A Study of the Pi Complexes of Tetracyanoethylene with Phenols, Aromatic Hydrocarbons, and Aryl Ethers GEORGE H. SCHENK, MILAGROS SANTIAGO, and PATRICIA WINES Deparfment o f Chemistry, Wayne State University, Detroit 2, Mich.
b The T complexes of tetracyanoethylene (TCNE) are useful for determining many pure organic molecules, analogous to the broad utility of metal-EDTA chelates in inorganic analysis. TCNE, like EDTA, is a fairly general complexing agent and reacts instantly under mild, room temperature conditions; it can b e made somewhat selective b y the use of other reagents; and equilibrium constants can b e used for analytical conditions; i.e., the absorbance of a solution of known concentrations of TCNE and T base can readily be calculated. Two types of plots which do not deviate seriously from Beer’s law can be obtained: type I, where the concentration of ir base is higher than that of TCNE, and type II, the opposite case. Solvent effects, temperature effects, and acid-base-salt effects are also investigated. To make TCNE complexing more selective, acetic acid, perchloric acid in acetic acid, acetic anhydride in chloroform, acidcatalyzed acetylation, and other reagents are utilized. The condensation reaction of TCNE with hydrazine yields a plot which strictly obeys Beer’s law.
U
SE OF ethylenediaminetetraacetic
acid (EDTA) as a general complexometric reagent requiring mild reaction conditions (room temperature, neutral pH) is one of the chief tools of the inorganic analytical chemist. Furthermore, it is possible to render EDTA complexation more selective by reduction, masking, kinetic masking, and pH control. Lastly, the chemist can readily calculate the effects of pH and masking. Until recently no similar reagent had the same potential for the organic analytical chemist. dlthough Sawicki (2 4 ) and coworkers have developed a number of excellent, useful reagents, such as piperonal, the isatins, and 3methyl-2-benzothiazolone hydrazone for colorimetric determination of a variety of organic compounds, the methods are not mild. They are also not amenable to calculation. I n 1958 h4errifield and Phillips (8) published a study of the intensely colored T complexes of tetracyanoethylene (TCYE) with aromatic hydrocarbons.
The potential of T C X E soon became apparent as Tarbell and Huang (19) used TCSE for color development in paper chromatography, Peurifoy and Nager ( 2 1 ) utilized T C N E for the estimation of nitrogen compounds, and Schenk and Ozolins (IO,16) used T C N E for the colorimetric estimation of anthracene and the titrimetric estimation of dienes via the Diels-Alder reaction. Recently, Bauer (3) showed that the substitution of unconjugated olefins could be determined with TCNE. Selectivity. T C S E n-a? thus shown analytically to compleu under mild conditions, b u t approaches such as reduction, pH control, and masking h a w not been established to eliminate interferences to make T C S E more specific. Phenols, aromatic hydrocarbons. aryl ethers, mercaptans, amines, thiophenes, and Diels-Alder-active dienes as well as many unconjugated alkenes are among the many organic species that react or 7 complex with T C S E . Because most of the spectral bands that result from these reactions are too broad to be eliminated by spectrophotometry, approaches such as solvent control, acidity. and functional group reactions are explored for possible elimination of interferences, and the results reported herein. Beer’s Law-I. A h o t h e r object of this Iyork was to study the conditions and make calculations for adherence to B e d s law for the .rr complexes formed according to the reaction: TCNE
+ aromatic base
complex (1)
T o obtain formation (or pseudo stability) constants for the above reaction, Merrifield and Phillips (8) plotted the molarity of T C N E divided by absorbance us. the reciprocal of the mole fraction of aromatic hydrocarbon (under conditions where the concentration of aromatic was large compared to TCXE) and obtained straight lines. This indicated that plots might be obtained which, although not strictly obeying Beer’s law, might not deviate seriously from Beer’s law over a small concentration range and under standardized conditions. Such plots indeed have been obtained for the colorimetric determination of aromatic rings where T C N E is present a t lower concentra-
tions than the aromatic ?T base. This method appears best for phenols and some aromatics where K is fairly large ( 320) since a small amount of a weak aromatic base does not form enough of the complex to measure spectrophotometrically. Beer’s Law-11. The more sensitive method involves T C S E a t concentrations appreciably higher t h a n the concentration of aromatic or phenol. The formation constant expression for this situation (16) is:
K =
(C 1
( ( B )- ( C ) } ( [ T C X E ]-
[m
(2)
in which (C) is the concentration of the complex a t equilibrium, ( B )is the initial concentration of the ?r base, [ T C N E ] is the initial mole fraction of T C S E , and [C] is the mole fraction of the complex a t equilibrium. Because most of the T complexes are very weak, [TCSE] - IC] will approximate [TCKE] within 5%, and Equation 2 rearranges to:
The absorbance, A, of a given solution in a 1-cm. cell will equal E ( C ) ,where E is molar absorptivity. The lower limit of ( B ) that can be measured accurately ( A = 0.1) will depend on t , which is of the order of 2000 for most of the complexes in methylene chloride, on [TCXE], which is 0.0032 for the convenient 0.05Alf TCNE solutions in methylene chloride, and on K . Substituting A/E for (C) in Equation 3 gives:
d 0.016 sK[TCNE] ‘v 7 (4)
Where K is of the order of 1, the lower limit of ( B ) that can be determined accurately is about 1.6 X 10-2;M; where K is of the order of 100, the lower limit is about 1.6 X 10-4M. Plots which did not deviate seriously from Beer’s law are also obtained for this method, but since K’s and molar absorptivities are not available to delineate all lower limits of ( B ) that can be accurately measured in different solvents and a t different concentrations of TCNE, still a further object of the VOL. 35, NO. 2, FEBRUARY 1963
167
~~
Table I.
