2092 Inorganic Chemistry, VoL 9,No. 9, 1970
SHIGENOBU FUNAHASHI AND MOTOI-IAKU TANAKA
previous work and therefore are not presented in detail. l 1 HorrocksZ1has presented an equation relating the contact shift to the free energy change when there is a diamagnetic e paramagnetic equilibrium in solution. Assuming that the monoadduct of Ni(dtp), is diamagnetic in the picoline systems, we have tried to fit the temperature dependence of the chemical shift when the experimental free energy change for the reaction is used. While the fit is adequate a t low temperatures (260290°K)) we find a divergence a t higher temperatures (290-340") between the experimental points and the calculated contact shift. l 1 While several hypotheses
are available to explain this behavior, we suggest that the high-temperature behavior might be accounted for by some kind of fluxional behavior. The fluxional nature of five-coordinate geometries is well known.2 3 Perhaps one geometrical structure is diamagnetic but as the temperature is increased, another structure with a triplet ground state becomes preferred. Acknowledgment.-Acknowledgment is made to the donors of the Petroleum Research Fund, administered by the American Chemical Society, for support of this research. (22) E. L. Rluetterties and II. Rosen, I n o i g . Chein., 7, 299 (1968). (2) M. Tanaka, S. Funahashi, and K . Shirai, i b i d . , 7, 57:3 (1968). ( 3 ) S. Funahashi a n d Xi. Tanaka, Bd1. Chena. SOC. J o p . , 43, 703 (1970) 14) S.Funahashi, S. Yamada, a n d M . Tanaka, zbtd., 43, i 6 Y (1970).
+
describes the result on the ligand substitution reactions SiII-EDDA SirI-NTA
+ PAR e XilI-PAIR + EDDX
(1)
+ PAR e SiII-PAR + NTA
( 21
The divalent anions of PAR and EDDA and trivalent anion of NTA are represented as R2-, EDDA2-, and NTA3-, respectively. Experimental Section Reagents. EDDA.-Reagent grade EDDS, obtained from Dojin-do Chemical Co., Kumamoto, Japan, was purified by recrystallization from aqueous methanol. Care was taken to use fresh solutions of EDDA because cyclization of the free ligand occurs.J NTA.-Reagent grade KTA, obtained from Dojin-do Chemical Co., was recrystallized twice from distilled water. Sodium Perchlorate.-Sodium perchlorate was prepared by dissolution of sodium carbonate in perchloric acid. Heavymetal impurities in sodium perchlorate were precipitated as hydroxides a t pH 9 and extracted as P A S complexes (I'XN: .
(,5) K. B. LeBlanc, A r i d . Cheriz., 31, 18-10(105'3).
LIGAND SUBSTITUT~ON REACTIONS OF Ni(1I) COMPLEXES 1-(2-pyridylazo)-2-naphthol)with purified carbon tetrachloride four times a t pH 9 and as oxinates with carbon tetrachloride four times at pH 7 . Sodium perchlorate was then recrystallized once from distilled water. Methods of preparation of the other reagents (sodium hydroxide, nickel perchlorate, boric acid, borax, and PAR) have been described previously .6 Kinetic Measurements.-The kinetics of the ligand substitution reaction between the nickel complex and PAR was studied 5pectrophotometrically a t 495 nm using a photoelectric spectrorneter, Type 6 (Hirama Rikakenkyujo, Kawasaki, Japan), equipped with a thermostated 2.0-cm cell compartment. A Coolnics circulator, Type CTE-1B (Komatsu Solidate Co., Komatsu, Japan), was utilized in circulating thermostated water through the block. The temperature in the cell was maintained a t 25 =k 0.1”. Supplementary spectra were taken on a JASCO Model ORD/UV-5 optical rotatory dispersion recorder. The procedure for a typical kinetic run is outlined as follows. The nickel complex solution was prepared by mixing nickel perchlorate solution with EDDA solution or NTA solution. Borate buffer solution was added to adjust pH of the solution. The ionic strength of the solution was held constant ( p = 0.10) by sodium perchlorate. A quartz cell with light path length of 2.0 cm containing the nickel complex solution was placed in a thermostated compartment of a spectrophotometer. The PAR solution was brought to temperature equilibrium in a thermostated bath. The reaction was started by mixing the nickel complex solution and the PAR solution. The transmittance a t 495 nm of the reaction solution was recorded as a function of the reaction time by a recorder (an X-Y recorder, Type VR-631; Matsushita Communication Industrial Co., Yokohama, Japan). All pH values were measured with a Radiometer PHM-4d (Copenhagen) with a Type G202B glass electrode and a Type K401 calomel electrode which was carefully calibrated with a standard buffer solution (potassium dihydrogen phosphate, 0.025 M ; disodium hydrogen phosphate, 0.025 M ) prepared as described by Bates?
