Metallic iron for safe drinking water worldwide 1 
2 
3 
4 
C. Noubactep 
Angewandte Geologie, Universität Göttingen, Goldschmidtstraße 3, D - 37077 Göttingen, Germany. 
Kultur und Nachhaltige Entwicklung CDD e.V., Postfach 1502, D - 37005 Göttingen, Germany. 
5 
6 
7 
8 
9 
10 
11 
12 
13 
14 
15 
16 
17 
18 
19 
20 
21 
22 
23 
24 
25 
e-mail: cnoubac@gwdg.de; Tel. +49 551 39 3191, Fax. +49 551 399379 
 
(CEJ-D-10-01279R1: Accepted 2010/09/29) 
 
Abstract 
A new concept for household and large-scale safe drinking water production is presented. 
Raw water is successively filtered through a series of sand and iron filters. Sand filters mostly 
remove suspended particles (media filtration) and iron filters remove anions, cations, micro-
pollutants, natural organic matter, and micro-organisms including pathogens (reactive 
filtration). Accordingly, treatment steps conventionally achieved with flocculation, 
sedimentation, rapid sand filtration, activated carbon filtration, and disinfection are achieved 
in the new concept in only two steps. To prevent bed clogging, Fe0 is mixed with inert 
materials, yielding Fe0/sand filters. Efficient water treatment in Fe0/sand filters has been 
extensively investigated during the past two decades. Two different contexts are particularly 
important in this regard: (i) underground permeable reactive barriers, and (ii) household water 
filters. In these studies, the process of aqueous iron corrosion in a packed bed was proven 
very efficient for unspecific aqueous contaminant removal. Been based on a chemical process 
(iron corrosion), efficient water treatment in Fe0 beds is necessarily coupled with a slow flow 
rate. Therefore, for large communities several filters should work in parallel to produce 
enough water for storage and distribution. It appears that water filtration through Fe0/sand 
filters is an efficient, affordable, an flexible technology for the whole world. 
 1
Keywords: Contaminant removal, Conventional treatment, Iron bed, Media filtration, 
Reactive filtration, Zerovalent iron. 
26 
27 
28 
29 
30 
31 
32 
33 
34 
35 
36 
37 
38 
39 
40 
41 
42 
43 
 
 
“The close link between scientific research and technical invention appears to be a new factor 
in the nineteenth century. According to Mumford, “the principal initiatives came, not from the 
inventor-engineer, but from the scientist who established the general law”. The scientist took 
cognizance both of the new raw materials which were available and of the new human needs 
which had to be met. Then he deliberately oriented his research toward a scientific discovery 
that could be applied technically. And he did this out of simple curiosity or because of 
definite commercial and industrial demands. Pasteur, for instance, was encouraged in his 
bacteriological research by wine producers and silkworm growers. …In the twentieth century, 
this relationship between scientific research and technical invention resulted in the 
enslavement of science to technique”. 
Jacques Ellul 1954 [1] 
 
 
 2
1 Introduction 43 
44 
45 
46 
47 
48 
49 
50 
51 
52 
53 
54 
55 
56 
57 
58 
59 
60 
61 
62 
63 
64 
65 
66 
67 
68 
Universal access to safe drinking water is a challenge to the scientific community to which 
the responsibility is incumbent on developing appropriate technologies. Safe drinking water 
must be globally available, e.g. drinking water for (i) citizens in all cities, (ii) holiday-makers 
and tourists in Bamena (Cameroon), Dighali (Bangladesh), or Fuheis (Jordan), and (iii) 
villagers in small communities worldwide. Adequate infrastructures for safe drinking water 
production certainly exist in most cities and in hotels for holiday-makers and tourists. In 
general, it can be roughly considered that water infrastructure is well developed in urban areas 
as opposed to rural areas where the infrastructure is either poorly developed or non-existent. 
Accordingly, the population to be urgently deserved with safe drinking water is the rural one 
[2-4]. 
Rural communities are not similar in their size, technical equipment and geographic 
distribution [5-11]. For example, in the Republic of South Africa close to five million people 
live in widely scattered remote small rural communities of often less than 100 people [11]. 
Similarly, areas of small island countries in the Pacific Ocean are made up of hundreds of 
scattered islands inhabited by few people [8,9]. In the developed world, it is a well-known 
fact that small and remote communities often lack adequate technical, managerial, and 
financial capacity for safe drinking water production [6,12,13]. For example, around 2003, 
there were about 1,000 small municipal drinking water utilities in the Province of Quebec 
(Canada) whose waters were reported to frequently violate the provincial drinking water 
standards [6]. 
In many rural communities in the developing world, most people do not have paid 
employment. As a rule, any income comes from kindred in the family network that have jobs 
in the city or abroad. Most people live a sustainable agricultural, fishing or hunting existence 
and this paradigm of living results in little global influence [8,9,14]. However, global issues 
do impact on the villages, as their water is potentially contaminated with various imported 
 3
69 
70 
71 
72 
73 
74 
75 
76 
77 
78 
79 
80 
81 
82 
83 
84 
85 
86 
87 
88 
89 
90 
91 
92 
93 
94 
manufactured substances including fertilizers, heavy metals, herbicides, insecticides, and 
pharmaceutics [15,16]. Despite billions of dollars in aid, technology transfer, and local 
spending, inadequate progresses have been made in recent years in improving access to safe 
drinking water in the developing world [8]. It is presently not certain, whether the United 
Nations Millennium Development Goal of "halving by 2015, the proportion of people without 
sustainable access to safe drinking water” in 1990 will be achieved [17,18]. Even upon 
achievement, present technologies may still leave up to 600 million people without access to 
safe water in 2015 [4,19]. 
The presentation above clearly shows that providing people with safe drinking water is a 
human need of universal relevance. Therefore, efficient but affordable technologies are 
needed as the communities in need (small municipalities and villages) are characterized by 
their low-income [4,8,10,20]. Various technologies for safe drinking water provision are 
available but they are either cost-intensive or not applicable without electricity [11]. 
Accordingly, suitable technology should use low cost materials which are readily available 
and match or exceed the capability of conventional water treatment technologies: cue metallic 
iron (Fe0). 
1.1  Metallic iron in drinking water treatment plants 
The last two decades have witnessed the establishment of Fe0 as a powerful water remediation 
material for several classes of pollutants [21-25]. Fe0 is currently used in groundwater 
remediation and wastewater treatment [25-29], and drinking water production at household 
level [30-36]. Despite existing patents on water treatment using Fe0 in treatment plants 
[12,37,38], no concept comparable to the one presented here could be found. The process of 
Meng and Korfiatis [38] involves at least two other technical issues: (i) a vibration device to 
increase filtration efficiency in Fe0 beds and (ii) the addition of oxidizing agents or coagulants 
to enhance the filtration efficiency in sand beds. The process of Santina [37] termed as 
“sulfur-modified iron” uses finely-divided Fe0 in the presence of powdered S0 (or MnS), 
 4
followed by an oxidation step. The technical issues in both processes could cause managerial 
shortness in small communities. The chemical-free process for arsenic removal using a 
Fe
95 
96 
97 
98 
99 
100 
101 
102 
103 
104 
105 
106 
107 
108 
109 
110 
111 
112 
113 
114 
115 
116 
117 
118 
119 
120 
0/sand filter presented by Gottinger [12] is closer to the one presented here. However, 
Gottinger’s invention is a pragmatic one and is theoretically designed for As only. 
