Jevs
05-01-04, 16:37
I was sent this today via e-mail (I think Les got a copy too). It might as well be written in russian to me cos it doesn't mean a lot. However, this long winded report may interest someone.
Fakultat III
Prozesswissenschaften
Institut fur Technischen
Umweltschutz
Fachgebiet
Wasserreinhaltung
Report
Studies on phosphorus removal
from fresh water and sea water
By commercial sorbents
Prof. Dr.-lng. M. Jekel
Dipl.-lng. Arne Genz
Uta Stindt
Technical University Berlin
Department of Water Quality Control
Sekr. KF 4
Strasse des 17. Juni 135
0-10623 Berlin (Germany)
Phone. ++49 +30 314 25058
Fax: ++49 +30 314 23850
e-mail: wrh@tu-berlin.de
Studies on phosphorus removal from fresh and sea water by commercial sorbents
Contents:
1. Introduction ………………………………………………………………………………. 2
2. Tested sorbents …………………………………………………………………………. 2
3. Methods ………………………………………………………………………………….. 5
3.1 Sorbent preparation …………………………………………………………………… 6
3.2 Sea water preparation ………………………………………………………………… 6
3.3 Isotherms ………………………………………………………………………………. 6
3.4 Elution studies ………………………………………………………………………… 8
4. Analytics ………………………………………………………………………………… 9
4.1 Sample preparation ………………………………………………………………….. 9
4.2 Orthophosphate ……………………………………………………………………… 9
4.3 Dissolved organic carbon …………………………………………………………… 10
4.4 Spectral absorption coefficient at 254 nm (SAC254) ……………………………… 10
4.5 pH, conductivity, turbidity …………………………………………………………… 10
4.6 ICP-MS, AAS ………………………………………………………………………… 10
5. Results ………………………………………………………………………………… 11
5.1 Results of isotherm studies ………………………………………………………… 11
5,2 Description with the Langmuir isotherm equation ……………………………….. 13
5.3 Results of elution studies ………………………………………………………….. 19
6. Conclusions …………………………………………………………………………… 22
7. Literature ……………………………………………………………………………… 24
1.
Studies on phosphorus removal from fresh and sea water by commercial sorbents
1. Introduction
Dissolved orthophosphate is the predominating phosphorus species in. fresh and
sea water. Typical concentrations in natural uncontaminated fresh and sea water
range from 5 to 30 ~g/L referring to phosphorus (Grobkopf 1998).
In aquaria these concentrations can be dramatically exceeded due to increased
supply with phosphate combined with insufficient fresh water exchange.
Main sources for phosphorus are the fish food, esp. when containing proteins
(mussel, krill), the excreted food and dying organisms, which contain phosphate in
their cell membranes. As a consequence an increased growth of algae can be
observed. Furthermore, increased phosphate concentrations influence the
synthesis of limestone due to precipitation of calcium phosphates and therefore
complicate the growth of corals.
In most cases the potential of water plants for removing dissolved phosphorus is
not sufficient to guarantee the required water quality. Frequent water replacement
may lead to satisfying phosphorus concentration but is a very expensive solution.
The irreversible removal of phosphorus from the water cycle is possible by using
solid adsorption media. The ideal media has a high capacity and a high specificity
regarding phosphate and should not influence other water parameters. '
2. Tested sorbents
Several commercial media are available for removal of phosphate from fresh and
sea water which are recommended to be used in the circulation filters of the tanks.
The studied media are iron-based (RowaPhos) or aluminum-based (ElimiPhos),
some in the form of zeolite (AntiPhos, PhosEx) or ceramics (Phosphate Sponge),
Table 1 gives a summary of the sorbent's characteristics.
2.
Studies on phosphorus removal from fresh and sea water by commercial sorbents
The following studies with 5 sorbents were carried out according to our offer
dated June 6th 2002 :
1. phosphate sorption isotherms for concentrations smaller than 1 mg/L P for
Berlin tap water and artificial sea water,
2. elution studies with subsequent determination of dissolved organic carbon
(DOC), pH, conductivity, heavy metals (Cu, Pb, Zn, Ni, Cd) and elements AI, Fe,
Mn.
Table 1: Characteristics of the tested sorbents
3.
Studies on phosphorus removal from fresh and sea water by commercial sorbents
Sorbent Phos Ex Phosphate Sponge
Producer JBL GmbH & Co. KG Kent Marine, Inc,
Postfach 17 1100 Northpoint Parkway
D-69137 Neuhofen Acworth, Georgia 30102
(USA)
Specifications; natural zeolite, specially treated ceramics
(acc. to producer)
add. Information 1lkg PhosEx binds 209 PO43- rinse with d.i. water before
use,
of producer 5-6 months efficient, PO43- is sufficient for 120 gal (454L)
reduced from 10 mg/L to
1 mg/L, long contact time
recommended
4.
Studies on phosphorus removal from fresh and sea water by commercial sorbents
5.
Studies on phosphorus removal from fresh and sea water by commercial sorbents
3. Methods
3.1 Sorbent preparation
For determination of sorption isotherms, all sorbents were ground in a mortar by
hand and sieved to a fraction < 63 ~m,
Because dried RowaPhos will lose its capacity, a special preparation process
was developed for this sorbent:
?? Rinsing with Millipore -water,
?? Adjustment of pH (8) with NaOH,
?? Drying at room-temperature for 16 h (water content -22 %),
?? Grinding per mortar and sieving to the fraction < 63 microns.
3.2 Sea water preparation
For first studies the commercial salt "Tropic Marin" was used for sea water
preparation (33 g/L Millipore -water). After one day storing this water showed a
high turbidity due to precipitation processes. Therefore "Tropic Marin" was not
usable for analytical purposes and was replaced by a salt produced by Dr.
Biener GmbH (36367 Wartenberg, Germany). This water showed a better
behavior during subsequent analytics and no precipitation was observed during
the whole period. The density was determined with 1.017 kg/m3 at 28°C.
3.3 Isotherms
Six isotherm points were determined for each sorbent at the original pH of sea
water (pH 8.0) and tap water (pH 8.1).
To evaluate adsorption isotherms, the 'method of adding different quantities of
sorbent (m) to solution volumes (L; L/m variable) of the same initial
concentration (c0= 1 mg/L P) was used. The initial concentration of 1 mg/L P
was produced by dilution of a phosphorus stock solution (c=1 g/L P) which
6.
Studies on phosphorus removal from fresh and sea water by commercial sorbents
buffered at the original pH of the sea water (3.7 parts KH2PO4 and 96.3 parts
Na2HPO4.2H2O in Millipore - water).
Defined quantities of sorbent (m= 5 - 800 mg dry matter) were added to several
glass bottles containing defined volumes of sea water/tap water (L= 200 ml). The
flasks were shaken for 96 h at 20 °C on a linear shaker and subsequently filtered
through 0.45 micron cellulose nitrate 'filter (Sartorius). The pH was measured
immediately. The filtrates were stored for P-analytics at 10°C.
