Normal-phase high-performance liquid chromatographic determination of phenols

Talanta, Vol. 36, No. 5, pp. 513-519, 1989 Printed in Great Britain. All rights reserved

0039-9140/89$3.00+ 0.00 Copyright 0 1989 Pergamon Press plc

NORMAL-PHASE HIGH-PERFORMANCE LIQUID CHROMATOGRAPHIC DETERMINATION OF PHENOLS S. N. LANIN and Yu. S. NIKITIN Department of Chemistry, Moscow State University, 119899 Moscow, USSR (Received 27 May 1987. Revised 23 September 1987. Accepted 17 November 1988)

Summary-Normal-phase high-performance liquid chromatography has been used for separation of phenol and its monoderivatives. Multi-component mixtures of hexane (non-polar component) with butan-l-01, chloroform, butyl bromide, butyl chloride or diethyl ether (polar additives) were used as selective eluents. Silica gel “Silasorb 600” with specific surface area of about 600 m*/g and average particle size of u 10 pm was used as the sorbent. Phenol and the o-, m-, p-isomers of cresol were concentrated by extraction with n-butyl acetate from aqueous solutions. A method for determination of microamounts of phenols in aqueous solutions in the presence of 160-fold amounts of aromatic hydrocarbons has been developed.

Phenol and its derivatives are among the most toxic and widely spread pollutants in industrial effluents and natural waters. Current methods (mostly spectrophotometrk-‘) usually fail to determine phenols without a preliminary separation. Unfortunately, the preliminary separation considerably complicates the analysis and does not always ensure accurate results, especially for isomers. Existing techniques determine only the total content of phenols in effluents and natural waters, and in many cases this approach is hardly satisfactory. It is often extremely important to identify individual pollutants and to determine individual concentrations, as the permissible concentration limits for various phenols and their isomers significantly differ, sometimes by factors of tens and even hundreds.4 High-performance liquid chromatography (HPLC) is especially advantageous for separation of organic mixtures, e.g., of phenols by either the normal-phase (polar sorbent, non-polar eluent) or reversed-phase (non-polar sorbent, polar eluent) technique.r-” In the present investigation normal-phase HPLC was applied. It is characterized by higher selectivity in separating O-, I?I- and p-isomers of organic compounds’4*‘s and better reproducibility of the surface adsorption properties for silica gels than for “bonded” phases (and the former are considerably cheaper than the latter). As the content of phenol derivatives in effluents and especially in natural waters is often rather low, preconcentration by adsorption or extraction is recommended. The latter has the advantage of transferring the phenols from the aqueous to the organic phase, as required for normal-phase HPLC.

EXPERIMENTAL Apparatus

A “Tswett-306” liquid chromatograph equipped with a spectrophotometric detector (2 200-700 nm) and a piston

pump giving eluent flow-rates in the range OS-5 (kO.1) ml/min was used. Stainless-steel columns (200 x 6 mm and 300 x 6 mm) were slurry-packed, using the suspension procedure, with silica gel Silasorb 600 (“Lachema”, Czechoslovakia), (specific surface area about 600 m’/g and average particle size -10 pm). The column efficiency in terms of p-nitrophenol, was ~5000 theoretical plates. A “Milichrom” microcolumn liquid chromatograph equipped with a spectrophotometric detector (1190-360 nm) and piston pump (2500 pl capacity) for eluent flow-rates in the range 2-600 pl/min was also used. Stainless-steel columns (120 x 2 mm) were slurry-packed with Silasorb 600 (-600 m*/g specific surface area, average particle size -6 pm). The column efficiency in terms of p-cresol was 9000-10000 theoretical plates.

Reagents

Samples were introduced into the chromatograph with an MS-10 microsyringe. Two-component mixtures (from 99:l to 30 : 70 v/v) of hexane or heptane (non-polar component) with butan-l-01, chloroform, butyl bromide, butyl chloride or diethyl ether (polar additives) were used as eluents. Phenol, and its o-, m- and p-substituted methyl, chloro, bromo, iodo, nitro, amino and hydroxyl derivatives were used as test materials. Phenol and m-cresol (pure) were distilled under reduced pressure (79-W/18 mmHg and 92-93”/12 mmHg, respectively). The purity of the substances was verified chromatographically and by refractometry. Absorption spectra of the phenol solutions in the eluents and in butyl acetate were recorded with a Hitachi-124 spectrophotometer. Procedure