The present theoretical study is an extension of the recently presented concept of iron beds for 
safe drinking water production at household level [33,39,40], and community scale [41]. The 
technology uses Fe0 to assist slow sand filtration. Contaminant removal is primarily due to 
size exclusion. Size exclusion is improved by volumetric expansion/compression cycles 
inherent to aqueous iron corrosion and the adsorptive nature of iron corrosion products [41]. 
The whole dynamic process of iron corrosion is responsible for the Fe0 bed efficiency. 
Accordingly, the water flow velocity will be a function of the reactivity of the used material 
and the extent of contamination (contaminant nature and concentration). It can be emphasized 
that the water flow velocity is necessarily slow making Fe0 filtration technology an 
appropriate technology for small communities [12,33,41]. For large communities several 
filtration units could be necessary to produce the daily required water volume. Alternatively, 
any large community can be subdivided in several small water districts (e.g. urban quarters). 
The presentation will start by recalling the main characteristics of natural waters that serve as 
source for drinking water production.  
2. Characteristics of drinking water sources 
Fresh waters or raw waters are natural waters (surface water, groundwater) commonly used as 
drinking water sources. The four main characteristics of raw waters are their content of: (i) 
pathogens (e.g. bacteria, fungi, helminths, parasites, protozoa, viruses), (ii) natural organic 
material (NOM), (iii) dissolved salt (salinity), and (iv) H+ ions (pH value). 
It is universally acknowledged that the greatest risks of waterborne disease are from 
pathogens. The microbial risk is mostly due to water contamination by human and/or animal 
feces [4,42-44]. The existence and dangers of pathogenic microbes in surface waters have 
 5
121 
122 
123 
124 
125 
126 
127 
128 
129 
130 
131 
132 
133 
134 
135 
136 
137 
138 
139 
140 
141 
142 
143 
144 
145 
146 
been recognized for more than a century [45]. On the contrary, groundwaters (e.g. wells and 
springs) are naturally protected against contamination by pathogenic microbes. The protection 
is attributed to the filtration properties of subsurface soils and geologic strata. Accordingly 
groundwaters are generally less charged with pathogens [45,46]. 
NOM is ubiquitous in natural waters and can be present in dissolved, colloidal and/or 
particulate forms [47-49]. The dissolved and colloidal forms of NOM (DOM or fraction 
passing a 0.45-μm filter) are the most problematic and undesirable fractions of NOM with 
regard to water treatment (e.g. with granular activated carbon). DOM is a heterogeneous 
mixture of complex organic materials including humic substances, hydrophilic acids, proteins, 
lipids, carboxylic acids, polysaccharides, amino acids, and hydrocarbons. Due to its 
heterogeneous nature, a surrogate parameter such as total organic carbon (TOC) or dissolved 
organic carbon (DOC) is generally used to quantify DOM concentrations in water. 
Salinity is a general term used to describe the levels of different salts such as sodium chloride, 
magnesium and calcium sulfates, and bicarbonates. The salinity (or hardness) of fresh waters 
is generally due to calcium and magnesium and may take values of up to 300 mg/L as CaCO3 
[47]. The salinity is primarily a measure of ionic conductivity of a water and thus is important 
for sustaining aqueous Fe0 corrosion (transport of Fe2+ from anodic sites). 
The pH value is the most important characteristic of natural waters for the treatment with Fe0. 
As a rule, Fe0 oxidative dissolution yields soluble FeII and FeIII species at pH < 4.0 - 4.5 and 
precipitates of FeII, FeIII and FeII/FeIII at pH > 4.5 [50]. Water treatment by Fe0 filters is only 
possible at pH > 4.5. [33] The pH value of natural waters may range from 6 to 10 [47]. 
Accordingly, dissolved iron precipitates as hydroxides or/and oxides in the vicinity of the Fe0 
surface forming an initially porous oxide film. Variations in pH can change both the surface 
charge distribution of iron hydroxides/oxides and the ionization of weak acid or bases (pKa 
values), including DOM with various functional groups in its structure. The pH value can also 
impact the conformation of DOM components, and thus their adsorption onto the oxide scale 
 6
on Fe0. Finally, if water is contaminated by metals and metalloids, their speciation, their 
complex formation tendency, solubility and thus affinity to iron oxides is strongly pH 
dependant. 
147 
148 
149 
150 
151 
152 
153 
154 
155 
156 
157 
158 
159 
160 
161 
162 
163 
164 
165 
166 
167 
168 
169 
170 
171 
172 
In general, natural waters may be contaminated by: (i) natural organic matter (NOM and 
DOM), (ii) anthropogenic organic substances (fertilizers, pharmaceutical products, solvents), 
(iii) anthropogenic and geogenic inorganic substances (As, F, Fe, Mn, P, U) and (iv) micro-
organisms (including bacteria and viruses). All these substances or substance groups should 
theoretically be efficiently removed in Fe0 filters [33,51-56]. This theoretical prediction is 
validated by Fe0-based SONO filters designed for As which currently free water from more 
than 23 different metallic species, ammonia, bacteria, chloride, nitrates, and total coliform in 
Bangladesh and Nepal [30,55-57]. On the other hand, You et al. [35] and Diao and Yao [34] 
have explicitly demonstrated the suitability of Fe0 to inactivate micro-organisms and removed 
them from water. 
3 Metallic iron for drinking water treatment 
Metallic iron is an emergent reactive material increasingly used for water treatment [25,29, 
33,36]. Fe0 is the most used reactive material in subsurface permeable reactive barriers 
[29,34,58]. It was originally used to remove redox-sensitive contaminants from groundwater 
[26,27,58-60]. It is commonplace to consider that the bare Fe0 surface reacts with the 
contaminants and converts them into non-toxic/less toxic species (Assumption 1). The 
validity of Assumption 1 automatically degrades all other reducing agents (co-reductants) to 
side-reductants. Co-reactants are primarily dissolved and adsorbed FeII and H/H2. Assumption 
1 further requires that the Fe0 surface must be accessible. Accordingly, the universal film on 
Fe0 is regarded as inhibitive for the process of contaminant reduction or reductive removal. 
The assumed inhibitive characteristic of the oxide film coupled to the large variation in the 
reactivity of used iron fillings (micro-scale Fe0) were amount the major reasons to introduce 
nano-scale Fe0 [61,62], and bimetallic materials [63,64]. Nano-scale Fe0 has been reported to 
 7
have the potential to overcome these two problems. In fact, both the initial rates and the extent 
of contaminant reduction per mole of Fe
173 
174 
175 
176 
177 
178 
179 
180 
181 
182 
183 
184 
185 
186 
187 
188 
189 
190 
191 
192 
193 
194 
195 
196 
197 
198 
0 is increased. However, despite infinitesimally small 
size, nano-scale Fe0 will also be covered by an oxide film such that expected, observed and 
reported increase reactivity is not necessarily coupled to direct reduction by Fe0 (electrons 
from Fe0). 
