lnrrrnarionai Biodeterioraiion & Riodegradrrrion(1996) 277-282 0
ELSEVIER
PII:
SO964-8305(96)00062-5
1997 Published by Elsevier Science Limited Printed in Great Britain. All rights reserved 0964-8305196 $15.00 + 0.00
Problems of Identifying Phenolic Compounds During the Microbial Degradation of Olive Mill Wastewater G. Knupp,“” G. Riicker,b A. Ramos-Cormenzanat S. Garrido Ho~os,~ M. Neugebaue# T. Ossenkopb
&
“Fachbereich fir Bauingenieurwesen, Labor fir Abfallwirtschaft, Siedlungswasserwirtschaft, Umweltchemie (LASU) , Fachhochschule Mtinster, D-48149 Mtinster, Germany ‘Pharmazeutisches lnstitut Poppelsdorf, Rheinische Friedrich- Wilhelms- Universittit Bonn, Kreuzbergweg 26, D-531 15 Bonn, Germany ‘Departamento de Microbiologia, Facultad de Farmacia, Universidad de Granada, 18071 Granada, Spain dDepartamento de Ingenieria Quimica, Facultad de Ciencias, Universidad de Granada, 18071 Granada, Spain
The main objectives of the presented Spanish-German collaboration are the purification of alpechin by biodegrading phenolic compounds and the investigation of metabolites during fermentation prior to its safe disposal. In addition to 12 well-known compounds, 3,4-dihydroxyphenylglycol was also identified in untreated Spanish and Italian alpechin samples using a GC/MS method. The qualitative composition of the Italian and Spanish samples differ. First results of degradation tests of reference substances are reported: Arthrobacter is capable of completely transforming added tyrosol to the corresponding 4-hydroxyphenylacetic acid and after 139h of fermentation no traces of tyrosol can be identified. In contrast, only traces of phenylacetic acid are produced by Bacillus pumilus after 139h of fermentation of tyrosol. 0 1997 Published by Elsevier Science Limited. All rights reserved
chemical and biological methods have been tested to purify the wastewater. A possible method of biotechnological purification through microbiological degradation of the phenolic compounds is being tested among others at the Departamento de Microbiologia, Granada in collaboration with the Fachhochschule Miinster and the Universitat Bonn. The main objective of this collaboration is the identification of might metabolites that occur during the fermentation to optimize the system and to make sure that no toxic intermediates are created. The efforts towards the identification of phenolic compounds in olive mill wastewater and first results of their possible biodegradation are reported here. For the purpose mentioned above the following investigations are planned and on-going. First, defined reference substances+affeic acid, tyrosol and 4-hydroxybenzoic acid-as selected by the Spanish partners, will be subjected to biodegradation in appropriate culture media, one by one, in order to get familiar with the degradation products to be expected. With more information about the fate of an individual type
Olive mill wastewater is produced in large quantities during the processing of olive oil. The Spanish generic term for this wastewater is alpechin. Alpechin causes severe problems in the environment due to its high chemical and and its biochemical oxygen demand antimicrobial and phytotoxic effects. These wellknown ecologically harmful properties are assigned to phenolic compounds such as aromatic acids, alcohols and aldehydes present in alpechin (Moreno et al., 1987; Capasso et al., 1992). An overview of the phenolic compounds as described in the literature by several authors (Fedeli, 1977; Balice & Cera, 1984; Martinez Nieto et al., 1992) and are given in Fig. 1. Many of them e.g. caffeic acid, tyrosol and 4hydroxybenzoic acid are well known precursors and products in the pharmaceutical, chemical and food industries. The undesirable aspects of the alpechin make the direct waste disposal or the reutilization of the rich organic and inorganic content difficult. Different physico*New address: Fachochschule Rhein-sieg, FB Chemie und Werkstofftechnik, D-53754, Sankt, Augustin. 277
