- Email: [email protected]
Comparison of a conventional extractor with two unconventional extractors
AnalyticaChimicuActa, 125 (1981) 175-178 ElsevierScientific Publishing Company, Amsterdam - Printed in The Netherlands Short Communication
COMPARISON OF A CONVENTIONAL UNCONVENTIONAL EXTRACTORS
WALTER
A. ~a*,
PAL~A
EXTRACTOR
P. WICRRAMANAYAKE
WITH TWO
and J~~RGEN MUELLER
Department of Chemistry. Dalhousie Uniuersity. Halifax. Now Scotia (Canada) (Received 9th September 1980)
Summary.
Two specially-made extractors are compared against a
standard
Soxhlet
by
conditions and containing coarse and fine particles. The unconventional extractors, one with a fiied thimble and the other with a removable thimble, proved considerably faster than the Soxhlet.
using
test samples simulating adsorptive and non-adsorptive
Among continuous extractors [1] , the Soxhlet ranks high as an efficient and widely used laboratory tool. It has thezefore been chosen as a standard against which to compare two specially-made high-purity extractors. These units can be built in a wide range of configurations and capacities [ 21. The ones tested here are shown in Fig. 1. They are designed to maintain a solvent head [3] and quasi-chromatographic flow aided by considerable hydrostatic Pull* The relative speed of extractors depends not only on solvent flow but also on various physical characteristics of the sample. The model systems examined here contained adsorbed or non-adsorbed solute on mati-ices of low or high flow resistance. Other parameters such as intraparticle diffusion or swelling were not investigated.
Experimental The matrices were a high-resistance (200-mesh) silica gel 62 with a surface area of about 300 m2 g-l and an average pore diameter of about 135 A, and a low-resistance (70-230 mesh) silica gel 60 with 550 m2 g-’ and 60 A. The solute was quinoline; the two solvents were hexane and ethanol. Quinoline was determined by ultraviolet absorption. It was strongly retained on silica gel when cbromatographed with hexane, but migrated with the solvent front when developed with ethanol. The sample was’ prepared by dissolving a known amount of quinoline in the extraction solvent and depositing it onto a known amount of silica by rotary evaporation_ The conventional extractor used was a Soxblet (Kontes model K-585100, size 21, in which hot solvent vapors ascend around the thimble chamber), operated with specially-made or commercial (Fisher) 25 X 80 mm thimbles (both thimbles gave the same performance) and a conventional Kontes model K456500 size 21 reflux condenser. 0003-2670/81/0000-0000/$02.50
0 1981 Elsetier Scientific Publishing Company
176
26 CLEARSEAL
‘-OLDFWGER
HOLE FOR 24140
CLEARSEAL
24140
CLEARSEA~
Fig. 1. Extractor models: (A) “-Integral thimble” unit; (B) “kemovable thimble” unit.
With one exception, 500~ml flasks were used for all extractors. These flasks had a side-arm through which, by means of a stationary PTFE capillary with outside Luer-lock, samples could be withdrawn for measurements and returned with a syringe. (This sampling port replaced the regular sidearm for nitrogen introduction that is shown in Fig. 1.) The solvents were heated by “Thermowells” (Laboratory Craftsman) at the percent of maximum power listed in Table 1. The cross-sections (and inner diameters) of the sample compartments were as follows: Soxhlet thimble 4.9 cm* (2.5 cm), Soxhlet thimble container 7.1 (3.0); model A thimble 10.8 (3.7); model B thimble 2.5 (1.8), and model B thimble chamber 4.5 cm* (2.4 cm).
“A
Go 180 240
Extroctlon time (min)
123 300
s:
2 z
B g5
C
2c
40
GO
80
0
l-
Extroctlon
time (min)
GO
120
I-
0
L Extraction
ttme (min)
GO
121
Fig. 4. Extraction curves for a nonadsorptive system with high flow resistance: quinoline on 200.mesh silica gel 62, extracted by ethanol close to boiling point. Curves: (7) Soxhlet, 66W326 W; (8) extractor A, 40%/326 W. Other data in Table 1,
Fig. 3. Extraction curves for a nonadsorptive system with low flow resistance: quinoline on 70-230 mesh silica gel 60, extracted by ethanol close to boiling point. Curves: (6) Soxhlet, 66%/326 W; (6) extractor A, 66%/326 W, Other data in Table 1.
