2)]( ClOd)a, the 018-enrichment was determined by preclpltatlon of the complex as the bromide, removal of the water by heat, equilibration of known amounts of the HzO and COZ,and measurement of the isotopic ratio in the C02.*O The results obtained in this manner agreed with t h a t calculated on the basis of the relative amounts of the starting materials assuming complete exchange. For all other aquo compounds the isotopic ratio was calculated from the known isotopic composition of the starting materials and the relative amounts used. The concentration of 0l8 in the enriched rvater was determined by accurate dilution with normal water, equilibration with C 0 2 and mass spectrometer analysis. The mass 46/(44 46) ratio of the N 2 0 was measured in a Consolidated Nier Isotope mass spectrometer. All samples were compared with a standard K20 sample prepared with normal reagents and were normalized to a value for the standard. of 2.00 X C. Rate of Exchange.-The rate of exchange of SO?with [Co(NH3)6X02](C104)2 was conducted a t 27.0" in 0.1 1V buffer solutions a t p H 5.45 and 6.35. The concentration of the labeled coordination compound (twice recrystallized from methanol-water solution. Anal. Calcd. for [Co(KH3)5K0~](C104)~: N , 21.62. Found: N , 31.52) was 2.5 X M/l. and of normal NaNOz, 5.0 X M/1. The solutions were protected from light and no decomposition was detected by measurement of the extinction coefficient a t 460 mp. The nitro complex was separated by addition of excess KaBr(s) and purified by solution in water and precipitation with NaBr(s) . Recrystallization was necessary since preliminary experiments indicated some N a S 0 2 was carried down with t h e solid. The pure complex was washed with alcohol and acetone and dried in z'acuo. Conversion to NzO was carried out by method C . D . Mutarotation of cis-[Co(en)~(N02)(ONO)]C10~.The preparation and resolution of this compound has been previously described.21 The rate of mutarotation was followed at 19" in pure water while the rate of conversion to the nitro form was carried out spectrophotometrically under the same conditions a t 440 mp.
+
+
111. Results and Discussion Since [ C O ( N H ~ ) ~ Of3,H ~(ROH2), ] exchanges its water with the solvent slowly a t room temperature, (19) T h e buffer was prepared by mixing equal volumes of concentrated Hap01 and H20 a n d neutralizing it with concentrated N a O H until a p H of 3.0 was obtained. Since the ionic strength and t h e N a ' concentration was high t h e p H has only qualitative significance. (20) M . Cohn and H. C . Urey, THISJ O U R N A L , 60, 679 (1938). (21) R.K e n t M u r m a n n , ibid., 77, 5190 (1955).
Oct. 5, 1956
FORMATION AND REARRANGEMENT OF NITRITOCOBALT(IIT) A~IMIKES
the preparation of O1s-enriched [CO(NH~)~ONO] +2, (RONO), from normal HzO, NOZ- and enriched (ROHZ)was thought possible if, during the reaction, the cobalt-oxygen bond is not broken. Preliminary determinations produced a small enrichment in the (RONO) which could not be explained without a partial retention of the cobalt-oxygen bond. Further work showed that exchange with the solvent was occurring during the conversion of (RONO) to NzO. Refined techniques reduced this exchange to a negligible amount permitting exact determinations on the mechanism to be carried out. Table I shows a portion of the results on the conversion of NOz-. Using hydrazine as the reactant a t p H 2 or 8 resulted in appreciable exchange. The high value of RN(HzO)in experiments 1-3 results from the enriched water used in the prep* aration of NOz-. I n later work this water was removed. Experiment 7 was carried out with an enriched, equilibrated NOZ--HZO solution and buffer of approximately the same 0l8 mole fraction. The NzO value in experiment 8 with the * same NOz-, converted in normal water and buffer, when compared with number 7, shows that little exchange occurs in this formation of N20 from NOz-. TABLE I OF NzO FROM NOsFORMATION RN 103
x
RN
x
1 2 3 4 5 6 7 8
8.43 8.43 8.43 4.29 9.65 12.02 13.92 13.92
2.81a 2.81a 2.81a 2.00 2.00 2.00 13.92 2.00
Ex-
RN X 108
io3
N o . ( K o a - ) (HaO)
Method NHaNHnpH2 NHaNHzPH8 Na-, b u f f e r , p H 3 . 8 Na-, b u f f e r , p H 3 . 8 N a - , buffer,pH3.8 N a - , buffer, p H 3.8 Na-, buffer, p H 3.8* N I - , buffer, p H 3.8