A stable solution of phenol (205 pg/ml) was preparedI by dissolving 0.0517 g of distilled phenol and 0.001 g of maleic acid in 96% ethanol and making up to volumes with the same solvent in a 250-ml standard flask. Phenol solutions in distilled water and in n-butyl acetate were prepared for chromatographic analysis by dissolving weighed amounts of phenol in the appropriate volume of solvent. The solutions were further diluted as required, for preparation of the calibration graph. The phenols were extracted into n-butyl acetate by a salting-out procedure in which 45 g of anhydrous sodium sulphate and 3 ml of n-butyl acetate were added to 250 ml of aqueous phenol solution and the mixture was shaken for 15-20 min, until equilibrium was attained. 573

S. N. LANZNand Yu. S. NIKE~N

574

stance (CC&) (ml) and VRsp,,the retention volume of phenol (ml). It follows from Tables 1 and 2 that the retention of phenols si~fi~ntly depends on both the nature and the position of the substituent. The retention of halogen- and methyl-substituted phenols is the weakest; it conside~bly increases with change of substituent, in the series Cl~CH,
REWL’I’S AND DRKW3SION

Phenols are polar compounds that are strongly adsorbed on silica gels. Consequently, in normal-phase liquid chromato~aphy phenols should have long retention times, and analysis of phenol-containing samples is slow, which is regarded as extremely undesirable in modern analysis. To reduce the retention time, a polar solvent (butan-l-01) with high elution strength (e” = 0.68) was added to a non-polar eluent (hexane) with low elution strength (6’ = O.Ol), to give a greater or smaller (depending on the concentration of the polar additive) deactivation of the surface silanol groups in the sorbent. Even a quite low butan-l-01 concentration considerably shortens the retention times and lowers the capacity factors of the phenols (Table 1). At low butan-I-ol con~n~ations in the eluent (,<2.5% V/Y) the retention times sharply increase, especially For the most polar phenol derivatives (amino-, hydroxy- and nitrophenols), and the chromatographic analysis becomes too complicated. The relative retentions of these substances (elution with 12.5% v/v butan-l-01 in hexane) are presented in Table 2. The selectivity (a), i.e., the retention relative to phenol is defined as a=

Vii-

vo

(1)

VR,ph- v,’

where VR is the retention volme of the test substance (ml), V, the retention volume of a non-sorbed sub-

Table 1. The corrected retention times (t;(, min)* and capacity factors (k’)’ of phenols on silica gel, as a function of butan-l-01 content in hexane (column 300 x 0.6 cm, eluent flow-rate (F) 1.34 ml/mint Butan-l-01 content, % u/u 1.0 Substance Phenol o-Cresol m -Cresol p-Cresol m-Cblorophenol p-Chlorophenoi

1.5 k’

rk 12.0 8.6 12.0 12.5 9.2 12.7

tk 8.1

2.93 2.10 2.93 3.05 2.24 3.10

z 81 ;4

2.0 k’ 2.31 1.54 1.86 2.31 1.83 2.48

G 5.4 4.0 4.9 5.3 4.2 5.7

2.5 k’

t;

k’

1.54 1.14 1.40 1.51 1.20 1.63

4.5 3.4 4.0 4.5 3.3 4.4

1.28 0.91 1.14 1.28 0.94 1.26

*t& = ta - to; k’ = (tR- tO)/rO,where ta is the retention time of the analyte, t, the retention time of a non-sorbed substance (Ccl,). tFor 1% v/v butan-l-01 content in hexane, F = 1.14 ml/min.

Table 2. Effect of the nature and position of a phenol substituent on the relative retention (a) of phenol derivatives, eluent 12.5% v/v butan-l-01 in hexane fi = dipole moment, Debye)

Substituent :IEL NO;

a

:z 0:69 0.71 2.26 23.2

para -

meta-

orthoa

a .

P 1.43

0.68

Ir 2.17

0.94 0.83

2.68 2.21

1.36 ::: 2.58 1.86

0.93 0.92 3.10 50.2

:*z 1:53 2.19

0.99 I .26 4.88 124.2

I.64 5.05 0

lr

HPLC determination of phenols

I

I

I

I

0

3

6

9

t,(mirJ

Fig. 1. Chromatogram of model mixture of benzene (1) and monosubstituted phenols: o-nitrophenol(2), o-chlorophenol (3), ocresol(4). Column Silasorb 600, 10 pm, 200 x 6 mm; mobile phase 2.5 v/v butan-l-01 in n-hexane; 25°C; spectrophotometric detector (254 nm).