Fig, 2, Extraction curve8 for an adsorptive system with low flow resistance: quinoline on 70-230 mesh silica gel 60, extracted by hexane close to boiling point. Curves, extractor models and heat inputs: (1) Soxhlet, 30%/326 W; (2) model B, 30%/326 W; (3) model A, 30%/326 W; (4) model A, 100%/600 W. Other data in Table 1.
20
E: 4c b (1
it
80
100
z 4
1?8
TABLE 1 Extraction
conditions
Run numb&
1
Extractormodel Quicoline (mp) Silicagel (g) Particlemesh solvent Fiask (r=H Heat setting (56) 100% power
Soxhlet
F&_ no.
3
2
14
15 701230 Hexane 500 30d 325 2
n”
Ac
14
15 701230 Hexane 500 3@ 325 2
5
4
13
15 701230 Hexane 500 36 325 2
A
c 16
15 701230 Hexane 1000 1OOf 500 2
So&let 12
15 701230 Ethanol 500 55d 325 3
6
7
8
A= 11 14 701230 Ethanol 500 55 325
Soshlet 12 7 200 Ethanol 500 55d 325
A= 11 7 200 Etha 500 40e 325
3
4
aCorresponds to curve numbers in Figs. 2-4. bModel B is the extractor with removable thimble shown in Fig_ lB_ =Model A is the extractor with integral- thimble shown in Fig. IA. “Maximum for Soxhlet before solvent backs up in reflux condenser. eSmall amount of overflow. fNecessary to restrict flow with stopcock to maintain solvent head.
Results and discussion Table 1 lists the extraction conditions and Figs. 2-4 present the resulting extraction curves. The conditions had been chosen to provide a fair comparison of performance, to accommodate the different model systems, and to include a high-speed run. Such extraction curves represent the best way to express quantitatively the results of this and similar studies. Clearly, the two unconventional extractors evaluated here were considerably faster than the quite efficient Soxhlet. This research was supported by NSERC grant A-9604. REFERENCES 1 L. C. Craig and D. Craig, in A. Weissberger (Ed.), Technique of Organic Chemistry, __ __ -~Vol. III, Interscience, New York, 1956, p_ 149. 2 W. A. Aue, M. M. Daniewski, J. Miiller and J. P. Laba, Anal. C&em., 49 (1977) 1465. 3 C. E. Browne, W. L. Buchanan and E. J. Eisenbraun, Chem. Ind., 1977, p_ 35.
4
COMPARISON OF A CONVENTIONAL UNCONVENTIONAL EXTRACTORS
WALTER
A. ~a*,
PAL~A
EXTRACTOR
P. WICRRAMANAYAKE
WITH TWO
and J~~RGEN MUELLER
Department of Chemistry. Dalhousie Uniuersity. Halifax. Now Scotia (Canada) (Received 9th September 1980)
Summary.
Two specially-made extractors are compared against a
standard
Soxhlet
by
conditions and containing coarse and fine particles. The unconventional extractors, one with a fiied thimble and the other with a removable thimble, proved considerably faster than the Soxhlet.
using
test samples simulating adsorptive and non-adsorptive
Among continuous extractors [1] , the Soxhlet ranks high as an efficient and widely used laboratory tool. It has thezefore been chosen as a standard against which to compare two specially-made high-purity extractors. These units can be built in a wide range of configurations and capacities [ 21. The ones tested here are shown in Fig. 1. They are designed to maintain a solvent head [3] and quasi-chromatographic flow aided by considerable hydrostatic Pull* The relative speed of extractors depends not only on solvent flow but also on various physical characteristics of the sample. The model systems examined here contained adsorbed or non-adsorbed solute on mati-ices of low or high flow resistance. Other parameters such as intraparticle diffusion or swelling were not investigated.
Experimental The matrices were a high-resistance (200-mesh) silica gel 62 with a surface area of about 300 m2 g-l and an average pore diameter of about 135 A, and a low-resistance (70-230 mesh) silica gel 60 with 550 m2 g-’ and 60 A. The solute was quinoline; the two solvents were hexane and ethanol. Quinoline was determined by ultraviolet absorption. It was strongly retained on silica gel when cbromatographed with hexane, but migrated with the solvent front when developed with ethanol. The sample was’ prepared by dissolving a known amount of quinoline in the extraction solvent and depositing it onto a known amount of silica by rotary evaporation_ The conventional extractor used was a Soxblet (Kontes model K-585100, size 21, in which hot solvent vapors ascend around the thimble chamber), operated with specially-made or commercial (Fisher) 25 X 80 mm thimbles (both thimbles gave the same performance) and a conventional Kontes model K456500 size 21 reflux condenser. 0003-2670/81/0000-0000/$02.50
0 1981 Elsetier Scientific Publishing Company
176
26 CLEARSEAL
‘-OLDFWGER
HOLE FOR 24140
CLEARSEAL
24140
CLEARSEA~
Fig. 1. Extractor models: (A) “-Integral thimble” unit; (B) “kemovable thimble” unit.