Results Ligand Substitution Reaction of the Ni’I-EDDA Complex with PAR.-Under the present experimental conditions, the Ni“-PAR complex forms the 1:2 complex, NiR22-,6the NiII-EDDA complex is NiEDDAJsand the dominant species of PAR is the monoionic species HR-, as evident from the stability constants of PAR.Q The molar absorption coefficients of NiR22- and HR- a t 495 nm are 7.85 X l o 4 and 1.11 X lo8, respectively. Reaction 1 is much favored t o the right so that the reaction of the Ni-EDDA complex with PAR goes to completion and the reverse reaction can be neglected in the kinetic study. The overall ligand substitution reaction is described as NiEDDA
+ 2HR- -3 NiRz2- + EDDA’
(31
where EDDA’ refers to EDDA not combined with nickel. PAR is in a sufficient excess over the NiEDDA complex, so that the reaction can be regarded as pseudo first order in NiEDDA. Plots of log ( E , E,) os. t are not linear until the later stage of the reaction (E is absorbance of the reaction system). Figure 1 (6) S.Funahashi and M. Tanaka, Inorg. Chem., 8,2159 (1969). (7) R. G. Bates, ”Determination of pH, Theory and Practice,” Wiley, New York, N. Y., 1964. (8) S. Chaberek, Jr., and A. E. Martell, J. Ameu. Chem. Soc., 74, 6228 (1952). (9) W. J. Geary, G. Nickless, and F. H. Pollard, A m l . Chim. Acta, 27, 71 (1962)
Inorganic Chemistry, Vol, 9, No. 9,1970 2093
600
I200
1800
t , sec Figure 1.-Typical first-order plot of the slower step using the final absorbance (E,) and the observed absorbance ( E t ) .
represents a first-order plot of the data. The points are taken from a continuous trace, and excellent linear plots were obtained for all points taken after about 200 sec. Extrapolation to zero reaction time gives a higher absorbance, Ei,than that of the reactants and indicates the rapid formation of an intermediate. T h e first part of the trace also gives a linear first-order plot as seen in Figure 2, the Ei value taken from the interr
10
20
30
t , sec Figure 2 -First-order plot of the faster step in Figure 1 using the intercept of Figure 1 to give E,.
cept of Figure 1 being assumed to correspond t o the absorbance a t equilibrium of the first part of the reaction. This fact suggests a consecutive reaction. We assume the reaction scheme NiEDDA
ki + HR- e RNiEDDA’ + H + ko,a(H)
RNiEDDA’
+HR-
NiRz2-
+ EDDA’
(4 1
(5)
where RNiEDDA’ indicates all species of intermediates which will be discussed later. A series of kinetic runs was performed a t various hydrogen ion concentrations in order to determine the effect of hydrogen ion. In Table I are given the conditional rate constant k o , l ( ~ ) for the fast reaction and the conditional rate constant k 0 , 2 ( ~for ) the subsequent slow reaction obtained under various experimental conditions. According to these results the conditional rate constant k c , l ( ~is) proportional to the concentration of PAR and independent of the other concentration terms, such as EDDA, hydrogen ion, and Ni-EDDA complex (runs no. 1-15 in Table I). Therefore, the rate equation for the fast reaction, that is, for the formation of an intermediate RNiEDDA’, is
2094 lnorgiinic Claemktry, Vol. 9,No. 9, 1970
SIIIGENOUU ~ k J N A H A S 1 1 1AND hfOTOIIAKU 'I'ANAKA
TABLE I THE
LIGAND SUBSTITUTION REACTION OF
THE ~ICKEL(I1)-~rHYLENEDIADlINE-N,S'-DIACETATE
COMPLEX JVITH 4-(2-PYRIDYLAZO)RESORCINOLn
Run no.