G. Knupp et al.
278 CCQH
H
-i-)’
HO-j=&
0
H”~cooH
R
Brenzcatechol 4-Hydroxyphenylacetic acid 3.4-Dihydroxyphenylacetic acid
R=H R = OH
R=H R=OH R = OCHs
Cumaric acid Caffeic acid Ferulic acid
OH H
HO
COOH
R RI =H Rj=OH [ RI = OH
R2 = H
Tyrosol
R2=H R2 = OH
Hydroxytyrosol 3,4-Dihydroxyphenylglycol ]
Fig. 1. Phenolic
compounds
of alpechin
of phenol, it is planned to design the system in a more complex manner to investigate interactions in particular, using synthetic alpechin as a culture medium. In a further step, phenol-free alpechin, mixed with three reference substances will be The long-range subjected to fermentation. objective will then be the identification of degradation products in natural alpechin after fermentation. For this purpose, at each step of the investigations, samples will be taken from the fermentation broth and analysed. A basic requirement for all these investigations is the development of a suitable analytical method for the rapid, sensitive and unequivocal identification of the phenolic compounds. In the literature, several methods have been described, e.g. TLC (Vazquez Roncero et al., 1974; Capasso et al.,
Oily residue: phenolic
R=H R=OH
4-Hydroxybenzoic acid Protocatechuic acid
R = OCH~ Vanillic acid
described
in the literature.
1994), HPLC (Martinez Nieto et al., 1992), and GC (Hamdi et al., 1992; Balice & Cera, 1984) or GC/MS (Lopez Aparicio et al., 1977). In the following, the GC/MS method used at Bonn and the difficulties that might occur when analysing untreated alpechin will be reported. Figure 2 shows a flowchart of this method. alpechin-adjusted to The pH4.5 with hydrochloric acid and centrifuged if necessary-is immediately extracted with ethyl acetate. After evaporating the organic fraction to dryness an oily residue is obtained. As the phenolic compounds themselves are too polar to be separated by gaschromatography, a suitable derivatization is necessary to enhance volatility and thermal stability. This is carried out using N-methyl-N-trimethylsilyltrifluoracetamide (MSTFA), a reagent which
I
Water residue
t Derivatization: MSTFA 85’C I5 min.
IrnD
k Compound
1
GCiMS
Fig. 2. Flowchart
of the analytical
method.
ident!fying phenolic compounds during alpechin microbial degradaation
transforms both hydroxylic and carboxylic functional groups into the corresponding TMSethers and -esters. It has been proven with the help of reference substances that no intermediate products due to non-quantitative reactions are produced using this reagent. The derivatized residue is then separated by capillary-CC and the corresponding mass spectra are recorded. When analysing alpechin in this way, a first idea of the compound identity can be obtained by comparing times in the retention gas chromatogram with those of reference substances. Then, the identities of the phenolic compounds are confirmed by comparing the mass spectra with those of authentic reference substances. This procedure is necessary because there are some cases described in the literature where retention times of compounds from alpechin correlate perfectly with those of reference substances although they are not identical (Lopez Aparicio et 1977). By using combined GC/MS, al., misinterpretations can be avoided. With this method, extracts of untreated alpechin from Spain were analysed. The Spanish alpechin sample was obtained from the Granada area. In addition, a sample of Italian alpechin from Borgomaro in Liguria was analysed. Both samples were kept frozen at -30°C until further some examination. Table 1 gives more information about these two samples. It can be seen that the Spanish alpechin has a chemical oxygen demand, a biochemical oxygen demand all of which are and dry matter content approximately double that of the Italian alpechin. So far, 12 phenolic compounds in alpechin from both origins have been identified using the GC/ MS method. Fig. 3 shows two typical gasof ethyl acetate extracts of chromatograms untreated alpechin, prepared in the same manner. Peak no. 1 in the chromatograms of Italian and Spanish alpechin was identified as the TMSTable 1. Spanish Sample-Important
COD (gl-I)“ BODs (gl-‘)d Dry matter (gl-‘) pH (24°C)
Alpechin Sample Parameters
vs
Italian
Alpechin
Spanish alpechin”
Italian alpechin’
49.0 4.2 35.1 4.9
80.4 11.5 73.0 5.2
“Origin: Area of Granada. bOrigin: Borgomaro, Liguria. ‘Chemical oxygen demand. “Biochemical oxygen demand.