575

Eluents containing methylene chloride (one of the most widely used mobile phase modifiers in normalphase chromatography) have lower selectivity for cresols (Fig. 5) than chloroform-hexane eluents do (Fig. 3). The nature and composition of the eluent can also considerably affect the ultraviolet-detection of the separated phenols. The eluent composition should be selected according to the absorption bands of the eluent and UV sample components, to ensure that the detector wavelengths correspond to the maximum absorbance for the sample components. In this way, the sensitivity of detection can be considerably increased. Moreover, detection at more than one wavelength makes it possible to identify the components of the mixture from the absorbance ratios. This is well illustrated by the LC separation of a phenolcresol mixture (Fig. 6). Analysis of such mixture is rather complicated when other analytical methods are

(Figs. 2, 3) in acceptable lengths of time. When the

sample contains phenols considerably differing in polarity and, consequently, in retention time, the method of stepwise gradient elution (Fig. 4) may be successfully applied. It significantly extends the possibilities for separating complex mixtures with instruments not provided with a gradient elution system. Not only the elution time, but also the selectivity and resolution may be controlled by changing the composition of the eluent. Eluents capable of separating mixtures of phenol with cresol isomers are given in Table 3, which makes it easy to choose the most suitable eluent, depending on whether all the components of the mixture are to be determined or only some of them (Fig. 3).

3

1

0

4

6

;

12

16

t,(mln)

, 0

6

10

I

I

15

20

t,(min) Fig. 2. Chromatogram of model mixture of nitrophenols and phenol: (1) o-Nitrophenol. (2) m-nitrophenol, (3) phenol, (4) p-nitrophenol. Column Silasorb 600, 10 pm, 200 x 6 mm; mobile phase 1% v/v propan-2-01 and 6% v/v diethyl ether in n-hexane; 25°C; spectrophotometric detector (270 nm).

Fig. 3. Chromatograms of model mixture of cresols and phenol: (1) o-cresol, (2) mcresol, (3) p-cresol, (4) phenol. Column Silasorb 600, 6 pm, 120 x 2 mm; mobile phase 40% v/v chloroform in n-hexane; 25°C; spectrophotometric detector (270 nm).

576

S. N. LANINand Yu. S. NIIUT~N 2

1

1

I

I

I

0

6

12

16

I

I

I

24

30

36

t,(min)

Fig. 4. Chromatogram of n-butyl acetate extract of phenol from aqueous solution. (1) n-butyl acetate, (2) phenol, (3) o-aminophenol, (4) m-aminophenol. Cohmm Silasorb 600, 10 pm, 200 x 6 mm; 25°C; spectrophotometric detector (270 nm). Stepwise elution: n-hexane-butan-l-01 (97.5: 2..5), after 12 mm (85: 15).

Besides the chromatographic separation itself, the preliminary extraction is another important stage in the determination of phenols in aqueous solutions, since it brings the phenols into the organic phase needed for injection for normal-phase chromatography and also provides preconcentration. The extraction should also eliminate possible interferences from any hydrocarbons present in samples. The extractant should provide high distribution coefficients for the analytes, have good solubility in the eluent, a retention time that is both minimal and different from those of the analytes, and minimum absorptivity at the detection wavelength. From the literature data,*gv20n-butyl acetate was selected as the extractant. Phenols often have to be determined in the presence

of large quantities of hydrocarbons, e.g., aromatic hydrocarbons in effluents from coke plants. To study this problem, a model phenol solution (5 pg/ml) was prepared in water saturated with benzene and naphthalene. At 20” the solubilities of benzene and naphthalene in water are 0.82 and 0.03 mg/ml respectively, so their total concentration in the solution was about 160 times that of the phenol. The degree of extraction was increased by use of anhydrous sodium sulphate (80 g/l.) as salting-out agent. The chromatographic data for the extract are presented in Table 4. Figure 4 and Table 4 demonstrate that phenol and its derivatives, extracted by n-butyl acetate from aqueous solutions, can be separated chromatographically and if the eluent composition is selected properly, the extractant and a large excess of aromatic hydro-

Table 3. Selectivity of phenol and cresol isomer separation (er) as a function of eluent composition (second component hexane) u

