The electrochemical behavior and structure of silver

0 downloads 0 Views 536KB Size Report
21. Tetracthylammonium tctrafluoroborate (TEAFB), prepared from tct- rafluoroboric acid and tetraethylammonium bromide (22). was re- crystallized from 95% ...
The electrochemical behavior and structure of silver salts of imides and carbo~amides"~

Can. J. Chem. Downloaded from www.nrcresearchpress.com by 192.30.84.250 on 10/04/16 For personal use only.

JEAN-YVESH U O T , DEN[S ~ SERVE,^ AND JEANLESSARD' Dkparternent de clzirnie, Universiti de Si~erbrooke,Sherbrookr (Q11i.), Catzadn J I K 2RI Received September 30, 1982 Can. J. Chem. 61, 1890 (1983). JEAN-YVES HUOT,DENISSERVE,and JEANLESSARD. Silver salts of amides (imides and carboxamides), containing silver cations and amidc anions in a one-to-one ratio, have been studied by electrochemical, potentiometric, and spectroscopic means. They have the structure [Amide-Ag-AmideIAg with a linear nitrogen-silver-nitrogen bond and a silver cation loosely bound to the oxygens of the amide moieties. For instance, ( i ) the electrochemical data show that half of the silver ions are tightly bound (morc difficult to reduce); (ii) only the loosely bound half can be titrated with chloride ions; (iii) according to 'H and "C nlnr spcctroscopy, the silver salt of succinimide ( I ) is symmetric. The selective reduction of the loosely bound silver ions in acetonitrile containing tetraethylammonium tetrafluoroborate consumes 0.5 F/mol and leads to the mixed salt [Amide-Ag-AmidcIEtrN; further reduction at a more negative potential leads to the formation of the tetraethylammonium salt, EtNAmide. 'The latter two salts derived from succinimide have been prepared by a chemical route. The silver, mixed silver-tetracthylammonium, and tetraethylammonium salts are electroactive in oxidation. Can. J. Chem. 61, 1890 (1983). JEAN-YVES HUOT,DENIS SERVEet JEANLESSARD. Nous avons Ctudii, par des mCthodes Clectrochimiques, potentiometriques et spectroscopiques, quelques sels d'argent d'amides (imides et carboxamides) formCs par une mtlange Cquimolaire de cations argent et d'anions d'amides. Les rCsultats permettent de diduire la structure [Amide-Ag-AmideIAg comportant une liaison azote-argent-azote lintaire et un cation argent faiblement liC aux oxygbnes des groupements amides. Par exemple, (i) les etudes Clectrochimiques montrent que la moitiC des ions argcnt sont fortement lies (plus difficle i rCduire); (ii) uniquement lcs cations argent faiblement lies peuvent &tretitrCs par les ions chlorures; (iii) selon les spectres rmn de proton et du carbone-13, le sel d'argent du succinimide ( I ) est symCtrique. La rkduction selective des ions argent faiblement lies, dans I'acCtonitrile contenant du tetrafluoroborate de tCtraCthylammonium, donne le sel mixte [Amide-Ag-Amide]Et,N; une riduction plus poussie a un potentiel plus nCgatif conduit au sel de tCtratthylammonium, EtNAmide. Ces deux derniers sels ont it6 prepares par voie chimique dans le cas du succinimide. Les sels d'argent, les sels mixtes d'argent et de tCtraCthylammonium et les sels de tCtraCthylammonium sont tlectroactifs en oxydation.

Introduction Our interest in the oxidation of imide and amide anions in a nonaqueous medium led us to prepare their silver salts which are more soluble in acetonitrile than the alkali metal salts. Furthermore, the silver salts should be readily converted to tetraalkylammonium salts by deposition of silver on a cathode in a medium containing a tetraalkylammonium salt as supporting electrolyte. The silver salt of formanilide was prepared by Comstock and Kleeberg ( I) in 1890 and those of benzamide, proprionamide, and acetamide by Titherley (2) in 1897. Conductometric (3) and electrochemical (4) studies of the silver salt of acetamide in melted acetamide led to the identification of a soluble 2 : 1 (amide anion - silver ion) complex, [(CH,CONH)2Ag]Na, and an insoluble 1 : 1 complex, CH,CONHAg. From studies on the site of alkylation of ambident anions, Stein and Tan (5) concluded that, in the silver salts of carboxamides, the silver ion ( a soft acid) is bound to the nitrogen (the softer site of the amide anion). Thermodynamic analysis of the silver salt of succinimide in water revealed the existence of a 1 : 1 and a 2 : 1 complex (6); 'Abbreviated from the M.Sc. Dissertation of J. Y. Huot, UniversitC de Sherbrooke, Sherbrooke, QuCbec. 1980. 'Presented in part at the 64th Canadian Chemical Conference, Halifax, Nova Scotia. June, 198l . 'NSERC of Canada (1978- 1980) and "Ministire de I'Education du QuCbec" (1978- 1980) predoctorate fellow. 4~rance-Quebec visiting Scientist (Sept.-Dec. 1977). Present address: Laboratoire de photochimie, UniversitC scientifique et medicale de Grenoble, 38041 Grenoble Cedex, France. ' ~ u t h o rto whom correspondence should be addressed.

the small contribution from entropy would be characteristic of a soft-soft (nitrogen-silver ion) interaction. According to potentiometric and thermodynamic data on silver complexes of uracil, the monoanion forms a 2 : 1 complex and the dianion, a I : I complex (7). In the dianion salt, one silver ion would be attached to the more basic nitrogen and the other to the neighboring carbonyl group (7). Recently, Guay and Beauchamp (8) have reported the X-Ray crystallographic structure of the silver salt of I-methyl thymine: two imide anions are bound to one silver ion through the nitrogens and in a linear fashion." This paper deals with the structure of some silver salts of iinides and carboxamides determined from electrochemical, potentiometric, and spectroscopic data.