1
2 3 4
5 6
"
8*
9 10 11 12 13
I4
C H , .If
CEDDA, 31
5,60 X 5.60 X 5,60 x 5.80 x L12 x 1.28 x 1,92 x 6.40 X 6.40 X 6.40 X 6.40 x 6.40 x 6.40 x 6.40 X 6.40 x 6.40 X 6.40 X 5.60 x 5.60 x 5.60 X 5 . 60 X 5.60 X 5.60 X 5.60 X
15 16 17 18 19 20 21 22 23 34 a CS, = 5.09 X 10-6 mined
10-6 10-0 10-6 10-3
10-fi 10-5
10-5 10-j 10-j 10-2 10F 10-5
M; fi
=
PH
8 . 0 0 X 10-2 9.21 8 . 0 0 x 10-5 9.45 8 . 0 0 x 10-5 9.61 8 . 00 X 10-6 9.73 3.99 x lo-^ 9.04 3.99 x 10-2 9,04 3 . 9 9 x 10-5 9,04 3 . 9 9 x 10-6 9.04 2.39 x 10-u 9.04 3.99 x 10-5 9.03 9.41 3.99 x 10-3 5.98 X IO-; 9.03 9.97 X 10-6 9.02 9.23 9.97 x 10-5 9.42 9 , 9 7 x 10-j 9 . 9 7 x 10-5 7.80 9 . 9 7 X IO-: 8.09 2.00 x 10-4 8.42 2 . 0 0 x 10-4 8.59 2.00 x 10-4 8.82 2.00 x 10-4 9.21 2 . 0 0 x 10-4 9.30 2 . 0 0 x 10-1 9.46 2.00 x 10-4 9.74 0.10 (NaC104); borate buffer: [B]
T h e conditional rate constant k 0 , 2 ( H ) , on the other hand, increases with decreasing pH and is independent of the other concentration terms such as PAR, EDDA, and Ni-EDDA complex. The plots of k 0 , 2 ( ~v )s . [ H + ] did not fall on a straight line but yielded a curve. It seems thus reasonable to assume that the increase in rate constant k0,2(H)with decreasing pH is attributed to the formation of the RNiHEDDA- complex, the protonated form of the intermediate RNiEDDA2-. Then the kinetic equation is expressed by
ka,i(R), sec-l
8.48 X 7.84 X 7.60 X 8.16 X 4.83 X 4.15 X 3.75 x 4.87 x 2.69 X 4.80 X 1.59 x 6.58 X 1.01 x 8.89 x 9.03 x
lo-* 10+
10-2 10-2
10-2 10-1 10-2 10-2
ki, &If-1
sec-1
1.06 x 0.98 X 0.95 x 1.02 x 1.21 x 1.04 X 0.94 X 1.22 x 1.12 x 1.20 x 1.15 x 1.10 x 1.01 x 0.89 x 0.91 x
108 lo3 103 108 108 10" IOJ 108 108 108 108 10" 108 10J 108
... ...
. . .c ...
...
I
.
/iu,?(H). S e C '
.
, . .
...
, . .
=
...
...
...
...
... ...
... ...
0.008 db; 25.0 =t0.1".
Cxi = 3.05 X
5.22 x 1 0 - 4 3.88 X 3 . 7 2 x 10-4 3 . 5 3 x 10-4 5.53 x lo-,' 4 . 9 5 x 10-4 4 . 5 7 x lo-' 5 4 8 X lo-* 4 . 4 8 x 10-4 5.37 x 10-4 3.76 X 5,83 X 6 . 2 9 x 10-4 4 . 9 1 x 10-4 3 . 3 0 x 10-4 3.09 x 10-8 2.32 X 1.72 x 10-8 1 . 2 8 x 10-3 9.48 x 10-4 5.57 x 10-4 4.67 X lo-* 4 . 4 5 x 10-4 3.49 x 10-4 df. riot deter-
Therefore, we can determine the stability coiistant and rate constants by the curve-fitting method. The plot of k 0 , 2 ( ~against , pH is compared with a normalized x) = f(1og x).~ The plot fits well function y = p / ( l one of a family of the normalized curves (Figure 3 ) .