It has already been demonstrated that the oxide film on Fe0 is beneficial for the process of 
contaminant removal [51-54,65]. In fact, the oxide film acts as contaminant scavenger and the 
contaminant may be further chemically transformed (oxidized or reduced if applicable). 
The presentation above clearly disprove the validity of Assumption 1 and corroborates results 
from other branches of science that iron corrosion is always coupled to oxide film formation 
at pH > 4.5. Iron corrosion continues under the film because the film is porous and permeable 
to water and other oxidizing agents [51,52,65-68]. More importantly, using Fe0 to assist sand 
filtration does not aim at inducing reductive transformations of contaminants, but to use the 
process of iron corrosion and the adsorptive properties of in situ generated iron oxides to 
sustain the filtration process. In other words, in Fe0 filters, Fe0 is oxidized by H2O and 
corrosion products are used as trap for contaminants which could be chemically transformed. 
Depending on their nature and concentration, selected contaminants may influence (inhibit or 
sustain) the process of iron corrosion. For example, it is well-established that the 
incorporation of a cation into the structure of iron (oxyhydr)oxides alters the nucleation, 
crystal growth, and transformation [69]. This impact of the alteration on the further iron 
corrosion should be carefully characterized in laboratory and field investigations. 
4 Mechanism of contaminant removal in Fe0 filters 
4.1 Aqueous iron corrosion 
Immersed reactive Fe0 corrodes due to differences in the electrical potential on anodic and 
cathodic sites on the Fe0 surface [70,71]. The metal oxidizes at the anode, where corrosion 
occurs according to equation 1: 
 8
Fe0 ⇔ Fe2+ + 2 e-      (1) 199 
200 
201 
202 
203 
204 
205 
206 
207 
208 
209 
210 
211 
212 
213 
214 
215 
216 
217 
218 
219 
220 
221 
222 
Simultaneously, a reduction reaction occurs at cathodic sites. The typical cathodic processes 
are: 
½ O2 + H2O ⇔ 2 e- + 2 OH-    (2) 
2 H+ + 2 e- ⇔ H2      (3) 
The electrons produced at anodic sites are conducted through the metal whilst the ions formed 
are transported via pore water (electrolyte). 
Fe2+ ions from Eq. 1 might be further oxidized (e.g. by O2, MnO2 or contaminants like CrO42-) 
to Fe3+ ions according to equation 4: 
Fe2+ ⇔ Fe3+ + e-      (4) 
On the other hand, Fe3+ from Eq. 4 is and oxidizing agent for Fe0 (Eq. 5): 
Fe0 + 2 Fe3+ ⇔ 3 Fe2+      (5) 
Lastly, generated Fe2+ and Fe3+ will form hydroxides according to equations 6 and 7 and the 
hydroxides will be progressively transformed to amorphous and crystalline oxides (Eq. 8): 
Fe2+ + 2 OH- ⇔ Fe(OH)2      (6) 
Fe3+ + 3 OH- ⇔ Fe(OH)3      (7) 
Fe(OH)2, Fe(OH)3 ⇒ FeOOH, Fe2O3, Fe3O4  (8) 
It is important to recall that, regardless from the size of the material, the formation and 
transformation of an oxide scale on Fe0 is an universal process (at pH > 4.5). Accordingly, the 
propensity of any Fe0 to aqueous corrosion is influenced by the nature of the oxide scale on its 
surface. The initial oxide scale is porous and permeable but may be transformed to an 
impervious layer mainly depending on the water chemistry [72]. In other words, an oxide 
scale begins to form immediately after Fe0 immersion and may facilitate or hinder corrosion, 
serving either as a barrier or as a path for ion exchange with the pore solution in the Fe0 filter. 
 9
It is also important to notice that the reactions after equations 2 through 8 are considered as 
side reactions in the discussion of the process of contaminant reductive transformation by Fe
223 
224 
225 
226 
227 
228 
229 
230 
231 
232 
233 
234 
235 
236 
237 
238 
239 
240 
241 
242 
243 
244 
245 
246 
247 
248 
0 
(Fe0 is ideally oxidized by the contaminant). This simplification is not acceptable for at least 
two reasons: (i) even though a contaminant (e.g. CrVI) may be a stronger oxidizing agent for 
Fe0 than Fe3+, H+ and O2, H2O (H+) is present in very large stoichiometric abundance; (ii) 
even if the universal oxide scale is considered as a path for contaminant transport, it is a 
reactive path containing species (e.g. adsorbed FeII) which are sometimes more powerful 
reducing agents than Fe0 [73]. The next section discuss the mechanism of contaminant 
removal in Fe0 filters. 
4.2 Mechanistic aspects of contaminant removal in Fe0 filters 
4.2.1 Contaminant removing processes 
Adsorption is characterized by the accumulation of substances at the interface between two 
phases (e.g. solid/liquid) due to chemical and physicochemical interactions. The solid on 
which adsorption occurs is called the adsorbent. In a Fe0/H2O system, there are theoretically 
several adsorbents including Fe0, Fe(OH)2, Fe(OH)3, FeOOH, Fe3O4, Fe2O3, and green rust. 
However, apart from Fe0 and well crystallized phases (Fe2O3, Fe3O4), all other solid phases 
are in transformation. During the precipitation of Fe phases contaminants can be enmeshed. 
Whether precipitates are pure phases or not, there is no defined surface on which an inflowing 
contaminants could adsorb. 
The transport of any contaminant (chemical, microbial and physical) in a Fe0 filter is 
primarily controlled by its physico-chemical characteristics, the composition of the water, the 
characteristics of available adsorbents, and the water flow velocity. Two key contaminant 
characteristics are size and surface electrostatic properties. Key properties of Fe0 beds include 
water flow velocity which is coupled to pore size distribution, temperature, pH, and 
chemical/mineral composition of water. From a pure physical perspective, contaminant size, 
bed porosity and relative surface electrical properties of Fe species and contaminants are the 
 10
most important factors governing the efficiency of a Fe0 bed. Accordingly, large contaminants 
with higher affinity to Fe
249 
250 
251 
252 
253 
254 
255 
256 
257 
258 
259 
260 
261 
262 
263 
264 
265 
266 
267 
268 
269 
270 
271 
272 
273 
274 
0 species should be readily removed in Fe0 beds. In other words, if 
adsorption was the most important mechanism, contaminant should be reduced (or oxidized) 
to yield species which should readily be adsorbed by Fe0 species. However, the technology 
was developed to reductively degrade polar halogenated organic species (RX), ideally to non-
polar organic species (RH). Actually, Fe oxides are mostly polar and will readily interact with 
other polar species. Given that reduced species are reported to be removed from the aqueous 
phase, it is obvious that adsorption on corrosion products alone can not explain the reported 
efficiency of Fe0 beds. 
It should be explicitly pointed out, that the goal of water treatment for save water production 
is not contaminant chemical transformation (reduction or oxidation) but contaminant removal. 
Accordingly, even reduced contaminants (e.g. RH) must be removed from the aqueous phase. 