changes,
(Nn0) 2.99 3.48 8.40 4.24 9.52 12.08 13.92' 1 3 . 8 8 i 0.12O
%
97
88 0.5 2.0 1.5 -0.6
.o
.4
Solvent from N O z - enrichment not removed. Buffer solution enriched; RN = 13.5. CAverdge of five determinations. a
Several methods of converting (RONO) to NOzwere tried before non-exchange of oxygen was obtained. Table I1 contains these experiments car* * ried out with (RONO) and with salt prepared from * (ROHz), and NOz- in dilute acid solution. The TABLE I1 FORMATION OF r';o2- FROM [ C O ( ru"a)sOXO] (C104)2 RN X 10a
KO. (Hn0)b
9 10 11 12
14.02 1.99 1.99 1.99
Method
R N X 103 (Nz0)
[Co(XH3h(?)Nd)]+2" A 14.00 A 13,91d SaN3, pH 8 9.63 C 13 99 f 0.08"
% Breaking Co-0 Bond
... 98.4 27.1 99.8
[ C O ( N H ~ ) ~ ( ~ N+28O ) ] 14.02 A 7.89 ... 1.99 A 7.74 97.7 100.0 1.99 B 7.88 1.99 C 7 .86c 99.8 1.99 NHzNHz-fiH 8 3.92 32.8 RN X l o 3 = 14.00-14.02. Solvent. CAverage for four determinations. d Highest value obtained. Average value 13.49-8.5% 0-N breaking. e RN X l o 3 = 13.761.99. 13 14 15 16 17
4889
reaction was allowed to proceed for no more than 4 min. a t 5-43' to reduce the amount of exchange of the solvent with the aquo complex. Since tl/, for these exchanges are in the range 5-30 hr. a t 25' very little replacement of 0 I 8 by 0 I 6 occurs. The nitrito complexes were recrystallized in cold normal water, washed with alcohol to remove the nitro compound and dried. Experiments 12, 15, 16 show that methods B and C are good methods of converting RONO to NOz- while experiments 10 and 14 indicate that alkaline substitution is primarily on the cobalt oxygen. Therefore 14-16 * show that the compound made from ROHz and NOZ- retains the Co-0 bond intact. Table I11 confirms these conclusions and extends them to other cations. The values listed are the highest obtained but are probably lower limits. TABLE I11 FORMATION OF R-ON0 FROM ROHZa %
x
103 [Co(NHs)sOHal +a 14.00 (14.06)' 9.86' 13.9qC 1.99b RN
Retention R N X 108 of Method (NzO) Co-0 Bond B 7.96 99.4 B 5.86 98.3 c 7.93 99.4 C 7.97e 99.3 RN % ReX 103 tention (NOeRN of and x 108 co-0 H20) b Method (iYz0) Bond 1.99 C 4.96 92.3 1.99 C 5.28 99.5 93.0 1.99 C 5.22 1.99 B 3.43 79.5 c 5.02 97.5 1.99
R N X 108 (NOz- and HzO) l.9gb 1.99b 1.99* 13.87d ( 1 3 . 9 9 p
RN X 108 (ComCompound plex)C cis-[Co(NHs)4(OHa)e]+8 8.43 cis-[Co(en)l(NOI)(OH,)]+a8 . 6 2 cis-[Co(en)z(NHa)(OHa)]+88 . 9 5 [Cr(NHa)sOHz1+8 5.62 cis- [Co(en)z(OHz)zl+a 8.22
a Reaction time 3 min., temp. 5-6". b Determined by equilibration with COZ. Value calculated from composition of equilibration mixture. d Determined by conversion to NzO. e Aveqage value of three determinations.
The experimental results clearly show, as was concluded by Basolo and Pearson,? substitution on the complex ion oxygen in the nitrosation of some aquo coordination compounds. The identity of the nitrosating agent is not clear but in solutions containing large amounts of NOz- it probably is (NdN62). It is of interest to point out that anhydrous N 6 C i does not react with [Co(NH3)6OHz](C104)3 after one hour a t -5" but this does not preclude the possibility of its participitation in solutions high in C1-. The change in mechanism under slightly alkaline conditions is noteworthy. The formation of the nitro complex is relatively slow and does not involve the nitrito intermediate. The reaction probably proceeds by two paths, the most important being dissociation followed by entry of a nitrite ion while displacement is less favorable. This type of substitution is probably not limited to the formation of nitrito compounds. For, if dissociation and displacement are slow, a labile oxygen containing ion may rapidly replace the labile hydrogen of a complex ion and form the product. The rate of exchange of oxyanions is often quite sensitive to environment and thus (in the nitrito system) raising the @ Hor changing to a less polar solvent leads to a much slower reaction which does not involve the nitrito intermediate. The