Results and discussion The stoichiometry of the silver salts of irnides and carboxamides corresponds to a one-to-one ratio of arnide anion and silver cation according to the method of preparation and to the analytical data (see the Experimental). The electrochemical behavior of the silver salt of succinimide, AgSucc ( I ) , on a platinum electrode and in acetonitrile containing tetraethylammonium tetrafluoroborate (TEAFB) as supporting electrolyte, is summarized in Table 1 and is representative of the behavior of the other silver salts. The linear voltammogram of 1 (Fig. 1, A ) shows three diffusion controlled waves having the same amplitude ( I , / C ) :one 6 ~ h X-ray e structure of gold(1) salt of N-methyl hydantoin also shows a nitrogen-gold(1)-nitrogen linear bond with two amide anions (9). A 2: 1 structure has been accepted sincc 1940 for the mercuric complexes of succinimide and acetamide (10). Likc the Agt ion, the Aut and ~ g "ions have a d'' configuration.

HUOT ET AL

1891

TABLEI. Elcctrochcrnical" and ultraviolct spcctsoscopic data for AgSucc (1) in acctonitrilc-TEAFB (0.1 M ) Lincar voltammctry" E Solution

(v)

II/C< (pA/mM)

After reduction at -0.50 V (0.49 F/rnol)

-0.18 - 1.30 ( + 1.48) - I .35 ( + 1.48)

7 6 7 6 7

1.2 0.5 0.3 0.3 0.3

Aftcr further rcduction at - 1.50 V (1.01 F/rnol)

( + 1.90)' ( + 1.69)'

13 13

0.7 I .O

Initial

Can. J. Chem. Downloaded from www.nrcresearchpress.com by 192.30.84.250 on 10/04/16 For personal use only.

I,~"

Cyclic voltnrnrnctry" E ,,'I

(v)

It,/c (pA/mM)

-0.25 -1.50 (0.00)" - 1.55 (-0.29)" (0.00)"

9 4 Vcry largc 5 3 Very large

at1

Elcctrodcposition of ~ g " E,,,,,," (V)

TI.,,,, (F/mol)

Yield"

233(4500)

-0.50

0.49

54

233(4500)

- 1.50

0.52

51

~,,,,,x(c)' (nm)

(%I

227( 1 1 000)

"The electrochen~icalanalyses were carried out on a Pt electrode using a 3.35 1nM solution. "On a rotating disc clcctrode ( w = 600 rpm). The anodic waves are given between parentheses. 'Range: 0.00 to -2.00 V . The anodic peaks are indicated between parentheses. "E VS. A ~ o / A ~0.01 - M. "The limiting current is diffusion controlled. 'For the uv spectra, 0.5 mL of the electrochen~icalsolution was diluted to 5 mL with dry acetonitrile. The extinction coefficients ( L mol-I cm-I) are based on the concentration of the succinitnide moieties (anions). 'Calculated from the weight of Ag" deposited on the cathode and based on the total amount of silver ions initially present in 1. "Oxidation of electrodeposited Ago. 'Forward scan. 'Backward scan.

t

A

f

-2,OOV

-1.50

- 1.00

-0.50

0.00

anodic wave and two cathoclic. waves.' The corresponding cathodic peaks of the cyclic voltammogram are shown in Fig. 2, A; the large anodic peak at 0 V is due to the oxidation of Ago deposited on the electrode during the cathodic sweep. Upon reduction at -0.50 V , 50% of the silver present in 1 is deposited. The first cathodic wave disappears while the second cathodic wave and the anodic wave remain unchanged (Fig. 1 , B). The cyclic voltammogram (Fig. 2, B) shows the second cathodic peak only and a new anodic peak at -0.29 V. There is no change in the uv spectrum (Table I). Further reduction at - 1.50 V leads to the deposition of the remaining 50% of the silver initially present in 1. In the uv spectrum, there is a small bathochromic shift and an hyperchromic effect. The linear voltammogram (Fig. I , C) shows the anodic wave only, which now appears at a more anodic potential; its amplititude has increased by a factor of two (Table I). The above results can only be accounted for if salt 1 has the I Awhich ~ there are two different structure [ S U C C - A ~ - S U C Cin types of silver ions in a one-to-one ratio. The silver ion tightly bound to two succinimide anions should be more difficult to reduce than the other. Therefore, the first cathodic wave ( E l l z = -0.2 V) corresponds to proccss [ ! and thi: secc)nil cathodic 121. wave ( E , / ? = -1.15 V?.to pi-.:.3:;

lb

0.00

+0.50

+

1.00

+1.50

+

2.00V

>

FIG. I. Linear voltarnrnetry of [Succ-Ag-SuccIAg (1) in acetonitrile-TEAFB (0. I M) on a Pt electrode (w = 600 rprn, sweep rate = 16 mV/s): A, before reduction; B, after reduction at -0.50 V; C, after reduction at - 1.50 V; D, residual current.

7The fact that the a t 1 value for the first cathodic wave is larger than I .O can probably be ascribed to the modification of the nature and surface of the electrode with the deposition of Ago. The same phenomenon was observed when reducing silver nitrate on platinum in acetonitrile-TEAFB (0.1 M ) (see Table 3, last entry). The a n values have been obtained by plotting log [(i, - i)/i] versus potential according to the relationship given in ref. 1 1 for irreversible processes.