Solubilities of
TCNE and
Solvent
Solubility
CHCla CHBr3 CHzClz CClo: CClZ CHClzCHClz CHzClCHzCl HOAc EtOAc Ac20 0.O6M AczO/EtOBc CHzCN
0 . O l i i '(slow) 0.5M"
... ...
EXPERIMENTAL
Apparatus. The transistorized Bausch & Lonib Spectronic 20 with t h e 11.7-mm. a n d 22-mm. test tubes a n d 10-mni. microcells as well as t h e Beckman-Spinco microspectrocolorimeter (for 1-ml, volumes) were used for spectrophotometric measurements. Reagents. Tetracyanoethylene ( T C N E ) was obtained from the Eastm a n Kodak Co., sublimed to obtain a white solid, and stored in a desiccator over sodium hydrovide pellets. Acetic anhydride, methylene chloride, ethyl acetate, a n d other solvents were ACS grade or reagent grade. E t h y l acetate contained no more t h a n 0.107, mater. ilromatic hydrocarbons and phenols were Eastman White Label or were purified by sublimation or recrystallization. T C S E solutions were prepared, for example, b y dissolving 6.4 mg. of T C S E per nil. with heating for 0.05X T C N E in methylene chloride, or b y dissolving T C N E with stirring a t room temperature for 0.5.11 T C K E in ethyl acptate or 0.331 T C S E in acetic acid. T C K E (0.1JI) in methylene chloride was prepared by the method of Schenk and Ozolins (15). The 0.05JI T C S E methylene chloride reagent remains colorless for over a week. The 0.06Jf acetylating reagent was prepared from ethyl acetate, acetic anhydride, and perchloric acid by the method of Schenk and Santiago ( 1 7 ) . If chloroform was used as solvent instead of ethyl acetate, 1.5 ml. of acetic anhydride per 50 ml. of chloroform was used t o acetylate the O.i5yOethanol in the chloroform. The 0.1JI perchloric acid in glacial acetic acid mas prepared by the method of Fritz (6). The 0.05-l4 periodic acid ANALYTICAL CHEMISTRY
0.012M durene, 0.002M TCNE, A 4 e ( b = 22 mm.) Approx. 2 (pptn.) 0.63 1.05 0.38
0.15M 0.31V >O. 5M
study was to determine some of these experimentally. For this purpose and for the purpose of investigating adherence to Beer's law, a five- to tenfold excess of T C K E over aromatic or phenol appeared sufficient to ensure a constant slope, A.A/A(B),for all but the strongest T bases.
168
Solvent Effects
Approx. 0.00211.1
...
p-Dioxane CHaOH
~
was made b y dissolving HsIOe in methanol and diluting with acetic acid so that the final concentration of methanol was 10%. Procedures. Solvent effects were determined b y mixing 1 ml. of 0.05V TCNE in methylene chloride with 0.3 mmole of durene a n d diluting to 25 nil. with t h e appropriate solvent. Temperature effects were investigated b y reading t h e initial absorbance of about 15 ml. of solution, refluxing t h e solution containing the complex i n a 100-ml. volumetric flask, and withdrawing 1-ml. aliquots after appropriate times for absorbance readings. Salt effects were investigated by coniparing absorbance readings of solutions of complexes identical except that a salt such as LiCl was present in one solution. Adherence to Beer's law was investigated b y first selecting chlorinated solvents for making u p the sample as well as dissolving TCNE. Solvents such as amines, dimethylformamide, nitromethane, or triethyl phosphate reacted with T C N E and could not be used. Where mixed solvents were used such as chloroform-methylene chloride, the proportions used were kept constant since Table I indicates wide variation in absorbance with a change in solvent. The T C N E solution was then added to the sample of aromatic or phenol and the absorbance was read after mi\;ing a t the appropriate wavelength. Since water hydrolyzes T C S C dowly even where the solvents have been rigorously dried ( I ) , both chloroform and methylene chloride were saturated with 0.5% of their volume of water and absorbance readings of the 2,6-dimethylphenol complex checked periodically. The effect of absorbance appears to be less significant for chlorinated solvents than for acetone ($1) and acetonitrile (1).