The equilibrium loading q in dependence on the reached equilibrium concentration c
was calculated using the mass balance of the system:
q = L/M ( c0 – ceq ) (1)
with q: solid-phase concentration [mg/g DM], co: initial liquid-phase concentration
[mg/L], Ceq: liquid-phase concentration [mg/L] at equilibrium, L: liquid volume [L]
and m: mass of sorbent [g DM].
The calculated loadings are related on the dry mass of each sorbent (Table 2:).
Table 2:
Water content and dry mass of all sorbents. For RowaPhos for the pretreated
material according to chapter 3.1 and the original material.
Sorbent RowaPhos AntiPhos ElimiPhos Phos Ex Phosphate
Sponge
Water 21.0 5.2 11.1 2.3 11.6
Content (pre-treated)
(%)
54.0
(original)
Dry Mass 79.0 94.8 88.9 97.7 88.4
(pre-treated)
(%) 46.0
(original)
7.
Studies on phosphorus removal from fresh and sea water by commercial sorbents
3.4 Elution studies
The elution studies were performed according to DIN-standard 38414 [1]. 50g of the
original sorbent (moist mass) were shaken with 0.5 L water for 24 h at 20°C.
For the parameters pH, conductivity, dissolved organic carbon (DOC) and UV
absorbance (SAC254), the elution was performed with Millipore water in order to
avoid buffering and interaction with water components. It has to be emphasized that
this elution procedure might not be comparable to the conditions in an aquarium in
which the water is well buffered, but helps to understand the composition of the
media.
Since most heavy metals are known as mobile cations at low pH, the pH is believed
to be the decisive parameter during elution. pH values lower than 6 or higher than 9
might lead to an increased dissolving of heavy metals. Therefore, the elution was
tested with tap water. Since the pH was kept constant between 6.5 and 8.5 (by
adding HCI/NaOH), this procedure allows reasonable comparison between the
different media. It can be assumed that in most cases the pH in an aquarium should
be within this range.
Subsequent to elution, the samples were filtered and stored for analytics,
8.
Studies on phosphorus removal from fresh and sea water by commercial sorbents
4. Analytics
4.1 Sample preparation
Samples for determination of DOC, SAC254 and Orthophosphate were filtered
through 0.45 micron cellulose-nitrate-filter (Sartorius). Samples were kept 10°C until
analyses.
For exemption from P all glassware were stored in an acid bath (HCI 10%) overnight
and rinsed with Millipore water.
4.2 Orthophosphate
For determination of orthophosphate a flow injection analyzer (FIAstar 5000, Foss
Tecator) was 'used according to ISO/DIS 15681:2001 part 1. The analysis bases on
the formation of heteropolyacids in the presence of phosphate and molybdate ions in
acidic solution, which are reduced in a second step by tin(II)chloride to a blue
colored molybdate complex. The detector measures two wavelengths (720 nm, 1000
nm) simultaneously. The reference wavelengths of 1000 nm is used to compensate
optical and electronic fluctuations. The signal is evaluated using the peak height. The
activated drift control excludes peaks with shape abnormality (tailing, fronting). Each
value is an average value of three measurements. All samples were measured at
room temperature.
Calibration was performed with sea water and tap water standards, respectively.
Measuring ranges of 0 – 100 micro g/L and 0 – 500 micro g/L P were utilized. The
lowest standard was 5 micro g/L for both calibration ranges. Daily fluctuations were
compensated by using daily factors determined with the highest standard of each
calibration (100 and 500 micro g/L P). As carrier were used Millipore-water for
isothermms in tap water and salt solution (33 g/L NaG!) for isotherms in sea water.
The detection limit .is specified with 0.5 micro g/L P with a relative standard deviation
of 0.7% related on a 100 micro g/L P standard (Foss Tecator 2001 ).
9.
Studies on phosphorus removal from fresh and sea water by commercial sorbents
4.3 Dissolved organic carbon
Dissolved organic carbon (DOC) was analyzed undiluted using a LiquiTOC analyser
with autosampler by wet oxidation, After acidification with phosphoric acid and
removal of inorganic carbon the sample is oxidized with peroxodisulfate, oxygen and
UV light. As oxidation product CO2 is measured using NDIR detection. Calibration
was performed with phthalate solutions ranging from 0.25 to 10 mg/L C. The
detection limit is specified with 0.40 mg/L with a standard deviation of 0.27 mg/L.
4.4 Spectral absorption coefficient at 254 nm (SAC254)
Spectral absorption coefficients SAC at 254 nm [1/m] were determined using a
Lambda 12 UV/VIS-spectrophotometer (Perkin-Elmer) and cuvettes of 1 cm length.
Because of the linear correlation between the SAC254 and the DOC in natural waters,
theSAC254 is used as a control measurement for the DOC values.
4.5 pH, conductivity, turbidity
The pH was measured by a WTW pH 340-meter using a Sentix 41-electrode (WTW).
Calibration was performed with technical standards (WTW). Conductivity was
measured using a WTW LF 323-A conductivity meter. Turbidity was determined
using a Dr. Lange LTP 5 turbidity meter measuring stray light at 90°.
4.6 ICP-MS, AAS
Metal ions were measured using a Quadrupol-ICP-MS ELAN 6000 (Perkin Elmer).
As detector a mass spectrometer with CEM was used. 0.1 ml Sc-solution per 10 ml
sample were added in order to compensate matrix effects. Calibration was
performed using a multi element standard IV (Merck). For further elution studies,
metal ions were measured using flame atomic absorption spectroscopy (SpectrAA
400, Varian).
10.
Studies on phosphorus removal from fresh and sea water by commercial sorbents
5. Results
5.1 Results of isotherm studies
The adsorption of phosphate on metal oxide surfaces has been well studied by
numerous authors. Phosphate is regarded as a specific adsorbing ion, which means
that phosphate can adsorb at pH values higher than the pHpzc (point of zero charge)
by the formation of complexes. Sigg and Stumm (1981) describe the formation of
inner spheric complexes on metal oxide surfaces. Persson et al. (1996) suggest the
formation of a mono dentate complex for sorption of phosphate on goethite. Rietra et
al. (2001) additionally point on the role of calcium for these sorption processes
Figure 1
Phosphate-isotherms, Berlin tap water, pH 8
(20°C, 96h shaking time, c0=1 mg/L P)
11.
Studies on phosphorus removal from fresh and sea water by commercial sorbents
In Figure 1 phosphate isotherms in Berlin tap water are shown. The isotherms in sea
water (Figure 2) show a similar trend as in tap water. The equilibrium loadings are
somewhat smaller than those in tap water but still in the same magnitude.
Figure 2
Phosphate-isotherms, sea water (Dr. Biener), pH 8,
(20°C, 96h shaking time, c0=1 mg/L P)
With RowaPhos the highest loadings were achieved. The reached maximum loading
of around 15 mg/g agrees with earlier studies at the Department of Water Quality
Control in pure systems (Driehaus 1994).