C A N . J . CHEM. VOL. 61. 1983

1892

A

Can. J. Chem. Downloaded from www.nrcresearchpress.com by 192.30.84.250 on 10/04/16 For personal use only.

f

for l b ) when based on the concentration of Succ-Ag-SLKC anions. Further confirmation of structure [Succ-Ag-SuccIAg for 1 is provided by the following potentiometric and amperometric titrations. First, as shown in Table 3, only half of the silver ions present in 1, the loosely bound ones, can be titrated with tetraethylanin~oniumchloride (eq. 131). The protonolysis of 1 with nitric acid (eq. 141) liberates the tightly bound silver ions

which can then be titrated with chloride ions (Table 3). Secondly, the potention~etrictitration of silver nitrate with the sodium salt of succinimide in 50% ethanol gives one end point at 2.0 equivalents (Fig. 3) and is thus described by reaction [5]. Thirdly, the amperometric titration of the tetraethylammonium salt of succinimide ( l b ) with silver nitrate (Fig. 4) shows clearly that the mixed silver-tetraethylammonium salt l a , where only the tightly bound silver ions are present, is formed first (eq. (61). lndeed when about 0.4 equivalent of silver nitrate has been added, the cyclic voltammogram shows a [6]

2Et,NSucc lb

171

+ AgNO, + [Succ-Ag-Succ]Et,N + Et,NNO, 1(1

[SUCC-Ag-Succ]Et,N+ AgNO, + [Succ-Ag-SuccIAg 1(1

+ Et,NNO,

1

FIG.2. Cyclic voltammetry of [Succ-Ag-SuccIAg (1) in acetonitrile-TEABF (0.1 M) on a Pt electrode (swcep rate = 200 mV/s): A, before reduction; B, after reduction at -0.50 V. The mixed silver tetraethylammonium salt l a and the tetraethylammonium salt l b have been prepared chemically (see the Experimental) and their electrochemical and uv spectroscopic behavior were identical to those of the species obtained from the electrochemical reduction of AgSucc (1) after passing 0.49 F/mol at -0.50 V, then 0.52 F/mol at -1.50 V, respectively. The mixed salt l a is oxidized at the same potential and gives a uv spectrum practically identical to that of the silver salt 1 in agreement with the conclusion that the same species, the Succ-Ag-Succ anion, is involved in both salt 1 and salt l a . Reduction of l a at -1.50 V liberates two succinimide anions (eq. [2]), which accounts for the two-fold increase of the amplitude of the anodic wave.' The succinimide anion (see 1b ) is more difficult to oxidize ( E = + I .90 V) and absorbs at a longer wavelength (A,,,, = 227 nm) than the Succ-Ag-Succ anion (see 1 or l a , E , / ? = + 1.48 V, A,,,,, = 223 nm). The extinction coefficient of 4500 L mol-' cm-' reported for 1 and l a (Table 1) is based on the concentration of succinimide anions; it becomes 9000 (as compared to I1 000

he

n value for the oxidation of the Succ-Ag-Succ

anion is 1 1 . 2 and that for the oxidation of the succinimide anion is 0.97 (see footqote 1 I).

cathodic peak at - 1.45 V, an anodic peak at -0.20 V, and the large anodic peak at 0 V; thus, this voltammogram corresponds to the cyclic voltammogram of l a (see Fig. 2, B). As more silver nitrate is added, the cathodic peak at -0.20 V due to the reduction of the loosely bound silver ions in 1 first appears, then increases in intensity,' while the intensity of the cathodic peak at - 1.45 V remains more or less constant (Fig. 4). The anodic peak at -0.30 V decreases in intensity and disappears when l a has been completely converted to 1 (eq. [7]). Finally, the amperometric titration of the silver salt 1 with the tetraethylammonium salt 1b followed by cyclic voltammetry (Fig. 5) shows clearly the formation of the mixed silvertetraethylammonium salt l a , according to eq. 181. As salt 1 is converted to salt l a , the amount of loosely bound silver ions

(limiting current of the cathodic peak at -0.20 V) decreases while that of the tightly bound silver ions (limiting current of the cathodic peak at - 1.45 V) increases. Surprisingly, when the loosely bound silver ions (salt 1) have completely disappeared, further addition of succinimide anions leads to a steeper increase of the limiting current at - 1.45 V; a plateau is reached when about 2 equivalents of l b have been added, as if the effective concentration of tightly bound silver ions (of l a ) was 'Since the peak potentials for the reduction of AgNO, and the reduction of the loosely bound silver ion of 1 are quite close, a single peak at ca. -0.2 V is observed when the two species are present, that is, when there is an excess of AgNO,.

HUOT ET A L

TABLE2 . ' H and "C nuclc;lr magnetic resonance and infrared data for succinimide and salts 1 , ltr, and 111

"C nmr"

' H nmr" Compound

Can. J. Chem. Downloaded from www.nrcresearchpress.com by 192.30.84.250 on 10/04/16 For personal use only.

HSucc Et,NSucc( 111)" 1Succ-Ag-Succ]Et,N ( 1 (1 )" [Succ-Ag-SucclAg (1)

CH, (PP~)

CH2 (ppnl)

C=O (ppm)

Infraredc v C=O (cm ' )

2.58 2.42 2.59' 2.55

30.5 32.9 32.4 32.4

181.6 198.3 194.5 194.7

1770, 1725 1650. 1600 1640 1650

"In CD,CN-D20 (9: I ) with TMS as reference. "In CH,CN-D,O (8:2) using CH,CN as reference. 'In dry CHCN. "In the "C nmr spectrum. the carbons of the tetraethylammonium cation iibsorb at 7.7 (CHI) and 53.1 pni (CH?). 'In CD,CN-D,O (X:?).