Both the use of acetic acid and 0 . 1 V perchloric acid as solvents were iiivestigated to prevent the interference of amines in complexation of aromatic hydrocarbons or phenols. Mixtures of aromatic or phenol and amine were diluted to within 5 ml. of the final volume with either acetic acid or 0.1JI
perchloric acid in acetic acid before solid T C N E or an acetic acid solution of T C K E was added, and the absorbance read at 520 mp, for instance, for 2,6-dimethylphenol. Acetic anhydride was used for the same purpose; 1 ml. of acetic anhydride was added to the above mixtures in 10 ml. of chloroform, and the solution was allowed to stand 30 minutes a t room temperature. T C N E in methylene chloride was added in excess of the aromatic or phenol or in smaller amounts than either. The solution was diluted to 25 ml. with chloroform, and the absorbance was measured a t the appropriate wavelength, such as 520 mp for 2,6-dimethylphenol. The blank (no amine) was treated similarly. The use of the 0.06~11acetylating reagent (0.007A1 in perchloric acid) was investigated to prevent the interference of amines, phenols, and mercaptans in the complexation of aromatic hydrocarbons or aryl ethers. A threefold excess ( 5 ml.) of the acetylating reagent was added to 0.1 mmole of amine, phenol, or mercaptan and 0.5 mmole of 1,3-dimethoxybenzene in 0 to 3 ml. of methylene chloride. After 20 minutes reaction a t room temperature, the solution was diluted to about 20 ml. with chloroform, 1 ml. of 0.05M T C K E in methylene chloride was added, the solution was diluted to 25 ml. with chloroform, and the absorbance was read at .Si5 mfi. The blank (no phenol, amine, or mercaptan) was treated similarly. The use of the same acetylating reagent was investigated to prevent the interference of other phenols in the complexation of 2,6 - di - tert - butyl4-methylphenol. A minimal amount (2 ml.) of the acetylating reagent was added to 0.05 mmole of other phenol and 0.2 mmole of 2,6-di-tert-butyl-4methylphenol in 0 to 1 ml. of methylene chloride or ethyl acetate. After 20 minutes reaction a t room temperature, i ml. of 0.1M T C N E in methylene chloride or i ml. of 0.5X T C S E in ethyl acetate was added, the solution n as diluted to 10 ml. with either solvent, and the absorbance was measured immediately a t 630 nip. This type of phenol 1 eacts slowly (the absorbance slowly changes) n i t h the acetylating reagent in contrast to the 20-minute reaction time for most phenols ( 1 7 ) . The serious interference of 2-tert-butylphenol was evaluated a t both 630 and TOO mp. The expense of using 0.5M T C N E was overcome by using 1-nil. volumetric cylinders, micropipets, and microcells for the Spectronic 20 or Spinco microcolorimeter. The condensation of hydrazine with T C S E t o form the yellow hexacyanodiazahesadiene dianion ( 7 ) was also investigated. To 1 mi. of an aqueous hydrazine solution was added 3 ml. of 0.005M T C N E in acetonitrile, dioxane, or tetrahydrofuran. -4 minimum volume (0.5 to 4 ml.) of phosphate buffer was added to keep the apparent pH a t i. The mixture was allowed to stand from 30 to 60 seconds and was timed so that all samples reacted for the same time. After dilution to 50 nil. with the respective organic solvent
and a timed interval of 10 to 20 minutes + 40 seconds, the absorbance was read at 440 mp (acetonitrile), 445 mp (tetrahydrofuran), or 456 mp (dioxane). -4 solvent blank of 3 ml. of 0.005M T C N E , the same volume of water as the sample, the same volume of phosphate buffer, and a dilution t o 50 ml. with the same organic solvent, was used.
DISCUSSION
Solvent Effects. Table I gives t h e solubility of T C S E in a number of solvents, a n d shows t h e effect of solvent on the absorbance of t h e durene-TCNE T complex. Solvent effects were also investigated for some solvents for the T complexes of 1,3dimethoxybenzene, 2,6-dimethylphenol, and naphthalene, and rough agreement was noted. This supplements the excellent work of Cram and Bauer (4) on the shift of absorbance maxima with solvent for 3,4-paracyclophane-TCNE T complexes. They suggested that one or two solvent molecules were bound in the T complex, giving rise t o the solvent effect. T h e change in absorbance from solvent to solvent for the TCNE-durene K complex reflects not only a change in K , b u t also possibly a change in e, as seen from Equation 4. The data of Merrifield and Phillips (8) indicate that only K decreases appreciably with a change from chloroform to methylene chloride but that K decreases appreciably and e increases appreciably with a change from chloroform t o diethyl ether. I n general, the most intense colors result in chlorinated or brominated alkane solvents although benzyl chloride also favored complex formation; unfortunately it also P complexes with T C X E . Although T C N E was added as 1 ml. of a n 0.05X solution in methylene chloride t o obtain the data, apparently this was not sufficient to preL ent precipitation in carbon tetrachloride or perchloroethylene, both of which favor complexation but arc Iioor solvents for TCSE. Where there is no x complexation possible to assist in solvation, the minimum requirement for solvation of T C N E appears to be compounds of the type HCR,, vhere the R's are all electron withdrawing groups-e.g., chloroform or bromoform. The hydrogen of bromoform is inductively less acidic and more sterically hindered than t h a t of chloroform, probably forms a w-eaker hydrogen bond (Equation 5 ) and is therefore a poorer solvent than chloroform, as shown in Table I. (?;C)&:C(CN)2
+ HCR,
CN (?;C)ZC : CC?; : * HCR,
(5)
Table II. Aromatic, 0.01M Naphthalene
TCKE, Jf 0.025
2,6-Dimethylphenol Phenanthrene 2,6-Dimethylphenol Naphthalene
0.03 0.03 0.03 0.02
Temperature Effects
Solvent(s) (0" C.) 60% CHsCIr50% CHC18 (about 50 ') EtOAc (77") EtOAc (77') CH2ClCH&1(84") HOAc (118")
Absorbance Initial 60 min. 0.23
0 24
0.29 0.15 0.375 0.06
0.30
2,GDimethylphenol
0.02
HOAc (118")
0.12
2,6-Dimethylphenol
0.02
BuOAc (122')
0.12
Hov-ever, the hydrogen of bromoform or chloroform can also hydrogen bond to a n aromatic T electron system (Equation 6 and Table I), as shown by Tanires (18). Thus chloroform, the better solvent for TCK'E, forms a stronger hydrogen bond t o T electron systems according to Tamres (It?), competes with T C N E more favorably than bromoform, and therefore favors T complexation of T C K E less than bromoform.