The isotherms of Phosphate Sponge and ElimiPhos run remarkable parallel for both
waters. Therefore, it is assumed that these media consist of identical structures,
supposedly on the basis of aluminum oxide. Altogether they showed a moderate
adsorption performance with loadings around 5 mg/g
12.
Studies on phosphorus removal from fresh and sea water by commercial sorbents
Though it was possible to remove all phosphate from the solution with AntiPhos
(which means that equilibrium concentration of 0 mg/L is reachable), the achieved
loadings are comparably low (smaller than 1 mg/g). The isotherm's shapes show only
weak affinity towards phosphate.
The most eye-catching characteristics showed the sorbent PhosEx. Even with the
highest sorbent dosage the equilibrium concentration does not drop below 450 micro
g/L. P. The achieved loadings are very low or almost zero.
5.2 Description with the Langmuir isotherm equation
For mathematical description of the isotherms and a better comparison, adaptation
was performed using the Langmuir isotherm equation:
KL . c
q = qm . 1+KL . c (2)
With q: loading [mg/g DM], qm: [mg/g DM], c: liquid-phase concentration (mg/L) at
equilibrium, KL: Langmuir constant [L/mgl
This equation was developed for the description of the adsorption of gases on solid
surfaces (Langmuir, 1918). qm characterizes the maximum capacity (monomolecular
loading) of the sorbent, KL the affinity which can be seen as the slope of the isotherm
curve. Equation (2) can be transformed into the linear function (3)
c / q = ( 1 / qm . KL ) + ( 1 / qm ) . c (3)
The two constants of the Langmuir isotherm equation (KL, qm) can be determined
from isotherm data by linear regression. In general, this method gives good results for
the isotherm points at higher equilibrium concentrations and leads to an accurate
description of qm.
13.
Studies on phosphorus removal from fresh and sea water by commercial sorbents
The adaptation was performed for the media RowaPhos, ElimiPhos and Phosphate
Sponge. The regression coefficients varied from 0.94 to 0.99. Due to the poor
adsorption characteristics of AntiPhos and PhosEx, a reasonable adaptation was not
possible for these two media.
The obtained isotherms are shown in Figure 3 (tap water) and Figure 4 (sea water).
For adaptation for RowaPhos in sea water, the isotherm points for the, highest L/m
ratio were excluded. These points might have been influenced by precipitation, most
likely of calcium phosphate.
For practically removing phosphorus from aquaria, not the maximum capacity qm
might be of relevance, but the reachable loading in the range of smaller phosphorus
concentrations. For this reason, the achieved loadings at equilibrium concentrations
of 50 and 100 micro g/L P were evaluated and compared. The higher are these
loadings, the steeper is the Isotherm curve and the more efficient is the P removal in
the small concentration range. The obtained parameters of adaptation are shown in
Table 3.
14.
Studies on phosphorus removal from fresh and sea water by commercial sorbents
Table 3:
Maximum capacity qm and capacity at equilibrium concentrations of 50 and 100 micro
g/L P (mg P/g DM), KL (L/mg) and R 2 for Berlin tap water and sea water according to
the Langmuir isotherm equation.
15.
Studies on phosphorus removal from fresh and sea water by commercial sorbents
Figure 3 Phosphate isotherms (acc. to Langmuir equation), Berlin tap water,pH
8, (20°C, 96h shaking time, c o = 1 mg/L P)
Figure 4 Phosphate isotherms (acc. to Langmuir equation), sea water (Dr.
Biener), pH 8 (20°C, 96h shaking time, c o = 1 mg/L P)
16.
Studies on phosphorus removal from fresh and sea water by commercial sorbents
For the 3 media a decrease of capacity in sea water could be observed compared to
tap water ranging from 17 to 35% of the maximum capacity. Therefore, even with a
high excess of competing ions (chloride, sulfate) in sea water, only a relatively small
decline in loading was found. Consequently, phosphate seems to adsorb mainly by
specific complex formation for all tested media. The sorbent surface should be
negatively charged at pH 8. As a consequence, the non-specific adsorbing ions like
chloride and sulfate do not show strong competition on the sorption of phosphate.
Likewise, effects by co-sorption of Calcium could not be observed since Calcium was
excessively present for both tap water and sea water.
RowaPhos showed in both waters the highest maximum capacity (qm). The qm for
ElimiPhos and Phosphate Sponge were around one third of RowaPhos for both tap
water and sea water.
With regard to q50 and q100, again the performance of RowaPhos was the best with
around 5 mg P/g DM. For ElimiPhos and Phosphate Sponge, q50 and q100 ranged
between 30 and 40% of the values for RowaPhos.
17.
Studies on phosphorus removal from fresh and sea water by commercial sorbents
For the practical application in a adsorption filter it might be relevant how much
phosphorus is removable with regard to a given adsorption filter volume. In Table 4
the mass of phosphorus that can be removed from tap water in a adsorption filter of 1
L volume is calculated for all sorbents using the respective maximum loadings
(chapter 5.1 and 5.2) and the bulk densities.
Table 4:
Removable phosphorus in a 1L adsorption filter
18.
Studies on phosphorus removal from fresh and sea water by commercial sorbents
5.3 Results of elution studies
In Table 5 the results of elution with Millipore water are shown. With regard to
conductivity, a remarkable shift was observed for AntiPhos up to 1500 micro
siemens/cm indicating increased dilution of inorganic salts. All other samples ended
up with a conductivity between 250 and 350 micro siemens/cm.
RowaPhos and AntiPhos cause a pH shift from pH 5.4 (Millipore) down to 4.8 and
4.9, respectively. ElimiPhos and Phosphate Sponge are alkaline media, causing a pH
shift from pH 5.4 (Millipore) up to 8.7 and 10, respectively. PhosEx is the only medium
that did not cause any shift in pH.
Generally, the DOC and SAC254 values show a parallel behavior. For ElimiPhos and
Phosphate Sponge an increased DOC (around 2.5mg/L) was measured. As found
during the isotherm studies and the pH values (s.a.), these two media again show a
parallel behaviour, which confirms the assumption, that these media have similar
structures and are supposedly produced in a comparable procedure. AntiPhos and
PhosEx showed a moderate elution with 1.2 and 1.7 mg/L DOC. For RowaPhos only
little elution of organics was observed. In general the found DOC values are
comparatively low when considering the high mass to volume ratio used.
19.
Studies on phosphorus removal from fresh and sea water by commercial sorbents
Table 6 gives the results of elution with tap water at pH values between 6.5 and 9.5.