T A B L E3. Elcctrochcmical" and potcntion~ctric"data for the silver salts of imides and carboxamides

Linear voltammetry'

S ~ l v e rsalts of

Concentration' (mM)

El/2/ (v)

II/C (k*.A/mM)

Potentiomctric titration of Ag' ion

C y c l ~ cvoltametry"

a11

El,' (v)

[I>/

c

(kA/mM) -

(%I

+H N O ~ " (%)

-

9

Succinimidc ( I )

4 Very large

5 4 Very large 8

Tetramethylsuccinimide (3)

8 Very large

Phthalimide (4)

10 8

2 Very large 8 5 Very large

Benzoylimide (5)

-

Formanilide ( 6 )

Very large Benzamide (7) AgOCOCH3 AgNO,

"The electrochemical analyses were carried out on a Pt electrode in acetonitrile-TEABF (0. I M ) . "The potentiometric titrations were carried out in acetonitrile with tetraethylamnioniuru chloride using a silver wire vs. Ag/Agt 0.01 M. 'On a rotating disc electrode ( w = 600 rpm except in the case of salt 7 where w = 4000 rpm). The anodic waves are given between parentheses. "Range: 0.00 to -2.00 V. The anodic peaks are indicated between parentheses. "The concentration is based on the formula (Ag Amide).,rH20 deternlined from tile analytical data (see Experimental). 'E vs. Ag/Ag' 0.01 M . "fter electrodeposition of all the silver ions, the amplitude of the nrlodic \tla>teincreases.by a hctor two and its half-wave potential changes with the direction of the scan; for 1, see Table I; for 2, reduction at - 1.60 V , E l / ? = 1.7 1 V (forward) and + 1.64 V (backward); for 3, reduction at - 1.60 V , E l l 2 = + 1.16 V (forward) and + 1.46 V (backward); for 4, reduction at - 1.50 V , El,, = +2.02 V (forward) and + 1.50 V (backward); for 5, reduction at -0.90 V, E l , ? = +0.93 V (no change with the direction of scanning); for 6 , reduction at - 1.50 V. El!? = +0.16 V (no change with the direction of scanning). "One equivalent of concentrated nitric acid with respect to AgSucc (1) was added. 'Ill-defined wave (see Fig. 6). 'The amount of salt added could not be solubilized completely. 'In the presence of I equiv. of succinimide (HSucc).

+

CAN. J . CHEM. VOL. 61, 1983

t

1.0

2.0

Can. J. Chem. Downloaded from www.nrcresearchpress.com by 192.30.84.250 on 10/04/16 For personal use only.

NaSucc (equivalents)

FIG. 3. Potentiometric titration of silver nitrate by the sodium salt of succinimide in 50% ethanol.

Et4NSucc ( equivalents) A ~ N O(equivalents) ~

r ~EtNSucc ) FIG. 4. Amperometric titration (cyclic ~ 0 ~ t X m n e tof .70 in ( l b ) (prepared by reduction of AgSucc (1) (3.40 mM) at acetonitrile-TEAFB (0.1 MI) by silver nitrate: A, current of the cathodic peak at - 1.48 V; m, current of the cathodic peak at -0.20 V; O,current of the anodic peak at -0.30 V .

twice as large in the presence of a 100% excess of succinimide anions. We have no satisfactory explanation for that phenomenon. in the cyclic The origin of the anodic peak at -o.30 voltammogram of the mixed salt l a (Fig. 2, B) can be ascribed to the oxidation of electrodeposited silver to a tightly bound silver ion in the presence of succinimide anions at the surface of the electrode (eq. [9]). Succinimide anions are liberated by [9]

A ~ '+ 2Succ-

' BF4-1101 ~ g +

0

-0.30 V

v

[Succ-Ag-Succl-

Ag+BF4-

+ e-

+ e-

F 1 c . 5 . Amperometric titration (cyclic voltammetry) of [Succ-Ag-SuccIAg (1) (4.39 m M ) by Et4NSucc ( l b ) (the equivalents of l b are calculated using the formula AgSucc for I ) in acetonitrile-TEAFB (0.1 M): A, current of the cathodic peak at ,45 V; m, current of the cathodic peak at -0.20 V; a, current of the anodic peak at -0.30 V; 0 , current of the anodic peak at 0 V .

the reduction of tightly bound silver ions (reduction of l a , eq. [2]) during the cathodic sweep. However, because reaction [8] is rapid, succinimide anions cannot be present at the electrode, during the anodic sweep, as long as there is still some salt 1 left. Indeed, as shown in Figs. 4 and 5, the anodic peak at -0.30 V appears only when there are no more loosely bound silver ions left. Furthermore, its limiting current increases with the concentration of succinimide, anions. In Fig. 4, the maximum value of the limiting current at -0.30 V corresponds to the highest concentration of mixed salt l a , as expected. In Fig. 5, however, the limiting current reaches a plateau when there is a 100% excess of succinimide anions with respect to salt 1, a phenomenon which is also observed for the cathodic wave at

Can. J. Chem. Downloaded from www.nrcresearchpress.com by 192.30.84.250 on 10/04/16 For personal use only.