In other words, any contaminant and all its transformation products have to be removed from 
the aqueous phase. For example, the reductive transformation of carbon tetrachloride by Fe0 
has been reported to produce hexachloroethane, tetrachloroethane, trichloromethane, 
dichloromethane, carbon monoxide, carbon dioxide, and methane [74]. Apart from non toxic 
CO2, all these reaction products should be removed from water to obtain safe drinking water. 
Removal mechanisms in this context are adsorption and co-precipitation (also termed 
enmeshment, entrapment or sequestration). 
This section hat demonstrated that, from a pure semantic perspective contaminants are 
removed in Fe0/H2O systems by adsorption and co-precipitation. 
4.2.2 Bed porosity and porosity loss 
A Fe0 filtration bed is composed of one or several reactive zones of granular sand and Fe0 
particles (Fig. 1). The compact Fe0:sand mixture has a random porous structure. The manner 
with which the pore space is formed depends mainly on the arrangement of the granular 
particles [75,76]. While packing uniform spheres, the least compact and most compact 
 11
arrangements are rhombohedral and cubic respectively. The pore size can be defined in terms 
of a length dimension (pore radius). Pore size in a packed bed is closely related with the size 
of the filter grains constituting the bed (e.g. Fe
275 
276 
277 
278 
279 
280 
281 
282 
283 
284 
285 
286 
287 
288 
289 
290 
291 
292 
293 
294 
295 
296 
297 
298 
299 
300 
0 and sand). The smaller the grains, the smaller 
the pore size.  
The most important feature of Fe0 filters is the evolution of the initial porosity with the extent 
of volumetric expansive Fe0 corrosion [77] and its consequence for the process of contaminant 
removal. It has been shown that if a filter contents less that 50 vol-% of Fe0, no clogging 
(residual porosity = 0) will occur upon Fe0 depletion [78]. In all the cases, a progressive 
diminution of pore radius will be observed. Assuming a purposeful selection of Fe0 and sand 
grain size and a relevant Fe0 volumetric ratio in a filter, the processes yielding contaminant 
removal in Fe0 beds are discussed below. 
4.2.3 Mechanism of contaminant removal in Fe0 filters  
In the conventional granular bed filtration (adsorptive filtration), contaminants have to be 
transported near the filter grains (e.g. activated carbon, metal oxide) by different transport 
mechanisms and then adhered to the grain surfaces by various attachment mechanisms for 
their successful removal [79,80]. Filtration is thus a complex process involving physico-
chemical mechanisms and essentially depending on four major various factors: (i) filtration 
rate, (ii) media grain size, (iii) affinity of contaminant to bed media, and (iv) contaminant 
concentration. Depending from the media grain size and the size of the contaminant, a 
filtration bed may work as pure sieve (size exclusion). Size exclusion is used for example in 
rapid sand filtration for water clarification. 
4.2.3.1 Adsorptive filtration and reactive filtration 
Conventional filters contains adsorptive media (e.g. iron oxides) with are inert in water and 
posses a given adsorptive capacity for any contaminant. Accordingly, a contaminant 
breakthrough is observed when the adsorptive capacity of the material in the filter is 
exhausted. In a Fe0/sand bed on the contrary, iron oxides for contaminant adsorption are 
 12
generated in-situ. Ideally, iron oxide generation through Fe0 oxidation H2O (or H+) occurs 
uniformly in the whole bed (Fig. 2). Therefore, although a reaction from exists due to dissolve 
O
301 
302 
303 
304 
305 
306 
307 
308 
309 
310 
311 
312 
313 
314 
315 
316 
317 
318 
319 
320 
321 
322 
323 
324 
325 
326 
2, salinity and probably contamination, virgin Fe0 can not be expected in a Fe0 filter 
(reactive filtration). Accordingly at any date contaminant removal occurs in the whole bed 
and iron corrosion proceeds in all three compartment of the bed. The best illustration for this 
is given by a experiment of Leupin and Hug [81]. The authors performed an As removal 
experiment with four identical filters in series containing each 1.5 g Fe0 and 60 g sand. The 
results showed that 36 L of water containing 500 μg As/L could be treated to below 50 μg/L 
arsenic. This performance resulted from multiple filtrations, showing that contaminant 
removal occurs in the whole bed. The difference between synthetic iron oxides and in-situ 
generated iron oxides (corrosion products) is excellently given by Sikora and Macdonald [82] 
and presented elsewhere in the context of safe drinking water production [33]. The further 
presentation will insist on the transformation of iron from its position in the metal lattice (Fe0) 
to its location in a crystallized corrosion products (e.g. FeOOH, Fe2O3, Fe3O4). 
4.2.3.2 The volumetric expansion/compression cycle 
The essential characteristic of a Fe0 filtration bed is the in-situ generation of very adsorptive 
iron hydroxides which are progressively transformed to amorphous and crystalline iron 
oxides. While filling the pore space, solid corrosion products necessarily reduce the pore 
radius, improving size exclusion but the most important feature is the dynamic nature of iron 
corrosion in the pore space (Fig. 3). Iron corrosion products could be regarded as 
“mercenaries” with the mission to trap contaminants in the pore space of the bed. 
Accordingly, contaminants should not be transported near the Fe0 grains to be removed.  
The cycle of a single atom (Fe0) in the process of iron corrosion can be given as follows: 
Fe0 ⇒ Fe2+/Fe3+(H2O)6 ⇒ Fe2+/Fe3+(OH)n ⇒ FeOOH ⇒ Fe2O3 ⇒ Fe3O4  (9) 
While only considering insoluble species the cycle is: 
Fe0 ⇒ Fe(OH)2/Fe(OH)3 ⇒ FeOOH ⇒ Fe2O3 ⇒ Fe3O4   (10) 
 13
The transformation can also be represented in terms of variation of the specific surface area 
(SSA in m
327 
328 
329 
330 
331 
332 
333 
334 
335 
336 
337 
338 
339 
340 
341 
342 
343 
344 
345 
346 
347 
348 
349 
350 
351 
352 
2/g - Fig. 3a). Selected representative values are given in parenthesis, Fe2(HO)6 
stands for ferrihydrite [83]. 
Fe0 (1) ⇒ Fe2(HO)6 (327) ⇒ FeOOH (55) ⇒ Fe2O3 (11) ⇒ Fe3O4 (2)  (11) 
The last alternative to represent the transformation is in terms volumetric expansion relative 
to Fe0 in the metal lattice. The coefficient of volumetric expansion given in parenthesis is 
equal to Voxide/VFe [77]. The following evolution is given (Fig. 3b): 
Fe0 (1) ⇒ Fe2(HO)6 (6.4) ⇒ FeOOH (3.0) ⇒ Fe2O3 (2.2) ⇒ Fe3O4 (2.1)  (12) 
The evolution of the surface area, the density and the coefficient of volumetric expansion 
clearly show that dissolved Fe first experiences an expansion than a compression. Focussing 
the attention on the initial stage (Fe0) and the final stage (FeOOH, Fe2O3 or Fe3O4) reveal an 
expansion which is definitively the reason for porosity loss. However, the whole dynamic 
process of iron corrosion should be considered. In particular, if there is not enough space for 
volumetric expansion, iron corrosion will stop. This is the very first argument against a 100 % 
Fe0 reactive zone of 100 % Fe0 filtration bed as used in the 3-Kolshi system [32,84,85]. 