+
Me4C6H2 HCR3 e T
( Me4CsHt) -*
HCR, (6)
A compound of the type H2CR,, such as methylene chloride, possesses less acidic hydrogens than the type HCR3, but does have a statistical advantage and a smaller steric bulk. The equilibria in Equations 5 and 6 are more favorable for methylene chloride than for chloroform, making it a better solvent but a less favorable media for x complexation of T C N E with an aromatic such as durene in Equation 6. Both acetic acid and ethyl acetate are better solvents than methylene chloride, which again reflects the statistical advantage of hydrogen bonding; Table I also indicates neither favors complexation as much as chlorinated solvents do. Dimethylformamide, dimethyl sulfoxide, and nitromethane also satisfy the requirements for solvation, and all three are good solvents, but all three apparently react with T C S E , producing yellow colors. Tarbell and Huang (19) indicate that the solubility of T C N E in chloroform is only 2 mg. per 100 ml. (0.00015AU),much less than that indicated in Table I. This could only be accounted for by assuming that their chloroform did not contain 0.75y0 ethanol as preservative. T C N E is extremely soluble in ethanol, and this may account for the slow dissolution of T C N E in chloroform containing ethanol as preservative. Temperature Effects. Peurifoy and Nager ( 1 1 ) reported that many
0.14 0.346 0 72 (10 min.) 0.59 (10 niin.) 0.20
T C S E-aromatic hydrocarbon colors disappear when such mixtures are heated to dryness at 110" C. in benzene or acetic acid. However, phenol-TCSE colors of red or blue faded to a light yellow or brown. I n solution, though, aromatic- or phenol-TCXE complexes appear stable in solvents boiling as high as 84" C. for as long as an hour, as shown in Table 11. Thus, the statement that "hydrocarbon-TCSE colors which persist at room temperature are unstable upon heating" ( 1 1 ) is subject to qualification. These colors do disappear rapidly in acetic acid a t 110" t o 118" C., but before they have completely disappeared, the solution becomes progressively more yellow, apparently as a result of the decomposition of T C X E . Acetic acid solutions of T C X E also become dark yellow after standing a t room temperature for more than a day. It a ould appear that Peurifoy and Nager's estimation of nitrogen compounds by formation of a yellow color after heating to dryness a t 110" C. would not be applicable to a homogeneous solution mhere T C N E forms a yellow color itself even in the presence of phenols or h>diocztrbons. Furthermore, it is not obvious that temperature control n-ill differentiate betn-een complexation of aromatics as opposed to phenols even tliough 2,6-dimethylphenol is reported t o condensr in the presence of pyridine and heat 11ith T C S E (15). Acid, Base, a n d Salt Effects. Table I indicates t h a t acetic acid 9s solvent favors complexation more th:in nonhalogenated solvents. Table I11 indicates t h a t 0.1-11 perchloric acid in acetic acid and 0 05M periodic acid in 90y0 acetic aci&lO% niethanol apparently do not disturb the equilibrium in acetic acid. (Tables VI1 and VI11 elaborate on the effect of perchloric acid in the presence of amines.) I n contrast, most bases tend to catalyze the decomposition of T C K E , giving rise t o yellow colors. K i t h VOL. 35, NO. 2, FEBRUARY 1963
169
alkyl amines, the color is probably due t o a condensation product of T C N E and the amine, but the color also occurs
Table 111.
Acid, Base, and Salt Effects
0.004M 2,6-dimethylphenol,
0.025M T C S E , HOAc ( 6 = 22 mm.)
Acid, base, or salt -45~0mp *4360mJL Sone (blank) 0 25 0 06 0 11Tf HClO4 0 258 1134 NaOAc 0 245 0 35 1X Urea 0 25 0 65 1.4M LiCl 0 23 1 2 0 7JI iXaC101 0 23 0 23 0 04Mn-BuSHz 0 30 1 7 0 002M 2,6-dimethylphenol, 0.02M TCSE, 487, CHC13-487, CH2Cli Base None (blank) 0 22 0 0004M Et3N 0 05 ca. 1 0 04M durene, 0 00211.1 TCXE, EtOAc Salt
-4480 m p
None (blank) 0 172 0 07 0 5M NaC104 0 163 0 24 0 012X durene, 0 002M TCXE, 90% HOAc-10% MeOH Acid A 4 8 0 rnp None (blank) 0.155 0.05JI HI04 0 14
Table IV.
Compound Pyrene
(6, mm.) 720 1 . .
Pvrene [trace anal. ( I $ ) ] 2,6-Dimethylphenol in 0 02M p-nttrophenol 3,S-Dimethylphenol in 0 02M o-nitrophenol 2,6-Di-tert-butylPmethylphenol 1,3-Dimethoxybenzene in 0.004M resorcinol Hydrazine (condensation)
- \
(11
720
in acetic acid with bases such as sodium acetate and urea, and to some extent with the salts, lithium chloride, and sodium perchlorate. As seen in Table 111, the absorbance of a peak occurring a t wavelengths appreciably longer than 360 mp, such as 520 mp for 2,6-dimethylphenol, is not significantly affected b y bases other than alkyl amines as long as T C N E is present in excess of the 2,6dimethylphenol. However, a t 360 mp, where the TCNE-2,6-dimethylphenol complex shows a trough in the absence of bases or salts, strong peaks occur in the presence of bases or salts. It was not immediately obvious from the latter spectra that a peak, characteristic of a possible TCKE-phenoxide anion r complex, was present. The data of Table 111 are in agreement with the observation of Peurifoy and llrager (11) that tertiary alkyl amines give intense yellow colors with TCNE, even though AIcKusick and coworkers ('7) reported no reaction. Peurifoy and Sager reported no color when T C N E was applied to paper containing n-butylamine; in acetic acid solution a strong yellow color is observed. Kinstein, Smith, and Darwish (22) have reported large salt effects on the rate of ionization of p-methoxyneophyl p-toluenesulfonate in less polar organic solvents. It was not anticipated that
Agreement with Beer's Law
Upper limit of concentration 5X
TCSE, Final M 0 07
0
(11.7)
520
(22) 520 (22)
0.07
70% CHzCl?, 30% CHCli
0.002
2% CHzClz, 987, CHC1,
0.002
2% CH~CIZ, 98% CEICl,
630
0.07
(22) 440 (10)
0.0003
(11.7) 575
Table V.