Polar additive to eluent,

% v/v

Instrument

m-/o-

p-lm-

CHCI,, 15 i-C, H,OH, I + C,HsBr, 6 I-CrH,OH, 1 + CHCI,, 6 i-C,H,OH, 1 CHCI,, 40 CHCl,, 50 CHCl,, 60

Tswett-306

1.75

1.07

phlp1.08

1.42

1.19

1.02

1.36 1.29 1.60 1.60 1.76

1.05

1.12 1.04 1.08 1.08 1.09

Milichrom

1.06 1.05 1.05 1.05

571

HPLC determination of phenols Table 4. Retention times (I~), corrected retention volumes (Vi) and capacity factors (k’) for the extractant, aromatic hydrocarbons and phenol on silica gel (column 200 x 0.6 cm, eluent 2.5% u/o butan-l-01 in hexane, flow-rate 1.4 ml/mm)

Substance

Retention time (ra), set

Ccl, (non-sorbed) Benzene Naphthalene n-Butyl acetate Phenol

Corrected retention time r&=ta-tO,

165 168 168 168 360

carbons do not interfere, since they have considerably smaller retention volumes and are eluted in one peak at the beginning of the chromatogram. The interference of aromatic hydrocarbons can

SIX

Corrected retention volume, V&=tk F, ml

Capacity factor, k’ = kn - t,Mn

0 0.07 0.07 0.07 4.55

0 0.02 0.02 0.02 1.18

0 3 3 3 195

also be eliminated by selecting the optimal detection wavelength. This is especially important for the determination of weakly retained phenol derivatives with retention times close to those of aromatic hydrocarbons. The resolution of their peaks may be rather bad (c$ o-nitrophenol and benzene with detection at 254 nm, in Fig. 1). When the detection wavelength is in the region for maximum phenol absorptivity (270 nm) the absorptivity of both n-butyl acetate and benzene is very low and their peaks almost disappear. This allows, for example, determination of o-nitrophenol in the presence of benzene (Fig. 7). The determination of o-cresol in the presence of phenol and m- and p-cresols may be taken as an example of HPLC analysis of aqueous solutions of phenols. The peak height (h, mm), is found to depend on the quantity of the o-cresol (M, pg). A typical calibration equation is

h = (49 f 5.O)M + (10 f 2.7)

(2)

for M = OS-3 pg. o-Cresol may only be determined if the distribution coefficient, D, and the degree of extraction, R, are known: D = C,,&

(3)

where C,, and C, are the concentrations of the substance in the organic and aqueous phases, respectively, and

where V, and V, are the volumes of the organic and aqueous phases, respectively. Table 5 lists the D and Table 5. Distribution coefficient (D) and degree of extraction (R) of ocresol between n-butyl acetate and water (saltingout with 80 g/l. Na,SO,)

1 15

0-

tatmin) Fig. 5. Chromatogram of model mixture of cresols and phenol: (1) otresol, (2) m-cm-sol, (3) p-cresol, (4) phenol. Column Silasorb 600, 6 pm, 120 x 2 mm; mobile phase 70% v/v methylene chloride in n-hexane; 25°C; spectrophotometric detector (270 nm).

Concentration of ocresol in water, rcalml 0.264 0.264 0.240 0.251 0.229

Concentration of o-cresol in n-butyl acetate, ux Iml 48 48 50 49 51

D

R, %

182 182 208 195 223

67.1 67.1 69.1 68.6 71.4

S. N. LANIN

578

(a)

0

and Yu. S. NIKITIN

lb)

15

10

5

20

0

t.(min)

(cl

10

5

15

t,(min)

0

10

5

15

t,(min)

Fig. 6. Chromatograms of a model mixture of cresols and phenol (1) o-cresol, (2) m- and p-cresols, (3) phenol. Column Silasorb 600,6 pm, 120 x 2 mm; mobile phase 1% v/v propan-2-01 in n-hexane; 29C; spectrophotometric detector: (a) 210, (b) 218, (c) 270 nm.

!

I I

1 n

I

I

I

0

3

6

9

t,(min)

Fig. 7. Chromatogram of model mixture of benzene and monosubstituted phenols: (1) benzene, (2) o-nitrophenol, (3) o-chlorophenol, (4) ocresol. Column Silasorb 600,lO pm, 200 x 6 mm; mobile phase 2.5% v/v butan-l-ol in n-hexane; 2ST; spectrophotometric detector (270 nm).