Table 6: Results of elution studies (elution with tap water ,pH 7-8, 100g
moist mass per L, 24 h shaking time, 20°C)
Rowa Phos Elimi-Phos Phosphate
Sponge
Anti-Phos PhosEx
pH 7.0 7.4 7.5 7.8 7.4
Cu (mg/L) <0.01 <0.01 <0.01 <0.01 <0.01
Pb (mg/L) <0.01 <0.01 <0.01 <0.01 <0.01
Zn (mg/L) <0.01 <0.01 <0.01 <0.01 <0.01
Ni (mg/L) <0.01 <0.01 0.01 <0.01 <0.01
Cd (mg/L) <0.01 <0.01 <0.01 <0.01 <0.01
Al (mg/L) <0.01 <0.01 0.01 0.4 0.3
Fe (mg/L) <0.01 <0.01 <0.01 <0.01 <0.01
Mn (mg/L) 0.01 <0.01 <0.01 0.53 0.29
For all sorbents, no elution of the harmful heavy metals Cu, Pb, Zn, Ni and Cd was
observed.
In general, dissolution of aluminum and iron is strongly connected with the respective
pH. Aluminum is almost insoluble at pH 6.0 to 7.5, iron at pH 6 to 10, but the
solubility increases again at pH values below or higher. For all media, no significant
iron dissolution was observed. For Phosphate Sponge, AntiPhos and PhosEx, only
slightly increased values for aluminum were found. This might indicate that these
media contain significant amounts of aluminum.
20.
Studies on phosphorus removal from fresh and sea water by commercial sorbents
Manganese was found in relative small concentrations for AntiPhos (0.53 mg/L) and
PhosEx (0.29 mg/L}. According to the WHO guidelines, manganese is regarded as
one of the least toxic elements. The WHO drinking water standard is 0.5 mg/L.
Considering the high mass/volume ratio used, the found manganese concentrations
are not regarded as dangerous.
21
Studies on phosphorus removal from fresh and sea water by commercial sorbents
6. Conclusions
For assessment of the studied media the capacity at higher and low equilibrium
concentrations and the price should be considered. These parameters strongly
depend on the initial P concentration and the target concentration. Additionally,
potential harmful impacts on the aquaristic flora and fauna should be taken into
account.
For the range of low P concentration, the iron-based medium RowaPhos shows the
best sorption characteristics in tap and sea water. This medium is recommendable
for low P target concentrations since it showed the highest q 50 and q 100 values.
ElimiPhos and Phosphate Sponge showed very similar sorption characteristics, but
the price for phosphate sponge is significantly lower. It is assumed that these media
mainly consist of aluminum oxide.
AntiPhos lies in the same price range but showed significant lower loadings. This
might be the reason why the producer recommends for application a 10 times
higher amount for 200 L sea water than for all other media.
PhosEx has the lowest price of all media and also the lowest performance in terms
of loading and affinity. Removal of phosphate down to concentration lower than 500
micro g/L was not possible. High sorbent dosages might be necessary to achieve a
low P concentration in aquaria, which might increase costs for the process design
(larger adsorbers etc. The producer specifies that PhosEx would reduce the
phosphate concentration from 10 mg/L to 1 mg/L (concentration probably given as
phosphate). Summarizing, the zeolite based media AntiPhos and PhosEx are not
recommendable for the range of low P-concentrations. It is not likely that these
media show a better performance at higher P-concentrations.
For all media an impact on the pH of Millipore water was found. Although sea water
will be buffering, it is recommendable to rinse esp. RowaPhos, AntiPhos and
Phosphate
22.
Studies on phosphorus removal from fresh and sea water by commercial sorbents
Sponge before application. It is believed that all tested media contain heavy metals;
therefore the pH should be kept between 7.0 and 8.5 to prevent elution. AntiPhos
showed additionally increased conductivity after elution.
For the application in a fixed bed adsorber not only the sorption equilibrium is of
interest but also the sorption kinetics. This has to be considered for the design
layout, esp. regarding the contact time in the filter. More advanced and expensive
testing in pilot or full scale would be necessary to assess these parameters.
Considering the German drinking water standards (Table 7), the found values for
Cu, Pb, Zn and Ni are below these standards. For cadmium, the standard is lower
than the detection limit. The PhosEx and AntiPhos eluates showed a slightly
increased Mn value, which might be connected with the indefinite composition of
these media. For assessment of the eluted metal concentrations, the applied mass
of sorbent in relation to the water volume used in the sea water tank has to be
considered.
According to the producer's specifications (Table 1), the recommended mass/volume
ratios are around 100 times smaller than those used in the elution studies.
Therefore, the measured concentrations can be roughly divided by 100 in order to
obtain the effective - concentrations during application. Consequently, all metal
concentrations measured after elution would be below these standards. However,
some special fishes may be more sensitive to heavy metals than humans.
Table 7:
German drinking water standards according to the Trinkwasser- Verordnung (2001 )
Metal Cu Pb Zn Ni Cd Al Fe Mn
Concentration 2 0.01 - 0.02 0.005 0.2 0.2 0.05
(mg/L)
23.
Studies on phosphorus removal from fresh and sea water by commercial sorbents
7. Literature
DIN 38414 Teil 4, Schlamm und Sedimente (Gruppe S), Bestimmung der
Eluierbarkeit mit Wasser (S4), Deutsche Einheitsverfahren zur Wasser-, Abwasserund
Schlammuntersuchung, Normenausschuss Wasserwesen (NAW) im DIN
Deutsches institut fur Normung e.V., 1984
Driehaus, W. (1994). Arsenentfernung mit Magnesiumdioxid and Elsenhydroxid in
der Trinkwasseraufbereitung. Fortschritt-Berichte VDl Reihe15 Nr. 33, VDl Verlag,
Dusseldorf
Foss Tecator. Application Note. Bestimmung von Orthophosphat im Wasser mit dem
Fiastar 5000 (2001)
Grobkopf, J. (1998). Phosphat in der meeresaquaristischenPraxis. Der Meerwasseraquarianer,
2. Jahrgang, Januar 1998, Seiten 10-15
lSO/DIS 15681-1. Water quality -- Determination of orthophosphate and total
phosphorus contents by flow analysis (FIA and CFA) -- Part 1: Method by flow
injection analysis (FIA)
Langmuir, I. (1908). The adsorption of gases on plane surfaces of glass, mjca and
platinum. J. Ameri. Chem. Soc. 40,1361-1403
Persson, P., Nilsson; N., Sjoberg, S. (1996). Structure and Bonding of
Orthophosphate Ions at the Iron Oxide-Aqueous Interface. Journal of Colloid and
Interface Science 177, 263-275
Rietra,.R., Hiemstra, T. Van Riemsdijk, W. (2001) Interaction between Calcium and
Phosphate Adsorption on Goethite. Environ. Sci. Techn. 35, 3369-3374.
Sontheimer, H; Frick, B.;Fettig, J.; (1985). Adsorptionsverfahren zur
Wasseraufbereitung DVGW- Forschungsstelle am Engler- Bunte-Insitut. Karlruhe.
Stumm, W., Sigg, L. (1981), The interaction of anions and weak acids with the
hydrous goethite surface. Colloids and Surfaces Vol. 2 Issue 2, 101-117
Trinkwasserverordnung (2001). Verordnung zur Novellterung der Trinkwasserverordnung
yom 21.Mai 2001. Bundesgesetzblatt Jahrgang 2001 Teil1 Nr. 24
24.