HUOT ET A L - 1.45 V and for which we cannot provide a satisfactory explanation as stated above. In the absence of succinimide anions, all the electrodeposited silver is oxidized according to eq. [ 101. Since there is a large excess of tetrafluoroborate anions, the rate of this process is limited by the diffusion of silver ions into the solution and the current of the anodic peak at 0 V is large; this current decreases as the peak at -0.30 V (the amount of succinimide anions) increases, as expected (see Fig. 4). The fact that there are two types of silver ions in silver salt 1 derived from succinimide eliminates structure A in which the two silver ions would be equivalent. The ' H and I3C nmr data recorded in Table 2 show clearly that the ISucc-Ag-SuccIAg complex (1) must be symmetric and have structure B in which the tightly bound silver ion is linked linearly to each nitrogen of the two succinimide anions. Indeed, the eight protons of the two succinimide moieties give rise to a singlet; the four carbons

FIG. 6. Lincar voltammetry of the silver salt of benzoylimide (5) (2.15 mM) in acetonitrilc-TEAFB (0. I M ) on a Pt electrode (w = 600 rpm, swcep ratc = 16 mV/s): A , before rcduction; B. after reduction at -0.25 V; C, after reduction at -0.90 V; D, rcsidual currcnt.

of the carbonyl groups give a single signal, as do the four carbons of the methylene groups. The loosely bound silver ions, which are more difficult to reduce ( E l / , = -0.20 V) than silver nitrate ( E l l , = -0.10 V), probably interact with the carbonyl groups of the two succinimide anions. In solution, this interaction is not strong enough to make the rotation around the N-Ag-N bond a slow process on the nmr time scale. Structure B is in agreement with the classical theory according to which an Ag-N bond should be stronger than an Ag-0 bond (12). It also agrees perfectly with the X-ray crystallographic structure of the silver salt of 1-methyl thymine determined by Guay and Beauchamp (8): half of the silver ions are bound linearly to the nitrogens of two imide anions and the other half interact with four carbonyl groups. T h e nmr and ir data for the mixed silver-tetraethylammonium salt l a (Table 2) are very similar to those for the silver salt 1 since the same species, the Succ- Ag-Succ anion, is responsible for the absorptions. W e have already pointed out the similarity of the uv spectra (Table I). As shown in Table 3, the electrochemical and potentiometric behavior of all the silver salts studied is perfectly analogous to that of salt 1 . Reduction at a potential (-E,,,,, = -Ell, + 200 mV) corresponding to the first cathodic wave leads to the deposition of only 50% of the silver ions and the remaining 50% are deposited upon reduction at a potential corresponding to the second cathodic wave. Thus, all the silver salts must have the same [Amide-Ag-AmideIAg structure. T h e silver salts and the corresponding mixed tetraethylammonium salts are oxidized at nearly the same potentials as in the case of salts 1 and l b derived from succinimide (see Fig. 6 for salts 5 and 5 a derived from benzoylimide). The tetraethylammonium salts of glutarimide ( 2 b ) and phthalamide ( 4 b ) are more difficult to oxidize than the corresponding silver ( 2 and 4) and mixed silver-tetraethylammonium ( 2 a and 4 a ) salts as in the succinimide case (see Table 1 and Fig. 1, C ) , whereas the reverse situation is found for the tetraethylammonium salts of tetramethylsuccinimide ( 3 b ) , benzoylimide (56) (see Fig. 6), and formanilide (6b). The oxidation of the silver salt of formanilide (6), a carboxamide, is much easier than that of silver salts of irnides because the energy of the H O M O of the Arnide-

[rutls- irutis

cis- irons

cis -cis

Ag-Amide anion must be lower when the anionic ligands are derived from an imide (two carbonyl groups) than when they are derived from a carboxamide (one carbonyl group) (13). We could not obtain a satisfactory oxidation wave for the silver salt of benzamide (7).1° T h e silver salt of benzoylimide (5) is the sole salt derived from an acvclic imide and is less stable than the other salts. W e but this also tried to prepare [(CH, c),N-A~-N(ccH,),]A~ salt was still less stable and was too sensitive to hydrolysis to be isolated and studied under the standard conditions. T h e electrochen~icalbehavior of salt 5 is somewhat different from that of the other silver salts since the two cathodic waves are separated by about 0 . 3 V as compared to 1.1- 1.3 V for the other salts. Furthermore, as shown in Fig. 6 , the second cathodic wave is not very well defined, being spread over a 0 . 5 V range. In salt 5 , the ligands (the benzoylimide anions) could a priori exist under the three planar conformations shown (Scheme 1). For the free anion, the cis-cis conformation should be much less stable than the two others because of steric interactions between the two phenyl rings which would tend to orient themselves into the OCNCO plane to allow some delocalization of the electrons of the imide anion onto the aromatic rings. However, if the cis-cis benzoylimide ((ciscis)Bzmide)anion is a better ligand to A g e than the two other conformers, the [(cis-cis)Bzmide- Ag-(cis-cis)Bzmide]anion could be present in significant amount together with the other c0&~1ex anions: [ ( c i s - t r a t ~ s ) ~ z m i d e - ~ g (cis-trans)Bzmide]-, [(trans-trans)Bzmide- Ag-(trans10 Further work on the oxidation of the silver salts of imides and carboxamidcs is in progress and will be reported in a forthcoming paper.

C A N . J CHEM VOL. 61. 1'183

Can. J. Chem. Downloaded from www.nrcresearchpress.com by 192.30.84.250 on 10/04/16 For personal use only.

TABLE4 Ultrav~olctspcctral data of thc salts of imitlcs and carboxam~dcsin acctonitrilc-TEABF (0.01 M)"

Silver salt of

Conccntration" (lo-'' M )

Bcforc reduction

Succinimidc (1)

3.35

223(4500)"

Glutarimide (2) Tetramcthylsuccinimide (3) Phthalimide (4)

1.02 2.34

23 l(9000) 227(4000)

1.58

Benzoylimide (5) Formanilide ( 6 )

0.28 0.95

230( 15 600) 236( 15 500) 226(9 100) 24 1 (28 000) 242( 10 500) 209(8900)

Aftcr rcduction 0.5 F/rnol

I F/mol

Parcnt amide'

"0.5 mL of the electrochemical solution was dilutcd to 5 mL with dry acetonitrile.