Solvent( s) 70% CH2C1,, 30% CHCI3
0.002
Lower Limits of Concentration ( A = 0.10, 6 = 1 cm.)
2,B-Di-tert2.6-
Solvent CHzClp (theory) CHX1,
CHC1, 70% CH2C1,-
Durene Pvrene (29.5) (TCXE) [TCKE] ( K ) = (54) 0.050 0.0032 3 . 5 x 10-4 1.0 x 10-3 0.050 0.01
30% CHCli 0.70 HOAc 0.30 EtOAc 0.50 170
0,0032 0.0008
0.0047
0,018 0.051
ANALYTICAL CHEMISTRY
3
x
sx 3 2
10-4 10-4
x'id-4
x
10-4
1
x
10-3
...
5 x 10-4 6 X 9 x 10-4
Dimethylphenol (-20)
x
10-4
sx
10-4 10-3
7 3
x
3 X'io-4
z x
10-4
but~l-4-
mechhylphenol (-0.5)
... x 10-3 3 x 10-2 1 x 10-2 2 x 10-3 3 x 10-3 9
salts would decrease the degree of association through ?r complexation, and the data of Table I11 appear to support this idea. I n acetic acid, many salts can act as Brgnsted bases and apparently the effect of lithium chloride and sodium perchlorate is to some degree the same as the effect of sodium acetate and alkyl amines. -4fter an hour, both the absorbance a t 520 and a t 360 mp had increased in lithium chloride-acetic acid solutions. Mercuric acetate (0.01-11) in ethyl acetate-niethanol-acetic acid had no apparent effect on the 2,6dimethylphenol complex although i t might he expected to compete as a x acid. This is being investigated further. Agreement with Beer's Law. Table I V lists various complev systems (containing a n aromatic, a phenol, or a n aryl ether) which do not deviate significantly from Beer's law from zero concentration t o the upper limit of concentration stated. =In absorbance of 0.07 is obtained for 5 X 10-4Alf pyrene and for 4 X 10-i31 2,6-dimethj-lphenol. Table IV illustrates both the use of T C S E in excess over the r base as well as the use of the aromatic ?r base in excess over TCNE (plots I1 and I). Several plots are listed in which interferences such as nitrophenols or resorcinol are present. Also listed is the determination of hydrazine by condensation with T C K E ; this does not involve complexation. Chloroform is used with methylene chloride to favor complexation. Lower Limit of Concentration. Table V contains data on lower limits of concentration for which absorbance can be measured accurately (0.10) for four different solvents and for one solvent mixture. The d a t a were found directly or calculated from data such as in Table I , using Equation 4 with the approximation that ( B )is equal to A / & [TCNE]. This corresponds to a n error of 16% for durene in methylene chloride when the exact solution is used in Equation 4. Since this is the strongest complex listed and since the data are the result of a single determination of &, the error for any value is estimated to be no more than 20 to 30%. T o find a more exact lovier limit, the graphical method of Merrifield and Phillips (8) would have to be used. The use of Equation 4 and a single measurement make possible a rapid calculation, for any concentration of TCNE, of the order of magnitude of the lower limit of concentration of an aromatic or phenol that can be spectrophotonietrically determined. The four compounds in Table V cover the range of strengths of common T bases, from a K of 54 t o a K of about 0.5 in methylene chloride. K will
vary from solvent to solvent as may the ratio of a n y two K’s. The data in Table V are calculated for 1-cm. cells in the Spectronic 20 since the theoretical lower limit is calculated on a 1-cm. basis using molar absorptivities determined on a prism spectrophotometer (Cary) (8). It is estimated that the data would be about 15% lower on a prism spectrophotometer such as a Reckman spectrophotometer because of correspondingly higher molar abqorptivities. The data in Table 1- indicate that T C S E in acetic acid or ethyl acetate is a more sensitive reagent than T C S E in chloroform or 0.0511 TCNE in methylene chloride. Although 0.1V T C X E in methylene chloride can be used, it is more difficult to make the final solution 0.131 in T C N E for methylene chloride than for acetic acid or ethyl acetate since methylene chloride solutions must he heated to effect rapid dissolution. The use of mixed solvents such as chloroform with methylene chloride improves sensitivity for at least the pvrene-TCNE complex and for other complexes not listed in the table. The 0.05J1 TCSE-methylene chloride rragent does have better color stability than T C N E solutions in acetic acid or ethyl acetate over a period of days. INTERFERENCES A N D SELECTIVITY
Because there are many possible approaches to removing interferences, they have been summarized in Table TI before being discussed separately. It ihould be stressed that these general methods may not give the desired selectivity in every case. particularly where there is a large difference in n basicities. Effect of Water. Since water hydrolyzes T C N E slowly even where a iolvent such as acetonitrile ( 1 ) has bern rigorouslv dried, the effect of mater on the absorbance of the TCNE2,6-dimethylphenol complex with time uaq checked. After an hour. the absorbance of a ‘‘wet” chloroform solution did not change although it decreased from 0.67 to 0.60 after 24 hours. The abqorbance of a similar “wet” methylene chloride actually increased from 0.50 to 0.56 in a n hour and to 0.61 in 24 hours. Obviously the effect of mater is small in chlorinated solvents; the uqe of acetic anhydride or acid-catalyzed acetylation qhould minimize hydrolysis of T C N E in any sol1 ent. Acetic Acid as Solvent. T h e use of glacial acetic acid as solvent pro\ ides no ~ e l e c t i r i t v b u t should retard t h e reactions of some amines with T C N E ; the more reactive amines qtill interfere somewhat as shown in Table VII. -1cctic acid prevents t h e interference of tertiary amines and secondary alkyl amines. This is useful rince tertiary amines cannot be
Table VI.