Fakultat III
Prozesswissenschaften
Institut fur Technischen
Umweltschutz
Fachgebiet
Wasserreinhaltung
Report
Studies on phosphorus removal
from fresh water and sea water
By commercial sorbents
Prof. Dr.-lng. M. Jekel
Dipl.-lng. Arne Genz
Uta Stindt
Technical University Berlin
Department of Water Quality Control
Sekr. KF 4
Strasse des 17. Juni 135
0-10623 Berlin (Germany)
Phone. ++49 +30 314 25058
Fax: ++49 +30 314 23850
e-mail: wrh@tu-berlin.de
Studies on phosphorus removal from fresh and sea water by commercial sorbents
Contents:
1. Introduction ………………………………………………………………………………. 2
2. Tested sorbents …………………………………………………………………………. 2
3. Methods ………………………………………………………………………………….. 5
3.1 Sorbent preparation …………………………………………………………………… 6
3.2 Sea water preparation ………………………………………………………………… 6
3.3 Isotherms ………………………………………………………………………………. 6
3.4 Elution studies ………………………………………………………………………… 8
4. Analytics ………………………………………………………………………………… 9
4.1 Sample preparation ………………………………………………………………….. 9
4.2 Orthophosphate ……………………………………………………………………… 9
4.3 Dissolved organic carbon …………………………………………………………… 10
4.4 Spectral absorption coefficient at 254 nm (SAC254) ……………………………… 10
4.5 pH, conductivity, turbidity …………………………………………………………… 10
4.6 ICP-MS, AAS ………………………………………………………………………… 10
5. Results ………………………………………………………………………………… 11
5.1 Results of isotherm studies ………………………………………………………… 11
5,2 Description with the Langmuir isotherm equation ……………………………….. 13
5.3 Results of elution studies ………………………………………………………….. 19
6. Conclusions …………………………………………………………………………… 22
7. Literature ……………………………………………………………………………… 24
1.
Studies on phosphorus removal from fresh and sea water by commercial sorbents
1. Introduction
Dissolved orthophosphate is the predominating phosphorus species in. fresh and
sea water. Typical concentrations in natural uncontaminated fresh and sea water
range from 5 to 30 ~g/L referring to phosphorus (Grobkopf 1998).
In aquaria these concentrations can be dramatically exceeded due to increased
supply with phosphate combined with insufficient fresh water exchange.
Main sources for phosphorus are the fish food, esp. when containing proteins
(mussel, krill), the excreted food and dying organisms, which contain phosphate in
their cell membranes. As a consequence an increased growth of algae can be
observed. Furthermore, increased phosphate concentrations influence the
synthesis of limestone due to precipitation of calcium phosphates and therefore
complicate the growth of corals.
In most cases the potential of water plants for removing dissolved phosphorus is
not sufficient to guarantee the required water quality. Frequent water replacement
may lead to satisfying phosphorus concentration but is a very expensive solution.
The irreversible removal of phosphorus from the water cycle is possible by using
solid adsorption media. The ideal media has a high capacity and a high specificity
regarding phosphate and should not influence other water parameters. '
2. Tested sorbents
Several commercial media are available for removal of phosphate from fresh and
sea water which are recommended to be used in the circulation filters of the tanks.
The studied media are iron-based (RowaPhos) or aluminum-based (ElimiPhos),
some in the form of zeolite (AntiPhos, PhosEx) or ceramics (Phosphate Sponge),
Table 1 gives a summary of the sorbent's characteristics.
2.
Studies on phosphorus removal from fresh and sea water by commercial sorbents
The following studies with 5 sorbents were carried out according to our offer
dated June 6th 2002 :
1. phosphate sorption isotherms for concentrations smaller than 1 mg/L P for
Berlin tap water and artificial sea water,
2. elution studies with subsequent determination of dissolved organic carbon
(DOC), pH, conductivity, heavy metals (Cu, Pb, Zn, Ni, Cd) and elements AI, Fe,
Mn.
Table 1: Characteristics of the tested sorbents
3.
Studies on phosphorus removal from fresh and sea water by commercial sorbents
Sorbent Phos Ex Phosphate Sponge
Producer JBL GmbH & Co. KG Kent Marine, Inc,
Postfach 17 1100 Northpoint Parkway
D-69137 Neuhofen Acworth, Georgia 30102
(USA)
Specifications; natural zeolite, specially treated ceramics
(acc. to producer)
add. Information 1lkg PhosEx binds 209 PO43- rinse with d.i. water before
use,
of producer 5-6 months efficient, PO43- is sufficient for 120 gal (454L)
reduced from 10 mg/L to
1 mg/L, long contact time
recommended
4.
Studies on phosphorus removal from fresh and sea water by commercial sorbents
5.
Studies on phosphorus removal from fresh and sea water by commercial sorbents
3. Methods
3.1 Sorbent preparation
For determination of sorption isotherms, all sorbents were ground in a mortar by
hand and sieved to a fraction < 63 ~m,
Because dried RowaPhos will lose its capacity, a special preparation process
was developed for this sorbent:
?? Rinsing with Millipore -water,
?? Adjustment of pH (8) with NaOH,
?? Drying at room-temperature for 16 h (water content -22 %),
?? Grinding per mortar and sieving to the fraction < 63 microns.
3.2 Sea water preparation
For first studies the commercial salt "Tropic Marin" was used for sea water
preparation (33 g/L Millipore -water). After one day storing this water showed a
high turbidity due to precipitation processes. Therefore "Tropic Marin" was not
usable for analytical purposes and was replaced by a salt produced by Dr.
Biener GmbH (36367 Wartenberg, Germany). This water showed a better
behavior during subsequent analytics and no precipitation was observed during
the whole period. The density was determined with 1.017 kg/m3 at 28°C.
3.3 Isotherms
Six isotherm points were determined for each sorbent at the original pH of sea
water (pH 8.0) and tap water (pH 8.1).
To evaluate adsorption isotherms, the 'method of adding different quantities of
sorbent (m) to solution volumes (L; L/m variable) of the same initial
concentration (c0= 1 mg/L P) was used. The initial concentration of 1 mg/L P
was produced by dilution of a phosphorus stock solution (c=1 g/L P) which
6.
Studies on phosphorus removal from fresh and sea water by commercial sorbents
buffered at the original pH of the sea water (3.7 parts KH2PO4 and 96.3 parts
Na2HPO4.2H2O in Millipore - water).
Defined quantities of sorbent (m= 5 - 800 mg dry matter) were added to several
glass bottles containing defined volumes of sea water/tap water (L= 200 ml). The
flasks were shaken for 96 h at 20 °C on a linear shaker and subsequently filtered
through 0.45 micron cellulose nitrate 'filter (Sartorius). The pH was measured
immediately. The filtrates were stored for P-analytics at 10°C.