279
derivative of benzoic acid. The intense peak no. 2 in the lower chromatogram belongs to brenzcatechol. Although the chromatogram of Spanish alpechin presents an intensive signal at the same scan, the corresponding mass-spectra revealed that this compound is not identical to brenzcatechol. Brenzcatechol was identified in low quantities in the peak’s tailing (signal no. 2). By comparing the massspectra of the TMS-derivatives, peak no. 3 in both chromatograms was identified as tyrosol. This substance is described in the literature as one of the major compounds (Capasso et al., 1994). Peaks Nos 4 and 5 in the lower chromatogram are the signals of derivatives of 4-hydroxybenzoic acid and 4hydroxyphenylacetic acid, respectively. These two compounds, clearly identified in the extracts of Italian alpechin could only be found in trace amounts in the Spanish alpechin. According to the retention time from a mixture of derivatized authentic reference substances, peak no. 6 of both chromatograms correlates with the derivative of veratric acid. However, comparison of the corresponding mass spectra revealed that the substance in these two chromatograms is not identical to veratric acid. Although veratric acid is described in the literature as a typical alpechin compound (Balice & Cera, 1984) it has not yet been identified in the analysed samples. In the leading edge of the dominant peak around scan no. 450 two further well-known compounds were identified in both extracts by their mass spectra: vanillic acid (peak no. 7) and hydroxytyrosol (peak no. 8). Peak no. 9 in the lower chromatogram belongs to the TMS-derivative of protocatechuic acid. Although the retention time of the corresponding intense peak in the upper chromatogram correlates with protocatechuic acid, the mass spectra are not identical. In contrast to being present in great amounts in extracts of Italian alpechin, only traces of protocatechuic acid were found in Spanish alpechin 9. The presence of 3,4at peak no. dihydroxyphenylglycol-a well known compound in olive fruits (Bianchi & Pozzi, 1994j-has not yet been described in alpechin. The TMS-derivative of this labile substance could be identified as peak no. 10. There are three other signals in the chromatograms worth mentioning: signal no. 11 in the lower chromatogram of Italian extract represents the TMS derivative of 3,4dihydroxyphenylacetic acid. No corresponding signal can be found in the upper chromatogram. Conversely phenylpropanoic compounds have been identified only in the Spanish alpechin so far. In the upper chromatogram the corresponding signals of
G. Knupp
280
et al.
100 1
100 1:45
Spanish sample
300 5:15
200 3:30
I
400 7:oo
500 6:45
600 10:30
500 6~45
6dO 10:30
700 12:15
600 Scans 14:00 Time [min]
Italian sample
8 \
I . 100 1:45
200 3:30
300 515
400 7:oo
Fig. 3. Chromatograms
the TMS-derivatives of cumaric acid (no. 12) and of caffeic acid (no. 13) can be found. A possible explanation for their absence in Italian alpechin might be a partial natural biodegradation of these compounds occurring prior to sampling. /IOxidation and/or m-hydroxylation could create protocatechuic acid. The strong intensity of the protocatechuic acid’s peak, as shown before in the would agree with this lower chromatogram, observation. In the following discussion, two mass spectra of tyrosol-TMS are given as an example of unequivocal identification. On the left side of Fig. 4 the TMS-derivative of a tyrosol reference
3 1, 15
I
600 Scans 14:OO Time [min]
of alpechin extracts.
substance and on the right side the analysed peak at scan no. 328 of Spanish alpechin are shown. After EI ionization the molecular ion occurs at m/z 282. The next peak at m/z 267 results from the loss of a methyl group. This typical M- 15 fragment ion is prominent in all mass spectra of trimethylsilylated alcohols and acids and can be used to detect the molecular ion. a-Cleavage creates the base peak at m/z 179 which belongs to a cyclic tropylium-TMS ether cation. The signals at m/z 73 and 103 are created by the TMS group. On the right, the mass spectrum of tyrosol recorded at scan no. 328 from the Spanish alpechin extract can be found.
281
Identifying phenolic compounds during alpechin microbial degradation Reference
Compound
substance
at peak # 3
179
179
73
262 267
1 I
IL L
SO
lb0
260
150
50
300
250
!