Approach
Approaches to Selectivity and Removing Interferences
Permits determination of
Literature and table reference
In the presence of
HOAc solvent
Table VI1
A4roniatics, aryl ethers, 2” alkyl amines, 3 ” phenols amines O . 1 M HClOd in Aromatics, aryl ethers, ,411 amines phenols HOAc AclO at room tem- .4romatics, aryl ethers lo, 2’ amines, water perature Aromatics, aryl ethers Most phenols, mercaptans, Acid-catalyzed acetylation at alcohols, 1’ and 2’ room temperature amines, water Acid-catalyzed 2,B-Di-tert-butylMost phenols except acetylation at phenols strong x bases such as room temperature 2,6-di-methylphenol Sitro- and halo-snhstiStrength of x bases Dialkyl-substituted phenols and other tuted phenols and other strong x bases weak x hasps: small amounts of thiophene Aromatics and possil)ly Large amounts of thioHg( O+c )z precipit’ation phenols phene and possihly mercap t ans Anthracene and other Cvclopentadiene and Maleic anhydride Diels-Alder aromatic dienes that react rapidly dienes at room temperature with maleic anhydride Bromination Possibly 2,4,6-trialkyl- Possibly unsubstituted phenols phenols and aryl amines HIOj in HO.4c Aromatics Possibly aryl ethers and phenols
acetylated and interfere in chlorinated solvents containing acetic anhydride (Table IX). Aniline apparently complexes JTith T C N E in acetic acid, or undergoes N-alkylation, and this increases the intensity of the colored solution enormously. Butyl amine also undergoes N-alkylation and interferes somewhat. Perchloric Acid in Acetic Acid. Table VI11 indicates t h a t 0.1M perchloric acid in acetic acid prevents the interferences encountered in acetic acid alone even when t h e interfering amines are present a t 3 t o 6 times t h e concentration used in Table VII. It appears t h a t protonation b y t h e perchloric retards S-alkylation of the primary and secondary amines as well as perhaps 4-tricyanovinylation of S,Ndimethylaniline. Uncatalyzed Acetylation. It is well known t h a t acetic anhydride reacts rapidly with most primary and secondary amines even without a catalyst. Vogel (20) uses pyridine as catalyst for their determination at room temperature in 30 minutes reaction time. The data in Table IX demonstrate that the interference of most primary and secondary amines can be removed after uncatalyzed acetylation (no pyridine) a t room temperature. With aniline, it is necessary to use a smaller concentration of T C X E than aniline to avoid interference. With alkyl amines, it is desirable t o use a
Table VII.
Table VI11 (20)
Table I X (17)
Table S (17)
Table S I Table SI1
(2-31
(16)
(6)
($1
Acetic Acid as Solvent
0 004111 2,6-dimethylphenol, 0.026.11 TCTU’E, absorbance constant for 40 min. 0 002V amine
None (blank) EtqN PhNMez BuzXH PhSHMe BuNH2 (0 00421) PhNHz (0,004-V)
(b
Am =
mp
22mm.)
0.25 0.24 0.265
0.25
0.31 0.38 Approx. 1
Table VIII. 0.1M Perchloric Acid/ Acetic Acid as Solvent 0 004.U 2,6-dimethvlphenol, 0 026M TCYE - 4 5 2 0 mp
0.012M amine
( b = 22 mni.)
None (blank) PhNHz BuSHz PhSHMe PhTMe2
0 26 0 265 0.24
0.245 0.27
higher concentration of T C N E than amine. Triethylamine interferes, but this could be avoided if it were present with priiiiary and secondary amines, by acetylation first and then addition of acetic acid, as in Table VII. Acid-Catalyzed Acetylation. Acetic VOL. 35, NO. 2, FEBRUARY 1963
171
reagent of Schenk and Santiago ( 1 7 ) . Not only does acetylation take place rapidly at room temperature, but only a minimal amount (0.0651) of acetic 2,6-Dimethylphenol (2,6-DMP), amine in anhydride is introduced so as not to Ac?O/CHClafor 30 minutes) disturb the complexation equilibrium in &20 rnp halogenated solvents (Table I). I n ( b = 22 mm.) 0.00441 amine addition, the perchloric acid (0.007M) 0.002M 2,6-DbIP, 0.02M TCNE, 48% has no effect on the complexation as CHCl3-48% CH,Cl? would the large amount of pyridine necessary for catalysis. Finally the Xone (blank) 0.22 PhKHz 0.30 solvent can be varied from chloroform 0.26 p-BrC8H4TH2 to ethyl acetate, etc., depending on the BuNH~ 0.22 sensitivity desired. PhNHMe 0.24 The results should not be taken to BQNH 0.22 PhQNH indicate that the interference of all 0.27 0.05 phenols may easily be removed since resorcinol is acetylated twice and its 0 008X 2,6-D111P90 00251 TCSE, 96% TT basicity is correspondingly reduced. CHC13 Stronger R bases such as p-methoxy0 11 Kone (blank) phenol may interfere somewhat after PhNHq 0 115 acetylation. Alcohols are also acetyp-BrCeHJH: 0 11 BuzTH About 0 lated easily by the reagent although they BuXHz 0.095 do not interfere unless a base such as urea, etc., is present. Tertiary amines obviously interfere if present in larger amounts than the perchloric acid. Table XI illustrates a further applianhydride alone a t room temperature cation of the acetylating