The equilibrium loading q in dependence on the reached equilibrium concentration c
was calculated using the mass balance of the system:
q = L/M ( c0 – ceq ) (1)
with q: solid-phase concentration [mg/g DM], co: initial liquid-phase concentration
[mg/L], Ceq: liquid-phase concentration [mg/L] at equilibrium, L: liquid volume [L]
and m: mass of sorbent [g DM].
The calculated loadings are related on the dry mass of each sorbent (Table 2:).
Table 2:
Water content and dry mass of all sorbents. For RowaPhos for the pretreated
material according to chapter 3.1 and the original material.
Sorbent RowaPhos AntiPhos ElimiPhos Phos Ex Phosphate
Sponge
Water 21.0 5.2 11.1 2.3 11.6
Content (pre-treated)
(%)
54.0
(original)
Dry Mass 79.0 94.8 88.9 97.7 88.4
(pre-treated)
(%) 46.0
(original)
7.
Studies on phosphorus removal from fresh and sea water by commercial sorbents
3.4 Elution studies
The elution studies were performed according to DIN-standard 38414 [1]. 50g of the
original sorbent (moist mass) were shaken with 0.5 L water for 24 h at 20°C.
For the parameters pH, conductivity, dissolved organic carbon (DOC) and UV
absorbance (SAC254), the elution was performed with Millipore water in order to
avoid buffering and interaction with water components. It has to be emphasized that
this elution procedure might not be comparable to the conditions in an aquarium in
which the water is well buffered, but helps to understand the composition of the
media.
Since most heavy metals are known as mobile cations at low pH, the pH is believed
to be the decisive parameter during elution. pH values lower than 6 or higher than 9
might lead to an increased dissolving of heavy metals. Therefore, the elution was
tested with tap water. Since the pH was kept constant between 6.5 and 8.5 (by
adding HCI/NaOH), this procedure allows reasonable comparison between the
different media. It can be assumed that in most cases the pH in an aquarium should
be within this range.
Subsequent to elution, the samples were filtered and stored for analytics,
8.
Studies on phosphorus removal from fresh and sea water by commercial sorbents
4. Analytics
4.1 Sample preparation
Samples for determination of DOC, SAC254 and Orthophosphate were filtered
through 0.45 micron cellulose-nitrate-filter (Sartorius). Samples were kept 10°C until
analyses.
For exemption from P all glassware were stored in an acid bath (HCI 10%) overnight
and rinsed with Millipore water.
4.2 Orthophosphate
For determination of orthophosphate a flow injection analyzer (FIAstar 5000, Foss
Tecator) was 'used according to ISO/DIS 15681:2001 part 1. The analysis bases on
the formation of heteropolyacids in the presence of phosphate and molybdate ions in
acidic solution, which are reduced in a second step by tin(II)chloride to a blue
colored molybdate complex. The detector measures two wavelengths (720 nm, 1000
nm) simultaneously. The reference wavelengths of 1000 nm is used to compensate
optical and electronic fluctuations. The signal is evaluated using the peak height. The
activated drift control excludes peaks with shape abnormality (tailing, fronting). Each
value is an average value of three measurements. All samples were measured at
room temperature.
Calibration was performed with sea water and tap water standards, respectively.
Measuring ranges of 0 – 100 micro g/L and 0 – 500 micro g/L P were utilized. The
lowest standard was 5 micro g/L for both calibration ranges. Daily fluctuations were
compensated by using daily factors determined with the highest standard of each
calibration (100 and 500 micro g/L P). As carrier were used Millipore-water for
isothermms in tap water and salt solution (33 g/L NaG!) for isotherms in sea water.
The detection limit .is specified with 0.5 micro g/L P with a relative standard deviation
of 0.7% related on a 100 micro g/L P standard (Foss Tecator 2001 ).
9.
Studies on phosphorus removal from fresh and sea water by commercial sorbents
4.3 Dissolved organic carbon
Dissolved organic carbon (DOC) was analyzed undiluted using a LiquiTOC analyser
with autosampler by wet oxidation, After acidification with phosphoric acid and
removal of inorganic carbon the sample is oxidized with peroxodisulfate, oxygen and
UV light. As oxidation product CO2 is measured using NDIR detection. Calibration
was performed with phthalate solutions ranging from 0.25 to 10 mg/L C. The
detection limit is specified with 0.40 mg/L with a standard deviation of 0.27 mg/L.
4.4 Spectral absorption coefficient at 254 nm (SAC254)
Spectral absorption coefficients SAC at 254 nm [1/m] were determined using a
Lambda 12 UV/VIS-spectrophotometer (Perkin-Elmer) and cuvettes of 1 cm length.
Because of the linear correlation between the SAC254 and the DOC in natural waters,
theSAC254 is used as a control measurement for the DOC values.
4.5 pH, conductivity, turbidity
The pH was measured by a WTW pH 340-meter using a Sentix 41-electrode (WTW).
Calibration was performed with technical standards (WTW). Conductivity was
measured using a WTW LF 323-A conductivity meter. Turbidity was determined
using a Dr. Lange LTP 5 turbidity meter measuring stray light at 90°.
4.6 ICP-MS, AAS
Metal ions were measured using a Quadrupol-ICP-MS ELAN 6000 (Perkin Elmer).
As detector a mass spectrometer with CEM was used. 0.1 ml Sc-solution per 10 ml
sample were added in order to compensate matrix effects. Calibration was
performed using a multi element standard IV (Merck). For further elution studies,
metal ions were measured using flame atomic absorption spectroscopy (SpectrAA
400, Varian).
10.
Studies on phosphorus removal from fresh and sea water by commercial sorbents
5. Results
5.1 Results of isotherm studies
The adsorption of phosphate on metal oxide surfaces has been well studied by
numerous authors. Phosphate is regarded as a specific adsorbing ion, which means
that phosphate can adsorb at pH values higher than the pHpzc (point of zero charge)
by the formation of complexes. Sigg and Stumm (1981) describe the formation of
inner spheric complexes on metal oxide surfaces. Persson et al. (1996) suggest the
formation of a mono dentate complex for sorption of phosphate on goethite. Rietra et
al. (2001) additionally point on the role of calcium for these sorption processes
Figure 1
Phosphate-isotherms, Berlin tap water, pH 8
(20°C, 96h shaking time, c0=1 mg/L P)
11.
Studies on phosphorus removal from fresh and sea water by commercial sorbents
In Figure 1 phosphate isotherms in Berlin tap water are shown. The isotherms in sea
water (Figure 2) show a similar trend as in tap water. The equilibrium loadings are
somewhat smaller than those in tap water but still in the same magnitude.
Figure 2
Phosphate-isotherms, sea water (Dr. Biener), pH 8,
(20°C, 96h shaking time, c0=1 mg/L P)
With RowaPhos the highest loadings were achieved. The reached maximum loading
of around 15 mg/g agrees with earlier studies at the Department of Water Quality
Control in pure systems (Driehaus 1994).