Therefore, Leupin et al. [81,86] suggested the admixture of inert sand to Fe0 as an efficient 
tool to ameliorate the efficiency of iron filters. Recent calculation [78] suggested that a 
reactive zone with more that 60 vol-% Fe0 should be regarded as pure material wastage 
because corrosion will stop because of lack of space to proceed. It is important to notice that 
mixing Fe0 and inert materials (e.g. sand) is a prerequisite for long term reactivity (and 
permeability). Accordingly, the resulting economy in investment costs (costs for the 
corresponding 30 vol-% Fe0) could be regarded as beneficial side effects. 
4.2.3.3 Expansion/compression cycles and contaminant removal 
The transformations accompanying Fe0 transformation to crystallized iron oxides may occurs 
in the presence of contaminants which may be trapped or enmeshed in the mass of corrosion 
products or be retained in the filter by size exclusion. The efficiency of a Fe0 filter for 
 14
contaminant removal can be summarized in the following metaphor: Instead of waiting for the 
contaminants to come to its surface, Fe
353 
354 
355 
356 
357 
358 
359 
360 
361 
362 
363 
364 
365 
366 
367 
368 
369 
370 
371 
372 
373 
374 
375 
376 
377 
378 
0 injects corrosion products into the pore space for 
rapid and effective contaminant removal. A further abstraction is to consider that the pore 
space is initially filled with porous amorphous iron hydroxides and oxides which are 
progressively transformed to more crystalline species. 
It is important to note that the presentation above has not considered the nature of the 
contaminants (bacteria or virus, chemical or microbial, organic or inorganic). Accordingly, 
even contaminants with less adsorptive affinity to iron oxides like MoVI [87,88], will be 
transported in the filter by gravity and removed by pure size exclusion. Specific laboratory 
researches are nevertheless needed for such contaminants. These studies may check the 
possibility to add a layer of adequate reactive materials (e.g. MnO2 or natural zeolithe for 
MoVI) for the specific removal of such contaminants before or after Fe0 filtration. 
5 The singularity of Fe0/H2O systems 
A natural water may contain three main types of contaminants that should be removed in any 
efficient treatment plant [19,43,44]: (i) chemical contaminants from natural or anthropogenic 
sources (organic and inorganic species), (ii) microbial contaminants which are the significant 
cause of water-borne diseases and is often associated with faecal matter, and (iii) physical 
contaminants such as taste, odour, colour, turbidity, and temperature. Physical aspects may 
not necessarily have any direct health effects but their presence in water may cause rejection 
by consumers. 
5.1 A Macroscopic view 
Given the large array of available contaminants (charged/non-charged, negatively 
charged/positively negatively charged, polar/non-polar, reducible/non-reducible, small/large) 
the question arises how a Fe0/H2O system could efficiently remove all these contaminants (at 
pH > 4.5). The answer is given by a profound observation of the evolution of Fe0 is aqueous 
solution, for example under anoxic conditions. Water (H2O) is the most abundant and 
 15
important oxidant and magnetite (Fe3O4) the main corrosion product. On the macroscopic 
scale, the original Fe
379 
380 
381 
382 
383 
384 
385 
386 
387 
388 
389 
390 
391 
392 
393 
394 
395 
396 
397 
398 
399 
400 
401 
402 
403 
404 
0 is progressively transformed to crystalline oxides (Fe3O4). The original 
Fe0 surface is progressively covered by a Fe3O4 film. Magnetite films a have a polycrystalline 
structure. The grain boundary diffusion coefficients of the reactant (H2O) and products (H2/H 
and FeII) moving through the corrosion film are between two and three orders of magnitude 
greater than the corresponding bulk diffusion coefficients. Therefore, grain boundary 
diffusion takes place rather than bulk diffusion ([89] and ref. therein). This is a major 
argument supporting the view that contaminant removal occurs within the oxide-film.  
The rate-limiting step in process of Fe0 oxidative dissolution at pH > 4.5 been the diffusive 
transport of reactants and products through the layer of Fe3O4, the nature of the contaminant 
necessarily plays a secondary role. The thickness of the Fe3O4 film influences the inward 
diffusion of oxidants (H2O, O2, contaminants) and the outward diffusion of corrosion products 
(H/H2, FeII), therefore, as the film grows, the rate of corrosion should become attenuated but 
also the rate of contaminant transport. Again, the nature of the contaminant is not yet 
addressed. Accordingly, contaminant are encapsulated within the oxide film irrespective from 
their intrinsic properties. It is evident that contaminants having stronger interactions with iron 
oxides will be sooner and stronger bounded. But chemical, microbial and physical 
contaminants are fundamentally removed. Properly dimensioning will certainly yield efficient 
water treatment. In other words, the relative affinity of contaminants for corrosion products 
may determine the bed thickness. To further understand the primarily unspecific nature of 
contaminant removal, the process or iron oxide generation will be considered on a 
macroscopic scale. 
5.2 A Microscopic view 
Iron oxyhydroxides (akaganeite, goethite, lepidocrocite), iron hydrous oxides (ferrihydrite, 
hydrohematite, maghemite) and iron oxides (hematite, magnetite) are known for their 
tendency to nucleate and grow on the surfaces of other phases [90]. In nature iron is leached 
 16
as FeII from iron minerals (e.g. pyrite), nucleates and grows as Fe phases on the available 
surfaces. In a engineered Fe
405 
406 
407 
408 
409 
410 
411 
412 
413 
414 
415 
416 
417 
418 
419 
420 
421 
422 
423 
424 
425 
426 
427 
428 
429 
0/H2O system, FeII is generated by the oxidative dissolution of 
Fe0, nucleation and growth occur at the Fe0 surface, the external or internal surface in-situ 
generated Fe phases (oxide scale) or the surface of additive materials (sand, pumice) [91].  
At neutral pH (7.0) and under anoxic conditions, leached FeII has a relatively high solubility. 
The saturation concentration of the FeII species approaches 0.5 M (28 g/L) at room 
temperature. However, dissolved Fe concentrations drop below 10-12 M (5.6*10-11 g/L) if FeII 
is polymerized e.g. [Fe(OH)2]n or oxidized to FeIII species [92]. This reaction results in the 
precipitation of Fe phases. FeIII species (and Fe phases) can also form under anoxic 
conditions, for example, when water contents oxidizing species like NO3- or MnO2. FeII 
oxidation and Fe phase formation is often catalyzed by microorganisms. 
FeIII phases which form from solution begin as small clusters that evolve into larger polymeric 
units with time, eventually reaching colloidal sizes [93]. Aggregation and/or crystal growth 
are necessarily coupled to the decrease in surface energy. All these processes occur millions 
of times in a Fe0 filter and in the presence of contaminants. As shown above contaminants are 
constrained to move to the oxide films of porous iron. Despite possible low affinity to the 
oxide film, contaminant could be retained by size exclusion. This size exclusion also happen 
hundreds of times in a Fe0 filters. Contaminants escaping in the entrance zone are possibly 
entrapped deeper in the filter. This argument supports the assertion that thicker beds will be 
necessary to satisfactorily remove contaminants with poor affinity to iron oxides. 