reagent to rea& rapidly only with amines, b u t the determination of 2,6-di-tert-butyl-4acetylation catalyzed b y pyridine at methylphenol which does not acetylate reflux temperatures (20) occurs with readily (17) and hence its R basicity phenols and mercaptans as well and is not reduced as is the basicity of other would be a way to some selectivity. unhindered phenols. However, T C K E is not necessarily Strength of R Bases. Some arostable at reflux temperatures, and pyrimatic hydrocarbons and phenols m a y dine would complex the T C N E strongly be determined in the presence of enough to prevent complexation in most others b y virtue of being much cases (8). stronger R bases. Table I V lists Table X illustrates the usefulness of Beer’s law conditions for the dethe micro acid-catalyzed acetylating termination of 2,6-dimethylphenol and 3,5-dimethylphenol in 0- and p-nitrophenol. Table XI1 lists other selective determinations that are possible. Table X. Acid-Catalyzed Acetylation The magnitude of these interferences a t Room Temperature would be expected to increase if a weaker TT base than 2,6-dimethylphenol (0.0ZM 1&dimethoxybenzene, 0.002M were being complexed. TCKE, 0.06.W Ac?O/CHCIa) Mercuric Acetate. Mercuric aceA515 mlr t a t e precipitates thiophene after heat( b = 22 mm.) 0.004M interference ing, and this may possibly be used S o n e (blank) 0 21 to remove large amounts of thiophene. Resorcinol (no 0 06M AcnO) 0 27 (28) has, in fact, used the Wronski Resorcinol 0 20 mercuration of thiophene as a n ana0 21 Aniline (filtered) Dodecyl mercaptan 0 22 lytical method, It may also be useful 0 225 p-hlethoxyphenol in precipitating mercaptans or some phenols. Maleic Anhydride. Schenk and Table XI. Acid-Catalyzed Acetylation Ozolins (16) have shown that maleic and 2,6-Di-fert-butyl-4-methylphenol anhydride can be used to react with cyclopentadiene a t room temperature 0.02X in the phenol, 0.073f TCXE, 207”0 0.06M AcnO/EtO$c, 807, CH2C1, to prevent its Diels-Alder reaction with TCNE. Other dienes react to varying 0.005M A630 m p degrees at room temperature with interference ( b = 11.7mm.) maleic anhydride. 0.20,O. 124 (700 mp) Xone (blank) Periodic Acid. Adler and Mag2,6-dimethylphenol nusson (2) have pointed out the (no 0.06M AczO) 0 . 4 2,6-Dimethylphenol 0 . 2 4 rapid oxidation of most phenols, but p-tert-Amylphenol 0.24 not phenol itself, and methyl aryl 2-lert-Butyl-6ethers with sodium metaperiodate in methylphenol 0.22,O. 132 (700 mp) Benzene 0.21 85% acetic acid. It thus appears feasible to oxidize phenols to quinones, Table
172
IX.
Uncatalyzed Acetylation a t Room Temperature
ANALYTICAL CHEMISTRY
Table Xll. Complexation of a Stronger T Base in Weaker T Bases
0.008M 2,6-Dimethplphenol, 0.002M
TCNE, CHCli
Weaker
?r
bases, A1
ib
A ~ mMp 22 mm.)
=
None (blank) Salicylic acid 0 01611 Benzene, 0 03231 m-ChloroDhenol, 0 00431 Anisole, 0 00431 Acetophenone, 0 032W p-Yitrophenol, 0 016M m-Nitrophenol, 0 016W o-Sitrophenol. 0 O l G M Thiophene, 0 02431
0 0 0 0 0 0 0 0 0 0
16
165 145 155 16 17 17 16 16
175
which do not complex with TCNE, and to determine phenol itself or aromatic hydrocarbons. Although the methanol-acetic acid solution of periodic acid oxidizes phenol as well as substituted phenols, i t does not affect aromatic hydrocarbons such as durene, as shown in Table 111. Thus, 0.024N naphthalene can be estimated in 0.006M p-methoxyphenol using 0.04Ji‘ T C N E in acetic acid-methanol where the final concentration of periodic acid is about 0.03M. Unfortunately, the p-benzoquinone produced necessitated spectrophotometric measurement a t 600 or 620 mp. Without periodic acid, the absorbance of such a solution a t 600 mp was 0.155; after an hour’s oxidation, the absorbance was 0.125. The absorbance of a similar solution without p-methosyphenol was 0.110. Bromination. Since unsubstituted phenols (6) and aromatic amines react rapidly with bromine to form the weakly R basic 2,4,6-tribromo derivatives, it might be possible to complex the more strongly basic 2,4,6-trialkyl substituted phenols and anilines, which would not brominate in such a mixture. The excess bromine might be removed by the addition of a slight excess of an olefin. This is under further study. LITERATURE CITED
(1) Abrahamson, E. A., E. I. du Pont de Nemours & Co., Wilmington, Del., private communication, 1962. (2) Adler, E., Magnusson, R., Acta Chem. Scand. 13, 505 (1959). ( 3 ) Bauer, R. H., ANAL. CHEY. 35, 107
(1963).
( 4 ) Cram, D. J., Bauer, R. H., J . Am. Chem. SOC.81, 5971 (1959).
(5) Fritz, J. S., “Acid-Base Titrations in Nonaqueous Solvents,” p. 13, G. F. Smith Chemical Co., Columbus, Ohio, 1952.
(6) Ingberman, A. K., AXAL. CmM. 30, 1003 (1958). (7) McKusick, B. C., Heckert, R. E., Cairns, T. L., Coffman, D. D., Mower, H. T., J . Am. Chem. Soc. 80, 2806 (1958).