The isotherms of Phosphate Sponge and ElimiPhos run remarkable parallel for both
waters. Therefore, it is assumed that these media consist of identical structures,
supposedly on the basis of aluminum oxide. Altogether they showed a moderate
adsorption performance with loadings around 5 mg/g
12.
Studies on phosphorus removal from fresh and sea water by commercial sorbents
Though it was possible to remove all phosphate from the solution with AntiPhos
(which means that equilibrium concentration of 0 mg/L is reachable), the achieved
loadings are comparably low (smaller than 1 mg/g). The isotherm's shapes show only
weak affinity towards phosphate.
The most eye-catching characteristics showed the sorbent PhosEx. Even with the
highest sorbent dosage the equilibrium concentration does not drop below 450 micro
g/L. P. The achieved loadings are very low or almost zero.
5.2 Description with the Langmuir isotherm equation
For mathematical description of the isotherms and a better comparison, adaptation
was performed using the Langmuir isotherm equation:
KL . c
q = qm . 1+KL . c (2)
With q: loading [mg/g DM], qm: [mg/g DM], c: liquid-phase concentration (mg/L) at
equilibrium, KL: Langmuir constant [L/mgl
This equation was developed for the description of the adsorption of gases on solid
surfaces (Langmuir, 1918). qm characterizes the maximum capacity (monomolecular
loading) of the sorbent, KL the affinity which can be seen as the slope of the isotherm
curve. Equation (2) can be transformed into the linear function (3)
c / q = ( 1 / qm . KL ) + ( 1 / qm ) . c (3)
The two constants of the Langmuir isotherm equation (KL, qm) can be determined
from isotherm data by linear regression. In general, this method gives good results for
the isotherm points at higher equilibrium concentrations and leads to an accurate
description of qm.
13.
Studies on phosphorus removal from fresh and sea water by commercial sorbents
The adaptation was performed for the media RowaPhos, ElimiPhos and Phosphate
Sponge. The regression coefficients varied from 0.94 to 0.99. Due to the poor
adsorption characteristics of AntiPhos and PhosEx, a reasonable adaptation was not
possible for these two media.
The obtained isotherms are shown in Figure 3 (tap water) and Figure 4 (sea water).
For adaptation for RowaPhos in sea water, the isotherm points for the, highest L/m
ratio were excluded. These points might have been influenced by precipitation, most
likely of calcium phosphate.
For practically removing phosphorus from aquaria, not the maximum capacity qm
might be of relevance, but the reachable loading in the range of smaller phosphorus
concentrations. For this reason, the achieved loadings at equilibrium concentrations
of 50 and 100 micro g/L P were evaluated and compared. The higher are these
loadings, the steeper is the Isotherm curve and the more efficient is the P removal in
the small concentration range. The obtained parameters of adaptation are shown in
Table 3.
14.
Studies on phosphorus removal from fresh and sea water by commercial sorbents
Table 3:
Maximum capacity qm and capacity at equilibrium concentrations of 50 and 100 micro
g/L P (mg P/g DM), KL (L/mg) and R 2 for Berlin tap water and sea water according to
the Langmuir isotherm equation.
15.
Studies on phosphorus removal from fresh and sea water by commercial sorbents
Figure 3 Phosphate isotherms (acc. to Langmuir equation), Berlin tap water,pH
8, (20°C, 96h shaking time, c o = 1 mg/L P)
Figure 4 Phosphate isotherms (acc. to Langmuir equation), sea water (Dr.
Biener), pH 8 (20°C, 96h shaking time, c o = 1 mg/L P)
16.
Studies on phosphorus removal from fresh and sea water by commercial sorbents
For the 3 media a decrease of capacity in sea water could be observed compared to
tap water ranging from 17 to 35% of the maximum capacity. Therefore, even with a
high excess of competing ions (chloride, sulfate) in sea water, only a relatively small
decline in loading was found. Consequently, phosphate seems to adsorb mainly by
specific complex formation for all tested media. The sorbent surface should be
negatively charged at pH 8. As a consequence, the non-specific adsorbing ions like
chloride and sulfate do not show strong competition on the sorption of phosphate.
Likewise, effects by co-sorption of Calcium could not be observed since Calcium was
excessively present for both tap water and sea water.
RowaPhos showed in both waters the highest maximum capacity (qm). The qm for
ElimiPhos and Phosphate Sponge were around one third of RowaPhos for both tap
water and sea water.
With regard to q50 and q100, again the performance of RowaPhos was the best with
around 5 mg P/g DM. For ElimiPhos and Phosphate Sponge, q50 and q100 ranged
between 30 and 40% of the values for RowaPhos.
17.
Studies on phosphorus removal from fresh and sea water by commercial sorbents
For the practical application in a adsorption filter it might be relevant how much
phosphorus is removable with regard to a given adsorption filter volume. In Table 4
the mass of phosphorus that can be removed from tap water in a adsorption filter of 1
L volume is calculated for all sorbents using the respective maximum loadings
(chapter 5.1 and 5.2) and the bulk densities.
Table 4:
Removable phosphorus in a 1L adsorption filter
18.
Studies on phosphorus removal from fresh and sea water by commercial sorbents
5.3 Results of elution studies
In Table 5 the results of elution with Millipore water are shown. With regard to
conductivity, a remarkable shift was observed for AntiPhos up to 1500 micro
siemens/cm indicating increased dilution of inorganic salts. All other samples ended
up with a conductivity between 250 and 350 micro siemens/cm.
RowaPhos and AntiPhos cause a pH shift from pH 5.4 (Millipore) down to 4.8 and
4.9, respectively. ElimiPhos and Phosphate Sponge are alkaline media, causing a pH
shift from pH 5.4 (Millipore) up to 8.7 and 10, respectively. PhosEx is the only medium
that did not cause any shift in pH.
Generally, the DOC and SAC254 values show a parallel behavior. For ElimiPhos and
Phosphate Sponge an increased DOC (around 2.5mg/L) was measured. As found
during the isotherm studies and the pH values (s.a.), these two media again show a
parallel behaviour, which confirms the assumption, that these media have similar
structures and are supposedly produced in a comparable procedure. AntiPhos and
PhosEx showed a moderate elution with 1.2 and 1.7 mg/L DOC. For RowaPhos only
little elution of organics was observed. In general the found DOC values are
comparatively low when considering the high mass to volume ratio used.
19.
Studies on phosphorus removal from fresh and sea water by commercial sorbents
Table 6 gives the results of elution with tap water at pH values between 6.5 and 9.5.