5.3 Reactive filtration on Fe0 beds 
The presentation above has strengthened the view that the efficiency of Fe0 filters is due to the 
progressive production of very reactive Fe phases that are in-situ further transformed. 
Accordingly, unlike in iron-coated system where the capacity of used iron oxides can be 
evaluated, a Fe0 filter is a system producing very reactive Fe0 which is in-situ transformed to 
 17
430 
431 
432 
433 
434 
435 
436 
437 
438 
439 
440 
441 
442 
443 
444 
445 
446 
447 
448 
449 
450 
451 
452 
453 
454 
455 
less reactive comparable to coated ones. The less reactive species are comparative to coating 
but contaminants are removed during their formation.  
5. Filter and plant design 
5.1 Filter design 
Fe0 filtration beds should remove trace amounts of chemical contaminants and pathogens 
from raw water to produce safe drinking water. The filter efficiency depends upon the 
purposeful selection of a reactive medium (Fe0) and the water flow velocity. The water flow 
velocity will depend on the intrinsic reactivity of Fe0 as the residence time should correspond 
to the time necessary to produce enough iron corrosion for contaminant removal by (i) 
adsorption, (ii) co-precipitation and (iii) size exclusion.  
It should be explicitly said that the goal should never be to select (or manufacture) the most 
reactive material but a material which is reactive enough to produce enough water for the 
community in need within a reasonable time. For example, a Fe0 material that is not reactive 
enough for a water plant in Freiberg (Sachsen) or Krebeck in Germany could be satisfactorily 
for a plant in a tropical village in Indonesia, Malawi, Nigeria, Peru, Senegal or Zambia. 
Nevertheless having readily reactive materials has the advantage to offer a flexibility in 
selecting the amount to be used in individual cases. For example, only 35 vol-% of a very 
reactive material could be used under tropical conditions and up to 55 vol-% under temperate 
conditions. Varying the reactivity of Fe0 materials by varying their particle size from fine 
powders to large granules and chips will be a tool in optimising filtration efficiency. 
5.2  Plant design 
A water treatment plant based on Fe0 bed filtration is very simple and similar to slow sand 
filtration for small communities. The simplest device is a single column containing layers of: 
(i) gravel and sand for water clarification (media filtration) and (ii) Fe0:sand for water 
treatment (reactive filtration). This device is similar to household Fe0-based SONO filters 
[30,94] which have been reported to function for more than five years. The long-term 
 18
efficiency of SONO filters is certainly due to the porous nature of used composites. In fact, 
using non-porous Fe
456 
457 
458 
459 
460 
461 
462 
463 
464 
465 
466 
467 
468 
469 
470 
471 
472 
473 
474 
475 
476 
477 
478 
479 
480 
481 
0 was efficient but not sustainable [94] and mixing non-porous Fe0 with 
inert materials has been theoretically [39] and experimentally [91] proven to sustain Fe0 
efficient filtration. 
In the first stage, a good target for Fe0 communities filters could be to design a filter which 
could be efficient for 12 months using locally available Fe0 materials, e.g. construction steel. 
For larger communities, a device might comprises one or two sand filters (media filtration) 
and one or several iron filters (reactive filtration) in series (Fig. 1, Fig. 4). A practical 
suggestion from experiences with 3-Kolshi filters in Bangladesh is to make small filters with 
20 L/h flow and connect several in parallel to scale-up (Hussam 2010, personal 
communication). For example if 100 L drinking water should be produce each hour (2400 
L/day), 5 such small filters should be used in parallel (Fig. 4). The production plan may 
comprise a total of 25 small filters to assure continuous water production. Some filters could 
be fixed while others worked, for example for maintenance, reparation or Fe0 replacement. 
Water for filtration could be first collected or pumped in a tank. Raw water is possibly 
chemically and microbially contaminated. Raw water is filtered through the Fe0/sand bed and 
contaminants are removed by several mechanisms including adsorption, co-precipitation, 
precipitation, reduction, size exclusion and combinations thereof. Filtered water could be 
stocked in a tank for distribution. In pilot plans for large communities several sets of raw 
water/drinking water tanks should be linked by various embodied filtration beds (Fig. 5).  
6 The economics of Fe0/sand beds 
Fe0 bed filtration as stand-alone remediation technology has already been proven affordable 
both at household [30,31,36] and at community [12,37,38] level. The costs are further reduced 
by admixing Fe0 with inert materials [39,91] and eliminating some technical steps as 
discussed in section 1.1. Accordingly the presentation will be limited in discussing the 
economics of Fe0 beds for a small community. 
 19
Cost is a major factor in implementing Fe0 filtration technologies and is necessarily site-
specific. Factors determining water treatment cost in Fe
482 
483 
484 
485 
486 
487 
488 
489 
490 
491 
492 
493 
494 
495 
496 
497 
498 
499 
500 
501 
502 
503 
504 
505 
506 
507 
0 beds include: (i) the quality of 
freshwater, (ii) plant capacity, and (iii) construction costs. The realistic costs of Fe0/sand 
filters given by Gottinger [12] could be adopted here. The estimation is based on the evidence 
Fe0 is the sole material to be bought. In Canada, Fe0 filings can currently be obtained for 
under $1.50 / kg (1.12 €/kg). The cost of manufacturing Fe0/sand filter is comparable to a 
biological activated carbon filter. The service life of a Fe0/sand filter (50 vol-% Fe0) was 
estimated to be approximately 40 months (3.3 years) [12]. This yield to a treatment cost of < 
0.01 $/L (< 0.01 €/L) and includes filter installation, media, operation and maintenance costs.  
It is not likely that any water treatment process addressing both chemical and microbial 
contamination could be cheaper than the own presented here. Calculations made for a model 
village in South Africa (125 inhabitants, 40 L water/person/day or 5,000 L/day) showed that 
200 kg Fe0 could be sufficient to produce safe drinking water for 3 years. The cost for the 200 
kg Fe0 is only 224 €. For comparison, a disinfection system recently presented as affordable 
for rural South Africa [11] has a capital cost of 900 R (82 €) for the same population size, the 
monthly running cost for disinfection was up to 150 R (1,800 R/year or 5,400 R for 3 years, 
that is 570 € for 3 years). Thereby disinfection by chlorination has been proven harmful for 
humans [95-99]. This comparison confirms that Fe0/sand filtration beds represent an efficient 
and economically feasible option for safe drinking water production for small-scale utilities. 