I

I

0

15

I

30

tR (min)

Fig. 8. Chromatogram of n-butyl acetate extract of phenol and cresols from aqueous solution: (1) ocresol, (2) m- and ptresol, (3) phenol. Column Silasorb 600, 10 pm, 200 x 6 mm; mobile phase 15% v/v chloroform in n-heptane; 25°C; Spectrophotometric detector (270 nm).

HPLC determination of phenols R values for extraction of otresol from aqueous solutions with n-butyl acetate and salting-out with sodium sulphate. A model solution of phenol, o-, mand p-cresol (each 1 pg/ml) was prepared by mixing. The ocresol. 2.55 ml of o-cresol solution (98 pg/ ml), 1.66 ml of m-cresol solution (150 pg/ml), 2.5 ml of p -cresol solution (100 pg/ml) and 1.25 ml of phenol (200 pg/ml), and diluting accurately to 250 ml. Then 20 g of anhydrous sodium sulphate were added to 200 ml of the mixture, and this solution was shaken with 3 ml of n-butyl acetate for 20 min. Finally 10 ~1 of the extract were injected by microsyringe into the chromatographic column. The chromatogram of the extract is given in Fig. 8. In this case we were mainly interested in the determination of o-cresol, so the eluent composition was selected (Table 3) to give maximum selectivity (c+, = 1.75) to separate the o-cresol from the m-cresol (the species eluted nearest to it). The concentration of o-cresol in the model mixture (C,) was calculated from the peak height and the equations: M

cm=-

v,

v s ( v+o D

1 >

or C,=d

1Ooiu v VsV,R

where Vs is the volume of the injected sample. The value obtained was 1.Ol pg/ml o-cresol, with a relative standard deviation of 7% (n = 3, range method). The results show that the proposed method of extractive preconcentration followed by LC analysis is fast and reliable for the determination of phenols and their isomers in aqueous solutions,

519 REFERENCES

1. Standardized Methodr of Water Analysis, Yu. Yu. Lurie (ed.), (in Russian), Khimiya, Moscow, 1973. 2. Yu. Yu. Lurie, Analytical Chemistry of Industrial Efpuents (in Russian), Khimiya, Moscow, 1984. 3. V. Leite, Determination of Organic Pollutants in Drinking Waters, Natural Waters and EJJIuents (in Russian), Khimiya, Moscow, 1975. 4. M. M. Senyavin, M. Ya. Belousova, M. S. Safronova, T. V. Avgul and T. G. Shlepnina, Zh. Analit. Khim., 1980,35, 1224. 5. K. Callmer, L. E. Edholm and B. E. F. Smith, J. Chromatog. 1977, 136,45. 6. J. F. Schabron, R. J. Hurtubise and H. F. Silver, Anal. Chem., 1978,50, 1911. 7. C. A. Chang and Cheng-Fan Tu, ibid., 1982,54, 1179. 8. E. Tesarova and V. Padkovl. Chromatoarauhia. 1983. 17, 269. 9. L. S. Ronald and J. P. J. Donald, J. Ckromatog. Sci., 1983, 21, 282. 10. H. A. McLeod and G. Laver, J. Ckromatog., 1982,244, 385. 11. C. A. Chang, Q. Wu and D. W. Armstrong, ibid., 1986, 354,454. 12. E. Nieminen and P. Heikkila, ibid., 1986, 360, 271. 13. M. Yoshikawa. Y. Taauchi. K. Arashidani and Y. Kodama, ibid., 1986, %2,425. 14. L. R. Snyder, in High-Performance Liquid Chromatography, Advances and Perspectives, S. Horvath @xl.), Vol. 1, p. 208. Academic Press, New York, 1980. 15. A. V. Kiselev, A. A. Aratskova, T. N. Gvozdovitch and Ya. I. Yashin, J. Chromatog., 1980, 195,205. 16. A. I. Belen’kaya and E. F. Goritskaya, K/rim. i Technol. Vody, 1983, 5, 233. 17. Ya. I. Korenman. E. M. Tishenko and T. A. Nefedova. Zh. Analit. Skim, 1982, 37, 911. 18. Ya. I. Korenman, N. G. Sotnikova, T. A. Nefedova and R. N. Bortnikova, ibid., 1984, 39, 2081. 19. Ya. I. Korenman. Extraction of Phenols (in Russian). ” p. 20. Volgo-Vyatski, Gorky, 1973. . 20. Idem, Distribution Constants of Organic Substances between two Liquid Phases (in Russian), No. 5, p. 13. Gorky, 1979. I_