Table 6: Results of elution studies (elution with tap water ,pH 7-8, 100g
moist mass per L, 24 h shaking time, 20°C)
Rowa Phos Elimi-Phos Phosphate
Sponge
Anti-Phos PhosEx
pH 7.0 7.4 7.5 7.8 7.4
Cu (mg/L) <0.01 <0.01 <0.01 <0.01 <0.01
Pb (mg/L) <0.01 <0.01 <0.01 <0.01 <0.01
Zn (mg/L) <0.01 <0.01 <0.01 <0.01 <0.01
Ni (mg/L) <0.01 <0.01 0.01 <0.01 <0.01
Cd (mg/L) <0.01 <0.01 <0.01 <0.01 <0.01
Al (mg/L) <0.01 <0.01 0.01 0.4 0.3
Fe (mg/L) <0.01 <0.01 <0.01 <0.01 <0.01
Mn (mg/L) 0.01 <0.01 <0.01 0.53 0.29
For all sorbents, no elution of the harmful heavy metals Cu, Pb, Zn, Ni and Cd was
observed.
In general, dissolution of aluminum and iron is strongly connected with the respective
pH. Aluminum is almost insoluble at pH 6.0 to 7.5, iron at pH 6 to 10, but the
solubility increases again at pH values below or higher. For all media, no significant
iron dissolution was observed. For Phosphate Sponge, AntiPhos and PhosEx, only
slightly increased values for aluminum were found. This might indicate that these
media contain significant amounts of aluminum.
20.
Studies on phosphorus removal from fresh and sea water by commercial sorbents
Manganese was found in relative small concentrations for AntiPhos (0.53 mg/L) and
PhosEx (0.29 mg/L}. According to the WHO guidelines, manganese is regarded as
one of the least toxic elements. The WHO drinking water standard is 0.5 mg/L.
Considering the high mass/volume ratio used, the found manganese concentrations
are not regarded as dangerous.
21
Studies on phosphorus removal from fresh and sea water by commercial sorbents
6. Conclusions
For assessment of the studied media the capacity at higher and low equilibrium
concentrations and the price should be considered. These parameters strongly
depend on the initial P concentration and the target concentration. Additionally,
potential harmful impacts on the aquaristic flora and fauna should be taken into
account.
For the range of low P concentration, the iron-based medium RowaPhos shows the
best sorption characteristics in tap and sea water. This medium is recommendable
for low P target concentrations since it showed the highest q 50 and q 100 values.
ElimiPhos and Phosphate Sponge showed very similar sorption characteristics, but
the price for phosphate sponge is significantly lower. It is assumed that these media
mainly consist of aluminum oxide.
AntiPhos lies in the same price range but showed significant lower loadings. This
might be the reason why the producer recommends for application a 10 times
higher amount for 200 L sea water than for all other media.
PhosEx has the lowest price of all media and also the lowest performance in terms
of loading and affinity. Removal of phosphate down to concentration lower than 500
micro g/L was not possible. High sorbent dosages might be necessary to achieve a
low P concentration in aquaria, which might increase costs for the process design
(larger adsorbers etc. The producer specifies that PhosEx would reduce the
phosphate concentration from 10 mg/L to 1 mg/L (concentration probably given as
phosphate). Summarizing, the zeolite based media AntiPhos and PhosEx are not
recommendable for the range of low P-concentrations. It is not likely that these
media show a better performance at higher P-concentrations.
For all media an impact on the pH of Millipore water was found. Although sea water
will be buffering, it is recommendable to rinse esp. RowaPhos, AntiPhos and
Phosphate
22.
Studies on phosphorus removal from fresh and sea water by commercial sorbents
Sponge before application. It is believed that all tested media contain heavy metals;
therefore the pH should be kept between 7.0 and 8.5 to prevent elution. AntiPhos
showed additionally increased conductivity after elution.
For the application in a fixed bed adsorber not only the sorption equilibrium is of
interest but also the sorption kinetics. This has to be considered for the design
layout, esp. regarding the contact time in the filter. More advanced and expensive
testing in pilot or full scale would be necessary to assess these parameters.
Considering the German drinking water standards (Table 7), the found values for
Cu, Pb, Zn and Ni are below these standards. For cadmium, the standard is lower
than the detection limit. The PhosEx and AntiPhos eluates showed a slightly
increased Mn value, which might be connected with the indefinite composition of
these media. For assessment of the eluted metal concentrations, the applied mass
of sorbent in relation to the water volume used in the sea water tank has to be
considered.
According to the producer's specifications (Table 1), the recommended mass/volume
ratios are around 100 times smaller than those used in the elution studies.
Therefore, the measured concentrations can be roughly divided by 100 in order to
obtain the effective - concentrations during application. Consequently, all metal
concentrations measured after elution would be below these standards. However,
some special fishes may be more sensitive to heavy metals than humans.
Table 7:
German drinking water standards according to the Trinkwasser- Verordnung (2001 )
Metal Cu Pb Zn Ni Cd Al Fe Mn
Concentration 2 0.01 - 0.02 0.005 0.2 0.2 0.05
(mg/L)
23.
Studies on phosphorus removal from fresh and sea water by commercial sorbents
7. Literature
DIN 38414 Teil 4, Schlamm und Sedimente (Gruppe S), Bestimmung der
Eluierbarkeit mit Wasser (S4), Deutsche Einheitsverfahren zur Wasser-, Abwasserund
Schlammuntersuchung, Normenausschuss Wasserwesen (NAW) im DIN
Deutsches institut fur Normung e.V., 1984
Driehaus, W. (1994). Arsenentfernung mit Magnesiumdioxid and Elsenhydroxid in
der Trinkwasseraufbereitung. Fortschritt-Berichte VDl Reihe15 Nr. 33, VDl Verlag,
Dusseldorf
Foss Tecator. Application Note. Bestimmung von Orthophosphat im Wasser mit dem
Fiastar 5000 (2001)
Grobkopf, J. (1998). Phosphat in der meeresaquaristischenPraxis. Der Meerwasseraquarianer,
2. Jahrgang, Januar 1998, Seiten 10-15
lSO/DIS 15681-1. Water quality -- Determination of orthophosphate and total
phosphorus contents by flow analysis (FIA and CFA) -- Part 1: Method by flow
injection analysis (FIA)
Langmuir, I. (1908). The adsorption of gases on plane surfaces of glass, mjca and
platinum. J. Ameri. Chem. Soc. 40,1361-1403
Persson, P., Nilsson; N., Sjoberg, S. (1996). Structure and Bonding of
Orthophosphate Ions at the Iron Oxide-Aqueous Interface. Journal of Colloid and
Interface Science 177, 263-275
Rietra,.R., Hiemstra, T. Van Riemsdijk, W. (2001) Interaction between Calcium and
Phosphate Adsorption on Goethite. Environ. Sci. Techn. 35, 3369-3374.
Sontheimer, H; Frick, B.;Fettig, J.; (1985). Adsorptionsverfahren zur
Wasseraufbereitung DVGW- Forschungsstelle am Engler- Bunte-Insitut. Karlruhe.
Stumm, W., Sigg, L. (1981), The interaction of anions and weak acids with the
hydrous goethite surface. Colloids and Surfaces Vol. 2 Issue 2, 101-117
Trinkwasserverordnung (2001). Verordnung zur Novellterung der Trinkwasserverordnung
yom 21.Mai 2001. Bundesgesetzblatt Jahrgang 2001 Teil1 Nr. 24
24.