7 Concluding remarks 
Fe0/sand filtration is an affordable technology for safe drinking water production at various 
scales: household, rural establishments (clinics, forestry stations, hospitals, hotels, schools), 
and small or large communities. Fe0/sand filtration is the ideal technology for remote villages 
in the developing world. Here inhabitants may lack money to purchase industrially 
manufactured Fe0 (no income) but they possess the ancestral iron-making technology 
[100,101]. However, manufactured materials should be tested for reactivity and rural 
 20
populations should be trained for filter design. It could be anticipated that, self produced safe 
drinking water will increase the self-confidence of these populations and contribute to reduce 
rural exodus. On the other hand the development and the implementation of the technology 
worldwide will render travel with bottle water superfluous. Moreover, Fe
508 
509 
510 
511 
512 
513 
514 
515 
516 
517 
518 
519 
520 
521 
522 
523 
524 
525 
526 
527 
528 
529 
530 
531 
532 
0/sand filters/beds 
are excellent candidates for safe drinking water in emergency situations (e.g. earthquakes, 
wars, tsunami) [78]. 
Fe0/sand filtration is equally a feasible option to successfully remove all target compounds 
from surface and groundwater: particles, natural organic matter, pathogens and micro-
pollutants (including so-called emerging contaminants). Therefore, efforts should be made to 
use this chemical-free technology as the first choice everywhere. It could be expected that 
using Fe0/sand filtration as standard technology will be very beneficial for water works as iron 
oxides (the products of iron corrosion) are easy to recycle to Fe0. However, given the large 
spectrum of toxic substances enmeshed in the mass of corrosion products, the recycling task 
should be carefully addressed. In particular used household filters should be collected and 
professionally disposed (or recycled). This approach has the great advantage to control filter 
residue in regions where water is contaminated by toxics species like arsenic or uranium. On 
the other hand, recycling Fe-based materials (not only filter residues) for Fe0 production will 
generate incomes in developing country while protecting the environment. 
The probably strongest argument for the development of Fe0/sand filtration technology is the 
simplicity of the system. One should not care in parallel for membranes, granular activated 
carbon and chemicals (including disinfectants) but only on the stock of iron and sand, and the 
regeneration of the former. Finally, it can be speculated the success of the Fe0/sand 
technology for safe drinking water production will depend on the capacity of researchers 
create new reactive Fe0 materials and their capacity to find ways to control material reactivity 
in an affordable way. 
 21
Currently, there is an ongoing discussion on the suitability of localized solutions for save 
drinking water production [102-104]. In the developed world, decentralized drinking water 
production units are progressively regarded as an efficient alternative to centralized 
production systems. This approach is discussed in analogy to the energy sector where 
decentralization has raised a surge in innovations and new market opportunities for a host of 
new and established companies [104]. In a similar way it could be expected that localized 
water treatment would produce new jobs, businesses, economic development, and quality of 
life. Efficient and affordable technologies for decentralized water treatment are needed to 
sustain the decentralization argumentation. The present communication and related works 
[12,33,39-41] have presented Fe
533 
534 
535 
536 
537 
538 
539 
540 
541 
542 
543 
544 
545 
546 
547 
548 
549 
550 
551 
552 
553 
554 
555 
556 
557 
0 bed filtration as a serious candidate to be systematically 
assessed. 
Acknowledgments 
Sven Hellbach (student research assistant) is acknowledged for technical assistance. The 
manuscript was improved by the insightful comments of anonymous reviewers from 
Chemical Engineering Journal. 
References 
[1] J. Ellul, The technological society. New York: Vintage Books. (1964), cited by ref. [20] 
[2] J. Albert, J. Luoto, D. Levine, End-User Preferences for and Performance of Competing 
POU Water Treatment Technologies among the Rural Poor of Kenya. Environ. Sci. Technol. 
44 (2010) 4426–4432. 
[3] G. Concha, K. Broberg, M. Grandr, A. Cardozo, B. Palm, M. Vahter, High-level exposure 
to lithium, boron, cesium, and arsenic via drinking water in the Andes of northern Argentina. 
Environ. Sci. Technol. 44 (2010) 6875–6880. 
[4] T.F. Clasen, Household water treatment and the millennium development goals: Keeping 
the focus on health. Environ. Sci. Technol. (2010) doi: 10.1021/es1018674. 
 22
558 
559 
560 
561 
562 
563 
564 
565 
566 
567 
568 
569 
570 
571 
572 
573 
574 
575 
576 
577 
578 
579 
580 
581 
582 
[5] J.B. Braden, P.C. Mankin, Economic and financial management of small water supply 
systems: Issue Introduction. J. Contemp. Water Res. Ed. 128 (2004) 1–5. 
[6] H.D. Coulibaly, M.J. Rodriguez, Development of performance indicators for small 
Quebec drinking water utilities, J. Environ. Manag. 73 (2004) 243–255. 
[7] J. Maras, Economic and financial management capacity of small water systems, J. 
Contemp. Water Res. Ed. 128 (2004) 31–34. 
[8] W.J. Smith Jr, The place of rural, remote and least-wealthy small islands in international 
water development: the nexus of geography–technology sustainability in Chuuk State, 
Federated States of Micronesia, The Geogr. J. 174 (2008) 251–268. 
[9] W.J. Smith Jr, Improving access to safe drinking water in rural, remote, and least wealthy 
small islands: Non-traditional methods in Chuuk State, Federated States of Micronesia, 
Int. J. Environ. Waste Manag. 10 (2009) 167–189. 
[10] M.N.B. Momba, C.L. Obi, P. Thompson, Survey of disinfection efficiency of small 
drinking water treatment plants: Challenges facing small water treatment plants in South 
Africa, Water SA 35 (2009) 485–494. 
[11] D.L. Key, J.D. Key, G. Okolongo and M. Siguba, Development of a small-scale electro-
chlorination system for rural water supplies. Report No. 1442/1/09 to the Water 
Research Commission, (ISBN 978-1-77005-916-0), 2010. 
[12] A.M. Gottinger, Master thesis, University of Regina (Saskatchewan, Canada), 2010. 
[13] A.M. Gottinger, D.J. Wild, D. McMartin, B. Moldovan and D. Wang, Development of an 
iron-amended biofilter for removal of arsenic from rural Canadian prairie potable 
water, In: Water Pollution X. A.M. Marinov and C.A. Brebbia, Eds.; WIT Press: 
Ashurst, Southampton, 2010, 333–344. 
[14] S.F. Bagi, Small rural communities’ quest for safe drinking water, Rural America 17 
(2002) 40–46. 
 23
583 
584 
585 
586 
587 
588 
589 
590 
591 
592 
593 
594 
595 
596 
597 
598 
599 
600 
601 
602 
603 
604 
605 
606 
607 
[15] K. Ballschmiter, R. Hackenberg, W.M. Jarman, R. Looser, Man-made chemicals found 
in remote areas of the world: The experimental definition for POPs, Environ. Sci. Pollut. 
Res. 9 (2002) 274–288. 
[16] O. Osibanjo, Organochlorines in Nigeria and Africa. In The Handbook of Environmental 
Chemistry Vol. 3, Part O, H. Fiedler (ed.), Springer-Verlag Berlin Heidelberg, 2003, 
321–354. 
[17] A. Loftus, Rethinking political ecologies of water, Third World Quart. 30 (2009) 953–
968. 
[18] E. Gerlach, R. Franceys, "Standpipes and beyond" - a universal water service dynamic. J. 
Int. Dev. (2009), doi 10.1002/jid.1567. 