(8) Merriheld, R. E., Phillips, W. D., Ibid., 80, 2 i i 8 (1958). (9) Middleton, VI'. J., Little, E. L., Coffnian. D. D., Enaelhardt, V. A4., Ibid., SO; 2795 (1958).110) Ozolins, X., Schenk, G. H., ANAL. CHEM.33, 1035 (1961). (11) Peurifoy, P. V., Nager, &I., Ibid., 32, 1135 (1960). (12) Rcilley, C. N., Crawford, C. M.> Ibzcl.> 27, 1716 (1956). (13) Yausen, G. N., Engelhardt, T'. 8., Iliddleton, JT. J., J . iltn. Chem. SOC. 80, 2815 (1958).
(14) Sawicki, E., Rec. Chem. Prog. 22, 249 (1961). (15) Schenk, G. H , Ozolins, &ASAL. I., CHEiV. 33, 1562 (1961). 116) Schenk. G. H.. Ozolins. hl., Talanta 8. 109 119Gli. (17) Schenk, G. H., Santiago, M., Microchem. J . VI, i 7 (1962). (18) Tamres, M., J . Am. Chem. Soc. 74, 3375 (1952). (19) Tarbell, D. E., Huang, T., J . Org. Chem. 24, 8 8 i (1959). (20) Vogel, A. I., L'Elernentary Practical Organic Chemistry," Part 3, p. 694, Longmans, Green, London, 1958. ~
(21) Wilson, J. R., Sutting, 11.D., Bailey, G F., ANAL.CHEV.34, 1331 (1962). (22) Winstein, S., Smith, S., Darwish, D., J . ilrri. Chem. SOC.81, 5511 (1969). Z. Anal. Chem. 174,280 (23) Wronski, &'I., (1960).
RECEIVED for review September 10, 1962. Accepted December 19, 1962. Work supported by Public Health Research Grant RG-7760 from the National Institutes of Health, Public Health Service. Presented in part at the 13th Pittsburgh Conference on Analytical Chemistry and Bpplied Spectroscopy, March 1962.
Neutron Activation Analysis of Aluminum Determination of Gamma-Emitting Impurities with Long Half Lives FRANCESCO GlRARDl and ROMANO PIETRA Servizio Chimica Nucleare, Centro Comune di Ricerche Euratom, lspra (Varese), Ifaly
A separation scheme i s presented for the systematic determination by neutron activation of 13 trace impurities in aluminum for nuclear reactor technology. The determinations are based on a group separation of the activated nuclides in the carrier free state b y ion exchange, isotopic exchange, and coprecipitation. The separated fractions are then analyzed b y y-ray spectrometry. All the impurities that give rise to long-lived gamma-emitting isotopes have been considered. Quantitative separation has been studied for the most common impurities and for the elements of greater importance from a nuclear standpoint. The presence of other impurities may be detected and their quantity may be roughly estimated b y analysis of the gamma spectra of the separated fractions. The rare earths are separated as a whole. The carrier free state makes the final separation of individual rare earths simpler and faster. The absence of gravimetric yield determinations results in a considerable saving of time. The complete analysis requires 7 hours, including the activity measurements. The automation of the ion exchange separations has decreased still further the operating time.
P
is frequently used in nuclear reactor technology for its low neutron absorption cross section and rapid decay of the induced activity. The presence of some impurities, even at the trace level, has a rather negative effect on the nuclear properties of aluminum. These impuriERE ALUMINUM
ties include fissile and fertile elements (uranium, thorium), elements Kith high neutron absorption cross section (boron, cadmium, and many rare earths), and elements giving, upon neutron activation, radioisotopes of long half life (cobalt, iron, chromium, scandium, cesium, etc.). The impurities t h a t give rise to long-lived gamma activity are of special interest from the standpoint of radiation hazard involved when equipment must b e irradiated for a long time and then extracted from the reactor and handled for further operation or disposal. The determination of these elements a t the trace level is therefore of great importance in establishing the nuclear properties of a material proposed for use in reactor experiments. Keutron activation analysis shows ai1 excellent sensitivity for all of these elements with the exception of boron, and i t is therefore frequently employed for these analyses. We have deyeloped a n analytical scheme for the above elements based on the following criteria: Most of the elements with an important influence on nuclear properties should be determined with a suitable sensitivity and precision; elements of minor significance should also be determined at least semiquantitatirely, if present at unusually high levels; the operation time should be kept sufficiently low t o permit the analysis of many samples in a reasonable time. Automation of the chemical separations and instrumental analysis should be used, whenever possible. Sequential separation schemes on many materials analyzed b y neutron activation are available from the literature. Blaedel, Olsen, and Buchanan (4)
have studied on synthetic mixtures of radioisotopes the possibilities of cation exchange for a group separation of metallic radioelements. The procedure is very rapid and it may be easily automated, but unfortunately for our purpose the greatest part of the elements of interest for reactor technology fall within one group. Separation schemes on a large number of elements h a r e been studied in neutron activation analysis of semiconductors (21) and water ( 5 ) . Other schemes have been studied for the separation of individual fission products in health physics research (6, 8). Also in these cases the range of elements studied does not cover many of the elements of interest in the nuclear field. Albert has developed the systematic analysis by neutron activation of high purity metals (aluminum, iron, zirconium) ; 49 elements are determined in a single sample. The first procedures were rather lengthy ( 2 ) ,but in his latest work only 2 working days are necessary for such a complete analysis ( 1 ) . Jervis and Mackintosh (11) have used ion exchange for determining impurities in reactor grade aluminum. Mackintosh (14) used radiochemical separations to study the distribution of impurities in the purification of aluminum by zone melting. Since none of the above separation schemes met our requirements, particularly that of rapidity, we have developed a new scheme. The elements determined with our procedure are reported in Table I. The procedure covers the range of gamma-emitting isotopes with half lives greater than approximately 1 day. VOL. 35, NO. 2, FEBRUARY 1963
173