"Based on the concentration of amide anions. ' I n dry C H K N without any supporting electrolyte. "Taken from Table I

tr~lns)Bzmidel-,[(cis-cis)Bzmide-Ag-(ci.~-trc~tls)Bzn~ide-, etc. The fact ( i ) that salt 5 is less stable than the other silver salts, and (ii) that the reduction of the tightly bound silver ion in 5 and 50 occurs at potentials lower than those observed with the other silver and mixed silver-tetraethylammonium salts, shows that the benzoylimide anion binds A g ' less tightly than a carboxamide anion or a cyclic imide anion. The presence of complexes, a mixture of benzoylimide-Ag-benzoylimide which differ from one another by the conformation of the benzoylimide anions and which have different reduction potentials, could explain the spreading of the cathodic wave due to the reduction of [Bzmide-Ag-Bzmidel- in 5 and 50. W e have included silver acetate in Table 3 for conlparison purposes. The low cathodic half-wave potential as well as the precipitation of 100% of the silver ions by tetraethylammonium bonds are weaker than Ag-N chloride shows that Ag-0 bonds and thus confirms the conclusions of the classical theory (12). The structure of silver carboxylates has been determined by potentiometry (14), linear voltammetry (15), and X-ray diffraction ( 16- 18) and corresponds to:

T h e one cathodic wave and the one cathodic peak observed in the voltammograms for silver acetate show that the silver ions are all equivalent, in agreement with the structure shown above. Table 4 summarizes the uv spectral data, in acetonitrile, of the various salts studied. The parent amides (imides and carboxamides) are included for comparison. As in the case of succinimide, the uv spectra before any reduction and after reduction of the loosely bound silver ions (0.5 F/mol) are identical because the same species, the amide-Ag-amide anion, is responsible for the absorption. The salts derived from imides bearing no aromatic ring(s) show a large hyperchromic effect when compared to the parent imide. However, since succinimide shows a strong band at 190 nm (E = 15 000) in acetonitrile (19), the formation of the imide anion must b e

accompanied by a ca. 35 nnl bathochromic shift of the 190 nm ~ F/mol), further reduction of salts 2 n band.'' As for salt I L(0.5 and 3c1 to give the corresponding tetraethylamnionium salts 2 b and 3 0 (1 F/mol) is accompanied by a small bathochromic shift. However, the hyperchromic effect was observed only in the case of salt I n .

Experimental Ger~ercrl Melting points were determined on a Buchi apparatus and are uncorrected. Infrared (ir) spcctra were obtained using a Pcrkin-Elmer 257 spectrophotometer. Ultraviolet (uv) spectra were recordcd with a Varian-Techtron 635 spectrophotometer. Proton magnetic resonance ( ' H nmr) spectra were measured using Bruker WP-60 and Varian A-60 instruments. Carbon magnetic resonance ("C nmr) spectra were recorded on a Bruker HX-90 spectrometer. Mass spectra were takcn on Hitachi RMU-6E instrument. Microanalysis were carried out by Mr. H. Scguin, National Research Council of Canada, Ottawa. Soh,ent orltl electrolvte Technical grade acctonitrile was purified as reportcd in ref. 21. Tetracthylammonium tctrafluoroborate (TEAFB), prepared from tctrafluoroboric acid and tetraethylammonium bromide (22). was recrystallized from 95% ethanol until thc silver ion test was negative. It was then dried in a vacuum oven for 2 days at 80°C. Electrocl~ettiictrI irlstrurrwntntior~cind rrletl~ods Solutions were deoxygenated with dry and oxygen-frce nitrogen saturated with acetonitrile. A nitrogen atmosphere was maintained over the solutions throughout the experiments. A three-electrode system and a PAR cell (type K0060) werc used for the voltammetric analyses and the electrodepositions of silver. The auxiliary electrode was a platinum wire. Potentials were measured with respect to an A g " / ~ g +(0.0 1 M ) electrode (23). The voltammetric analyses were carried out with a homemade potentiostat. A Tacussel rotating electrode type ED1 (with Controvit speed controller) having a platinum 20,mm' disc was used at rotation rates of 500 to 5000 rpm for the linear voltammograms; the swecp rate was 16 mv/s. The same electrode servcd as a stationary electrode for the cyclic voltammograms; the sweep rate was 200 mV/s. The plat-

el he ionization of phenol to phenolate causes a 25 nm bathochromic shift of the 2L0 nm band (20).

1897

HUOT ET AL.

inum disc was washed with nitric acid, rinsed with water, dried. then polishcd on a Carbimet-600A paper (Buchler) before each usc. The eleetrodepositions of silver were performed with a Tacussel type ASA- 100 potentiostat having both a Tacussel type 1G-5 integrator and an anlrnetcr connected in series within the auxiliary circuit. The working electrode was a 18 cm' platinum plate. After the deposition of silver, the electrode was washed with acetonitrile, dried. and the amount of silver deposited was determined by weight.

Can. J. Chem. Downloaded from www.nrcresearchpress.com by 192.30.84.250 on 10/04/16 For personal use only.