+
351
d[NiRz2-1 ____ = kz[RSiEDDA42-]f ~ ~ . H [ R S ~ H E D I ) ~( 7 \ -) ] dt
From the equilibrium relationship KHILsIj+EUD,\
[ R Xi H E D I)A -1 -___ -[ R X E D D 4 * -1 [ H -I- J
= --
~
(8)
and mass balance, we have [RNiEDDA']
+ [RNiHEDDA-]
=
[RSiEDDA2-]
=
[RP3EDDX2-](l
+ K H ~ ~ i ~ c ~ u . i [ H(9)+ 1 )
1 7.5
ao
a5
9.0
9.5
PH
Combining eq '7, 8. and 9, we obtain
Figure 3.--I-'lots of k a , z c ~ os. , pH. The line is the theoretical curve calculated with the values obtained by the curve fitting. The points are experimental a t 25.0'. T n o dotted lines show the sec-I). range of deviation of k2.a (k2,H = (5.1 J 0.4) x =
kO,Y(H) [RSiEDDA']
(10)
The rate constants a t p = 0.10 and 25.0" are k l = (1 0 (kz ~ ~ , H K ~ B ~ ~ H E D L ) A [ H ' I )0.3) ( ~ X i o 3 Jf-l sec--', k z = (2.7 f 0 . 2 ) X sec-I, and k 2 , ~= (5.1 =k 0.4) X lop3 sec-'. The K H ~[H+])-' ~ is i a conditional ~ ~ first-order ~ ~ rate ~ is 108O 2 *o.lo a t p = 0.10 stability constant KHRN,aEl,,,A constant composed of rate constants, stability constant, and 25.0". T h e uncertainty of constants indicates the and hydrogen ion concentration. The conditional range. rate constant k O , Z ( H )can be rewritten as Determination of the Stability Constant of the Intermediate Ni"-EDDA-PAR Complex.-\Then the
where
~ u , z ( H )=
+
+ *
Inorganic Chemistry, Vol. 9, No. 9, 1970 2095
LIGAND SUBSTITUTION REACTIONS OF Ni(I1) COMPLEXES equilibrium expressed by eq 4 is established, we have the relationships
+ [RNiEDDAZ-] + [RNiHEDDA-] + [RNiEDDA2-] + [RNiHEDDA-] Ei = CHR[HR-]+ ~ R N ~ ~ u u A [ R T \ ; ~ E D +D A ~ - ] CNi
=
CR
[NiEDDA]
(12)
[HR-]
(13)
=
Ligand Substitution Reaction of the NiII-NTA Complex with PAR.-The overall ligand substitution reaction for the NiII-NTA system is described as NiNTA-
+ 2HR- +NiRz2- + NTA'
(15)
where NTA' refers t o NTA not combined with nickel A t the beginning of the reaction, any jump in absorbance a t 495 nm (Ei) observed in the NiII-EDDA system did not occur. The substitution reaction is of the first order in NiII-NTA and PAR, respectively. The second-order plots were linear a t least for over 90% of the reaction. Then the second-order conditional rate constant K o ( H ) was determined from the slope of the straight line. Some conditional rate constants a t various concentrations of NTA and a t various pH's are given in Table 111.
~ R N ~ E E D D ~ R N ~ H E D(14) D A - ] ion.
where CNi and CR represent the total concentrations of nickel and PAR, respectively; E refers to the molar absorption coefficient of relevant species; Ei is the absorbance a t equilibrium of the reaction system which is estimated by the extrapolation to zero reaction time in Figure 1. The Ei value increases with increasing pH as apparent from equilibrium 4. I n the present experimental condition, the E i value is constant a t p H higher than 9.3 where the intermediate is formed quantitatively. Thus we have the molar absorption coefficient of RNiEDDA2-; ~ R X ~ E D D Ais 3.83 X lo4 a t 495 nm. Protonation of the intermediate RNiEDDA2- is considered to occur a t the most basic site of RNiEDDA, i e . , a t one of secondary amines in the coordinated EDDA (see Figure 7). So the molar absorption coefficient of RNiEDDA2- does not seem to differ appreciably from that of RNiHEDDA-, the value of E R N ~ H E D D A is assumed to be the same as ERNiEDDA. Then combining eq 13 and 14, we obtain [RNiEDDA2-]
+ [RNiHEDDA-]
=
E, - ~HRCR - eHR
€RNiEDDA
From eq 12 and 13, we have the concentrations of NiEDDA and PAR. Therefore we can calculate the conditional equilibrium constant K i with knowledge of the concentration of hydrogen ion K, =
([RNiEDDAz-] f [RNiHEDDA-])[H+] [KiEDDA][HR-]
T A B L E 111 THEL ~ G A NSUBSTITUTION D REACTION OF THE KICKEL(II)-SITRILOTRIACETATE COMPLEX WITH
4- (2-PYRIDYLAZO)RESORCIXOLa
5.40 5.40 5.40 5.40 5.40 5.40
X X X X X X
pH
M-1
sec-1
Ei
PH
9.97 9.97 9.97 3.99 3.99
0.332 0.269 0.244 0.143 0.182
9.02 8.09 7.80 8.03 8.54
u C K i = 5.09 X 10-6 M; CEDUA= 6.40 (NaC104); 25'.
Log K ~ R N E D D
X
8.3 8.1 8.0 8.3 8.3 Av 8 . 2 3 0~. 2 10-6 M i c./ = 0.1
J - 1
9.12 8.07 8.28 8.72 9.11 9.12
M;
p
3.78 2.45 2.87 3.38 3.53 3.52 = 0.10
i 0.1".
The conditional rate constant kO(H) is independent of the concentration of NTA and increases with increasing pH. We assume the following reaction scheme on the basis of the results for the NiII-EDDA system
RNiNTA8-
l C a C R , A