With one exception, 500~ml flasks were used for all extractors. These flasks had a side-arm through which, by means of a stationary PTFE capillary with outside Luer-lock, samples could be withdrawn for measurements and returned with a syringe. (This sampling port replaced the regular sidearm for nitrogen introduction that is shown in Fig. 1.) The solvents were heated by “Thermowells” (Laboratory Craftsman) at the percent of maximum power listed in Table 1. The cross-sections (and inner diameters) of the sample compartments were as follows: Soxhlet thimble 4.9 cm* (2.5 cm), Soxhlet thimble container 7.1 (3.0); model A thimble 10.8 (3.7); model B thimble 2.5 (1.8), and model B thimble chamber 4.5 cm* (2.4 cm).
“A
Go 180 240
Extroctlon time (min)
123 300
s:
2 z
B g5
C
2c
40
GO
80
0
l-
Extroctlon
time (min)
GO
120
I-
0
L Extraction
ttme (min)
GO
121
Fig. 4. Extraction curves for a nonadsorptive system with high flow resistance: quinoline on 200.mesh silica gel 62, extracted by ethanol close to boiling point. Curves: (7) Soxhlet, 66W326 W; (8) extractor A, 40%/326 W. Other data in Table 1,
Fig. 3. Extraction curves for a nonadsorptive system with low flow resistance: quinoline on 70-230 mesh silica gel 60, extracted by ethanol close to boiling point. Curves: (6) Soxhlet, 66%/326 W; (6) extractor A, 66%/326 W, Other data in Table 1.
Fig, 2, Extraction curve8 for an adsorptive system with low flow resistance: quinoline on 70-230 mesh silica gel 60, extracted by hexane close to boiling point. Curves, extractor models and heat inputs: (1) Soxhlet, 30%/326 W; (2) model B, 30%/326 W; (3) model A, 30%/326 W; (4) model A, 100%/600 W. Other data in Table 1.
20
E: 4c b (1
it
80
100
z 4
1?8
TABLE 1 Extraction
conditions
Run numb&
1
Extractormodel Quicoline (mp) Silicagel (g) Particlemesh solvent Fiask (r=H Heat setting (56) 100% power
Soxhlet
F&_ no.
3
2
14
15 701230 Hexane 500 30d 325 2
n”
Ac
14
15 701230 Hexane 500 3@ 325 2
5
4
13
15 701230 Hexane 500 36 325 2
A
c 16
15 701230 Hexane 1000 1OOf 500 2
So&let 12
15 701230 Ethanol 500 55d 325 3
6
7
8
A= 11 14 701230 Ethanol 500 55 325
Soshlet 12 7 200 Ethanol 500 55d 325
A= 11 7 200 Etha 500 40e 325
3
4
aCorresponds to curve numbers in Figs. 2-4. bModel B is the extractor with removable thimble shown in Fig_ lB_ =Model A is the extractor with integral- thimble shown in Fig. IA. “Maximum for Soxhlet before solvent backs up in reflux condenser. eSmall amount of overflow. fNecessary to restrict flow with stopcock to maintain solvent head.
Results and discussion Table 1 lists the extraction conditions and Figs. 2-4 present the resulting extraction curves. The conditions had been chosen to provide a fair comparison of performance, to accommodate the different model systems, and to include a high-speed run. Such extraction curves represent the best way to express quantitatively the results of this and similar studies. Clearly, the two unconventional extractors evaluated here were considerably faster than the quite efficient Soxhlet. This research was supported by NSERC grant A-9604. REFERENCES 1 L. C. Craig and D. Craig, in A. Weissberger (Ed.), Technique of Organic Chemistry, __ __ -~Vol. III, Interscience, New York, 1956, p_ 149. 2 W. A. Aue, M. M. Daniewski, J. Miiller and J. P. Laba, Anal. C&em., 49 (1977) 1465. 3 C. E. Browne, W. L. Buchanan and E. J. Eisenbraun, Chem. Ind., 1977, p_ 35.
4