Amities nrrd i1nide.c. Succinimide, glutarimide, phthalamide, and benzanilide were obtained from commercial sources. Tetrorrrethj~lsuc~cir~ir?licle I t was prepared according to the method described by Biekel and Waters (24) in a 22% yield, mp 186°C (lit. (23) mp 187- 188°C); v,,,,, (CHCIj): 35 10,3220, 1790 and 1740 cm- I : 6 (CDCI,): 1.18 and 1.21 (not resolved, 12 H) and 8.56 ppm (m, NH). Berrzoylirnide It was obtained as described by Titherley (25) in a 10% yield; mp 144-145°C (lit. (24) mp 144°C); v,,,.,,(CHCI,): 3440, 3380, 1750, 1690. 1600, and 1580 cm- '; 6 (CDC13):7.40 (m. 6H) and 7.75 ppm (m.4H), the NH absorption could not be seen. Forrnc~rrilide I t was prepared in a 72% yield by refluxing equ~molarquantities of aniline and formic acid for 5 h while rcmoving water w ~ t h a Dean-Stark separator, mp 47°C (lit. (26) mp 50°C); v,,,.,,(CHCI,): 3420, 3380, 1690, and 1600 cm-': 6 (CDC13): 7.15-7.48 (5H), 8.32 (d, J = 1.7 Hz, O7H), and 8.66ppm (d, J = 11.4Hz, 0.3H). Thus the conformer with C-H and N-H in a cis relationship predominates.

Silver salts Silver acetate is commercially available. The silver salts of imides and arnides were prepared according to the following general procedure (27). Equinlolar amounts of silver nitrate and imide or carboxamide were dissolved in 50% ethanol (between I and 2 mL per mol of amide except in the case of benzoylimide where 6 mL per mol of imide was used). The solution was heated to reflux and an equimolar quantity of sodium hydroxide (a l .OO M or a 5.00 M aqueous solution) was then added dropwise and slowly. The precipitate formed was filtered, washed with 95% ethanol, then with ether. It was dried at room temperature and in the dark in a vacuum desiccator containing phosphorus pentoxide. The ir spectra of all the silver salts in KBr show a broad band in the 1600 cm-' region for the C=O stretching vibration and a broad absorption around 3400-3500 cm-' due to the presence of water in the crystals. Except for the silver salt of succinimide ( I ) , the silver salts of imides and carboxamides were generally not soluble enough in acetonitrile to yield satisfactory ir (dry CH,CN) and nmr (CD,CN-D20) spectra. Salt I (succirrimide) It was isolated as a greyish solid, dec. at 250°C (72% yield); v,,,,, (C.H,CN): 3640 and 3540 (water), 1650 (C=O) cm-I; 6 (CH,CN, 20% D20): 2.56 ppm (s). Anal. calcd. for C,H,NO~A~.:H~O:C 22.34, H 2.37, N 6.51, Ag 50.18; found: C 22.72, H 2.27, N 6.68, Ag 49.30. Salt 2 (glutarimirle) It was isolated as a brown solid, dec. at 220°C (81% yield). Anal. calcd. for C6H6N02Ag:C 27.29, H 2.75, N 6.37, Ag 49.04; found: C 27.04, H 2.79, N 6.52, Ag 48.88. Salt 3 (tetrarnerhvlsuccinimide) It was isolated as a brown solid which melted above 280°C (69% yield). Anal. calcd. for C,H,2N02Ag: C 36.64, H 4.58, N 5.34, Ag 41.20; found: C 36.17, H 4.43, N 5.17, Ag 40.70. Snit 4 (phthalirnide) It was isolated as a beige solid melting above 345°C (92% yield). Anal. calcd. for C B H , N O ~ A ~ . $ H ~CO36.53, : H 1.92, N 5.33, Ag 41.02; found: C 37.01, H 1.89, N 5.40, Ag 40.39.

Sol! 5 (berlzoyliniide) It was isolated as a brownish solid, mp 265°C (91% yicld). Arlnl. calcd. for C,,HI,,NO2Ag-fH20:C 49.46. H 3.27, N 4.12, Ag 31.73; found: C 49.67. H 3.22. N 4.04. Ag 3 1.48. Scrlt 6 ~or.rr~nr~ilicle) It was isolated as a greyish solid, mp 182- 185°C (90% yield). Ann/. calcd. for C , H ~ N O A ~ - ~ HC~35.47, O: H 2.98. N 5.91. Ag 45.51; found: C 35.97, H 2.75, N 6.02, Ag 47.01. Scilt 7 (berrznrniele) The orange-brown solid obtained was not pure according to the elemental analysis and was most probably contaminated with silver oxide. Arral. calcd. for C7H6NOAg.$H20 i ~ g ' O :C 30.50, H 2.20, N 5.08, Ag 52.18; found: C 30.06, H 2.18, N 5.36, Ag 53.73.