100
150
Fig. 4. EI/MS
of silylated
Ir
532
I300
lmouritv
48
24 h
HO~jCDOH
J
372
f
h
600 10:30
600 14:00
72
Scan Time [mini
10:30
h
I.;.bosca” 400 7:oo
200 3:30
10:30
14:o0
Time [min]
139 h
h
ho
i7:oo
.
250
tyrosol.
330 “0~“”
3:30
200
,
fermented under usual conditions. Four samples were taken after 24,48, 72 and 139h of fermentation and extracted immediately. Figure 5 shows four chromatograms monitoring this fermentation over 139h. With the GC/MS method described above it can be shown that Arthrobacter is capable of oxidizing the added tyrosol to the corresponding 4hydroxyphenylacetic acid. Although no quantitative
As the two spectra show good correlation within their fragmentation patterns, the compound identity could be confirmed. In the next step of the Spanish-German collaboration, tyrosol was subjected to fermentation by Arthrobucter for 139h. For this purpose tyrosol was added to a typical culture medium in a concentration of lgl-’ and then aerobically
400 7:oo
,“‘I
m/z
m/z
.L
-
14:00
Fig. 5. Biodegradation
400 7:oo
200 3:30
Time [min]
of tyrosol
over
139h.
600 10:30
scan 14:00
Time [min
282
G. Knupp et al.
measurements have been carried out to date it can be seen in the chromatograms that the peak at scan no. 330 belonging to tyrosol decreases while the corresponding signal of 4-hydroxyphenylacetic acid increases over the monitored time. In the fourth the chromatogram tyrosol has completely disappeared after 139h of fermentation. Now the only prominent peak at scan no. 374 belongs to 4hydroxyphenylacetic acid. Both, retention times and fragmentation patterns of the mass spectra match those of our authentic reference substances. In contrast, Bacillus pumilus was unable to completely biodegrade tyrosol over the monitored time. So far, next to the remaining tyrosol only traces of phenylacetic acid could be identified. Extracts of the fermentation broths of 4-hydroxybenzoic acid are currently being analysed.
REFERENCES Balice, V. and Cera, 0. (1984) Acidic phenolic fraction of the olive vegetation water determined by a GC method. Grasas y Aceites, 35, 178-180. Bianchi, G. and Pozzi, N. (1994) 3,4-Dihydroxyphenylglycol,
a
major
C&z
phenolic
in
Olea
europaea
fruits.
Phytochemistry, 35, 1335-1337.
Capasso, R., Cristinzino. G., Evidente, A. and Scognamiglio, F. (1992) Isolation, spectroscopy and selective phytotoxic effects of polyphenols from vegetable waste waters. Phytochemistry, 31, 4125-4128.
Capasso, R., Evidente, A. and Vista, C. (1994) Production of hydroxytyrosol from olive oil vegetation waters. Agrochimica, 38, 165-17 1. Fedeli, E. (1977) Lipids of olives. Prog. Chem. Fats Other Lipids, 15, 57-74.
Hamdi, M., Garcia, J. L. and Ellouz, R. (1992) Integrated biological process for olive mill wastewater treatment. Bioprocess Engng, 8, 79-84.
Lopez Aparicio, F. J., Garcia-Granados Lopez de Hierro, A. and Rodriguez Alonso, M. (1977) Estudio de1 contenido en Qcidos carboxilicos de1 alpechiin de la aceituna, y evolution de 10s mismos. Grasas y Aceites, 28, 393401.
Martinez Nieto, L., Ramos Cormenzana, A., Garcia Pareja, M. P. and Garrido Hoyos, S. E. (1992) Biodegradation de compuestos fenolicos de1 alpechin con Aspergillus terrreus. Grasas y Aceites, 43, 75-8 1. Moreno, E., Perez, J., Ramos-Cormenzana, A. and Martinez, J. (1987) Antimicrobial effect of waste water from olive oil extraction plants selecting soil bacteria after incubation with diluted waste. Mikrobios, 51, 169-174. Vazquez Roncero, A., Graciani Constante, E. and Maestro Duran, R. (1974) Componentes fenolicos de la accituna. I Polifenoles de la pulpa. Grasas y Accites, 25, 269-279.