reaction of MOO*" or WO4" with cis-[Co(en)Z( H z O ) ~ ]differs + ~ from the NOz- example in that it is rapid even under neutral or slightly alkaline conditions.22a Unlike NOz- both of these oxyions undergo rapid exchange with water in this p H range.22b It is suggested that these ions may also react by siibstitution on the cobalt oxygen. In order to study the conversion of (RONO) to the (RN02) form, i t was necessary to demonstrate the non-exchange of NOz- with (RNOZ). This was carried out a t 27" in the dark by determining * the 018-contentof NOz- and (RNOz) in a solution containing these ions in a molar ratio of 2 to 1. Less than 2% exchange was observed after 50 hr. a t a pH of 5.45 and 6.35. Table IV contains the TABLE IV CONVERSION OF [ C O ( S H ~ ) ~ ( O N O +*)TO ] [Co(NH3)6(S O Z ) ]+* RN x 10' [Co(NH!pON01
Vol. 78
R. KENTMURMANN AND HENRYTAUBE
4890
RN X 108
RN X 10'
(HnO) (NzO)
Exchange,
Conditionso
%
14.06 1.99 p H 6 . 3 , 6 O o , 0 . 5 h r . 0.0 13.99 2.00 p H 5 . 0 , 6 0 ° , 0 . 5 h r . .1 . . 7.59 Solid, 60", 1 hr. .. 1.99 7.60 pH 5.5, 60°, 0.5 hr. .o 1 . 9 9 7.54 pH 6.0, 0.5/1 ratio .7 NOz- to complex, 25", 12 hr. 1.99-11.41 . . 6.70 Solid, 60", 1 hr. .. 1.99-11.41 1.99 6 . 6 5 p H 6 . 0 , 6 0 ° , 0 . 5 h r . 1.0 1.99-1 1 . 41b 1.99 6.69 p H 5.0, 5/1 ratio NO2- 0.2 to complex, 14 hr. a t rm. temp. 0 Carried out in water solution unless otherwise noted. Carried out in the presence of NOz-; R N ( N O ~ - )= 2.00 x 10-3.
1.99-1.99 1.99-1.99 13.19-1.99 13.19-1.99 13.19-1 .9gb
evidence for an intramolecular rearrangement. Neither the cobalt oxygen nor the oxygen attached only to the nitrogen exchanges with the solvent or with added NOz- during the transformation. Preliminary results with cis- [Co(en)z(NOz)(ONO)]+ also show negligible exchange during rearrangement even in the presence of NOz-. Thus, in a t least two cases [ C O ( N H ~ ) ~ O N Oand ] + ~cis[Co(en)z(NOZ)(ONO)]+, the rearrangement is certainly an intramolecular one as concluded by AdellZ3 and by Basolo and Pearson.' Furthermore, optically active cis-[Co(en)z(NOz)(ONO)] + mutarotates a t the same rate as the nitrito(22) (a) R. Kent Murmann, unpublished results: (b) N. F. Hall and 0. R. Alexander, THISJOURNAL, 61, 3455 (1940); G. A. Mills, ibid., 62, 2833 (1940). (23) B. Adell, Swcnsk. Kem. Tids., 66, 318 (1944); 67,260 (1945); Acta Chcm. S c a d . , 6 , 941 (1951).
nitro conversion and ultimately leads to the active dinitro complex without racemization or inversion. 21 It is suggested that if the -OXO group is released, i t must immediately attach its nitrogen before i t leaves the sphere of influence of the complex. This is essentially identical with a heptacoordinated activated state. R6-Co-OXO
--.
I
o
---
N\o;
+ jRs-Co(
j
i + Rs-CONO~
I
_--
The rate of rearrangement of [ C O ( N H ~ ) ~ O N+zO ] in acid solutions has been measured by Ade11.z3 No decomposition but a decreased rate compared to neutral solutions was observed. We have observed the same phenomenon but find that a t the higher acid concentrations (0.07 N ) nitrogen bubbled through the solution removes traces of either oxides of nitrogen or nitrous acid while small quantities of the aquo compound are formed. We interpret this in terms of a n equilibrium involving the aquo complex. Since the rate of water exchange of [ C O ( N H ~ ) ~ O+ 3H ~is] essentially independent of [H+] and [SO4yIz4i t would seem that direct replacement by NOz- is not a probable mechanism. Thus it is suggested that the lowered rate of formation of [ C O ( N H ~ ) ~ N O Z in] +slightly ~ acid solutions is due to an equilibrium which removes a portion of the nitrito complex: the reaction going to completion by a shift of the equilibrium concentrations on depletion of the nitrito complex. H + , (H&) [ R ~ C O O H4~ ]HOSO
[R,Co-OSO]
--f
[RbCoS021
The constitution of the nitrosating agent in the absence of excess NOz- has not been elucidated. The anion catalysis parallels that observed in the aquation of [ C O ( N H ~ ) ~ S O25~and ] + in the exchange of [Cr(H20)6]+3 26 with water which may indicate a strong ion pair effect. We wish to express our appreciation to Howard Baldwin for some of the isotope ratio measurements. STORRS, CONNECTICUT CHICAGO, ILLINOIS (24) H. Taube and A. C. Rutenberg, J . Chem. P h y s , 20, 825 (1952). (25) H. Taube and F. Posey, THISJOURNAL, 76, 1463 (1953) (26) H.Taube and R. A. Plane, J . Phys. Chcm , 66, 33 (1982)