+

Mixed silver- tetrnethylc~rnrrrorriurnsnlt c!f 'slrccir~irrlide( I a ) To a stirred mixture of succinimide (2.00 g , 20.2 mmol), tetraethylammonium hydroxide ( 14.4 mL of 1.40 M aqueous solution, 20.2 mmol), and water (15 mL) was added dropwise to a solution of silver nitrate (1.72 g, 10.1 nlnlol) in water (30 mL). No precipitation occurred. Thc reaction mixture took a greyish-blue color. The solvent was removed in a rotatory evaporator and the resulting greyish-blue solid was washed with methylene chloride, rnp 190- 192°C (3.26 g, 69%); v,,,;,,(CH,CN): 3620. 3540, and 3420 (watcr). 1640 (C=O) cm-'; 6 (CD,CN, 20% D20): 1.20 (t oft, J = 8 and 2 Hz, 12H), 2.59 (s, 8H), and 3.1'6 ppm (q, J = 8 Hz, 8H). Anril. calcd. for Cl,H2,Nj0,Ag.2H20: C 40.85, H 6.87, N 8.94; found: 40.35, H 6.82, N 8.85. Tetrnetlr~lnmrnor~iur~~ scllt of succ,irlinride ( I b) Tetraethylammonium bromide (42.0 g, 0.20 mol) was dissolved in the minimum amount of hot 95% ethanol. Silver sulfate (3 1.2 g, 0. I0 mol) was added. Silver bromide was removed by filtration and the solvent was removed with a rotatory evaporator. The residue was recrystallized from ethanol-acetone to afford tetracthylammonium sulfate, mp 164- 165°C (5.5 g. 15%). The sodium salt of succinimide was then prepared from succinimide (0.458 g, 4.6 mmol) and sodium ethoxide (30 mL of 0.15 M solution in absolute ethanol, 4.5 mmol). Tetraethylammonium sulfate (0.800 g, 2.25 mmol) in absolute ethanol (4 mL) was added to the previous solution which became clouded. Part of the solvent was removed in a rotatory evaporator. The addition of acetonitrile caused the quantitative precipitation of sodium sulfate which was removed by filtration. The solvent was stripped off in a rotatory evaporator, then the residue was dried in a vacuum desiccator over phosphorous pentoxide to afford salt 10 as a white solid, mp 72-73°C (sealed tube) (0.806 g, 76%); v,,,;,,(CH,CN): 3600 (b, water), 1650 and 1600 (C=O) cm-I; 6 (CD,CN): 1.22 (t o f t , J = 7 and 2 Hz, 12 H), 2.33 (s, 4H), and 3.20 ppm (q, J = Hz, 8H). Sodiurn snlr of succirlitnide The salt was prepared from succinimide and sodium ethoxide (see above). It was also obtained ir~situ for the arnperometric titration of silver nitrate by adding sodium hydroxide (1.30 mL of a 1.0 N solution, 1.30 mmol) to a solution of succinimide (0.127 g, 1.28 mmol) in 50% ethanol (250 rnL).

Acknowledgements We are grateful to the "Ministkre de 1'Education du Quebec" and to the Natural Sciences and Engineering Research Council of Canada for financial assistance. We thank Professor F. M. Kirnrnerle for helpful discussions and Mr. H. Seguin of the National Research Council of Canada for rnicroanalyses. I. 2. 3. 4. 5.

w'. J. COMSTOCKand F. KLEEBERG. Am. Chem. J. 12, 493 (1890). A. W. TITHERLEY. J. Chem. Soc. 71, 490 (1897). G. JANDER and G. WINKLER. J. Inorg. Nucl. Chem. 9 , 2 4 (1959). S. GuloT and B. TREMILLON. J. Electroanal. Chem. 22, 147 (1969). A. R. STEINand S. H. TAN.Can. J. Chem. 52, 4050 (1974).

Can. J. Chem. Downloaded from www.nrcresearchpress.com by 192.30.84.250 on 10/04/16 For personal use only.

1898

CAN J CHEM VOL 61. I983

Thcrnmochirn. Acta, 30, 6. M. J. BLAIS,0. ENEA,and G BERTHON. 37 ( 1979). 7. J. R. DEMEMBER and F. A. WALLACE. J Am. Chem Soc. 97, 6240 ( 1975). 8. F GUAYand A. L. BEAUCHAMP J. Am. Chcnm. Soc. 101, 6260 ( 1979). 9. N. A. MALIK.P J. SADLER. S NEIDLE, and G. L. TAYLOR. J. Chem. Soc Chem Comrnun. 7 1 1 ( 1 978). and T. R. SESHADRI. Proc Indian Acad. SCI. 10. N. V . SUBBARAO Scct. A, 1 (1940). I I . Z. GALUS.Fundarncntals of electrochemical analys~s.Ell~sHorwood Ltd., Westergate, G . B. 1976. p. 249 12. M. ISRAELI and L. D PETIT. J. Inorg. Nucl. Chern. 37, 999 (1975). 13 1. FLEMING. Frontier orb~talsand organic ehem~calreactions. John W~lcyand Sons Ltd., New York. NY. 1976. 14. R. ALEXANDER, E. C F. KO, Y C. MAC.and A. J. PARKER. J. Am. Chern. Soc. 89, 3703 (1967). J . Am. Chcrn. 15. S. H. COHEN, R. T. IWAMOTO, and J. KLEINBERG. Soe. 82. 1844 (1960).

16. R. L. G R I F F J.~ Chem. ~. Phys. 11, 499 (1943). ~ ~ J . J. Am. Chcrn. Soc. 78,3029 17. A . E. B L A K E S L E EL.~ HOARD. (1956). 18.. P. COGGON and A. T. MCPHAIL. J. Chcrn. Soc. Chcm. Cornrnun. 91 (1972). J. Chern. Soc. 4555 (1957). 19. D. W. TURNER. 20. A. 1. SCOT. Ultraviolet spectra of natural products. Per,~ a r n o n Press, Ltd., Ncw York. 1964. p. 93. 2 1. D. BERUBE and J. LESSARD. Can. J. Chern. 60, 1 127 (1982). 22. H. 0 . HOUSE,E. FENB,and N. P. PEET.J . Org. Chcrn. 36,2371 (1971). J . Electroanal. Chern. 1, 486 (1959160). 23. J . P. BILLON. 24. A. F. BICKEL and W. A. WATERS. Recl. Trav. Chirn. Pays-Bas, 69, 1490 ( 1950). 25. A. W. TITHERLEY. J . Chcrn. Soe. 85. 1673 (1904). 26. R. C. WEAST(Erlitor). Handbook of Chemistry and Physics. 53th ed. CRC Prcss, Cleveland. 1972. p. C-303. 27. R. P. BELLand J. P. BROWN.J. Org. Chern. 1, 1520 (1936): J. Org. Chcm. 6, 489 (1941). L. I. SMITHand J. NICHOLS.