[19] WHO/UNICEF, Progress on sanitation and drinking-water 2010 update, Geneva, 
Switzerland: WHO/UNICEF 2010. 
[20] W.J. Smith Jr., Geographic research in water resources: a vibrant research agenda for the 
next 20 years, J. Contemp. Water Res. Ed. 142 (2009) 83–89. 
[21] R.W. Gillham, Cleaning halogenated contaminants from groundwater, U.S. Pat. 
5266213, 1992. 
[22] D.W. Blowes, C.P. Ptacek, System for treating contaminated groundwater. U.S. Patent 
5362394, 1994. 
[23] S.F. O'Hannesin, R.W. Gillham, Long-term performance of an in situ "iron wall" for 
remediation of VOCs Ground Water 36 (1998) 164–170. 
[24] M.M. Scherer, S. Richter, R.L. Valentine and P.J.J. Alvarez, Rev. Environ. Sci. Technol., 
Chemistry and microbiology of permeable reactive barriers for in situ groundwater clean 
up, 30 (2000) 363–411. 
[25] A.D. Henderson, A.H. Demond, Long-term performance of zero-valent iron permeable 
reactive barriers: a critical review, Environ. Eng. Sci. 24 (2007), 401–423. 
 24
608 
609 
610 
611 
612 
613 
614 
615 
616 
617 
618 
619 
620 
621 
622 
623 
624 
625 
626 
627 
628 
629 
630 
631 
632 
633 
[26] A.B. Cundy, L. Hopkinson, R.L.D. Whitby, Use of iron-based technologies in 
contaminated land and groundwater remediation: A review, Sci. Tot. Environ. 400 
(2008), 42–51. 
[27] R. Thiruvenkatachari, S. Vigneswaran, R. Naidu, Permeable reactive barrier for 
groundwater remediation, J. Ind. Eng. Chem. 14 (2008), 145–156. 
[28] J.P. Muegge, P.W. Hadley, An evaluation of permeable reactive barrier projects in 
California, Remediation 20 (2009), 41–57. 
[29] D.H. Phillips, T. Van Nooten, L. Bastiaens, M.I. Russell, K. Dickson, S. Plant, J.M.E. 
Ahad, T. Newton, T. Elliot and R.M. Kalin, Ten year performance evaluation of a field-
scale zero-valent iron permeable reactive barrier installed to remediate trichloroethene 
contaminated groundwater, Environ. Sci. Technol. 44 (2010), 3861–3869. 
[30] A. Hussam, A.K.M. Munir, A simple and effective arsenic filter based on composite iron 
matrix: Development and deployment studies for groundwater of Bangladesh, J. 
Environ. Sci. Health A 42 (2007), 1869–1878. 
[31] T.K.K. Ngai, R.R. Shrestha, B. Dangol, M. Maharjan, S.E. Murcott, Design for 
sustainable development – Household drinking water filter for arsenic and pathogen 
treatment in Nepal, J. Environ. Sci. Health A42 (2007), 1879–1888. 
[32] E. Awuah, R.T. Morris, P.A. Owusu, R. Sundell, J. Lindstrom, Evaluation of simple 
methods of arsenic removal from domestic water supplies in rural communities, 
Desalination 248 (2009), 42–47. 
[33] C. Noubactep, A. Schöner, P. Woafo, Metallic iron filters for universal access to safe 
drinking water, Clean 37 (2009), 930–937. 
[34] M. Diao, M. Yao, Use of zero-valent iron nanoparticles in inactivating microbes, Water 
Res. 43 (2009), 5243–5251. 
[35] Y. You, J. Han, P.C. Chiu, Y. Jin, Removal and inactivation of waterborne viruses using 
zerovalent iron. Environ. Sci. Technol. 39 (2005), 9263–9269.
 25
634 
635 
636 
637 
638 
639 
640 
641 
642 
643 
644 
645 
646 
647 
648 
649 
650 
651 
652 
653 
654 
655 
656 
657 
[36] M.I. Litter, M.E. Morgada, J. Bundschuh, Possible treatments for arsenic removal in 
Latin American waters for human consumption, Environ. Pollut. 158 (2010), 1105–
1118. 
[37] P.F. Santina, Method for removing toxic substances in water, US Pat. 5575919, 1996. 
[38] X. Meng, G.P. Korfiatis, Iron powder and sand filtration process for treatment of water 
contaminated with … - US Pat. 6,942,807, 2005. 
[39] C. Noubactep, S. Caré, Enhancing sustainability of household water filters by mixing 
metallic iron with porous materials, Chem. Eng. J. 162 (2010), 635–642. 
[40] C. Noubactep, S. Caré, Dimensioning metallic iron beds for efficient contaminant 
removal, Chem. Eng. J. 163 (2010) 454–460. 
[41] C. Noubactep and A. Schöner, Metallic iron: dawn of a new era of drinking water 
treatment research? Fresen. Environ. Bull. 19 (2010), 1661–1668. 
[42] N.J. Ashbolt, Microbial contamination of drinking water and disease outcomes in 
developing regions, Toxicology 198 (2004), 229–238. 
[43] WHO, Guidelines for drinking-water quality. vol. 1. 3rd ed. Geneva, Switzerland: World 
Health Organization, 2004. 
[44] UNICEF, Unicef handbook on water quality. United Nations Children's Fund, New 
York, 2008. 
[45] J.B. Robertson, S.C. Edberg, Natural protection of spring and well drinking water against 
surface microbial contamination, I. Hydrogeological parameters. Crit. Rev. Microbiol. 
23 (1997), 143–178. 
[46] S.C. Edberg, H. Leclerc, J. Robertson, Natural protection of spring and well drinking 
water against surface microbial contamination. II. Indicators and monitoring parameters 
for parasites. Crit. Rev. Microbiol. 23 (1997), 179–206. 
 26
[47] T. Karanfil, Activated carbon adsorption in drinking water treatment, In Activated 
Carbon Surfaces. In Environmental Remediation T.J. Bandosz (editor). Interf. Sci. 
Technol. 7 (2006), 345–373. 
658 
659 
660 
661 
662 
663 
664 
665 
666 
667 
668 
669 
670 
671 
672 
673 
674 
675 
676 
677 
678 
679 
680 
681 
682 
[48] P. Le Cloirec and C. Faur, Adsorption of organic compounds onto activated carbon – 
applications in water and air treatments, In Environmental Remediation T.J. Bandosz 
(editor). Interf. Sci. Technol. 2006, 7, 375–419. 
[49] W. Zhu, J. Graney, K. Salvage, Land-use impact on water pollution: elevated pollutant 
input and reduced pollutant retention, J. Contemp. Water Res. Ed. 139 (2008), 15–21. 
[50] G.W. Whitman, R.P. Russel, V.J. Altieri, Effect of hydrogen-ion concentration on the 
submerged corrosion of steel. Indust. Eng. Chem. 16 (1924), 665–670. 
[51] C. Noubactep, Processes of contaminant removal in “Fe0–H2O” systems revisited. The 
importance of co-precipitation, Open Environ. J. 1 (2007), 9–13. 
[52] C. Noubactep, A critical review on the mechanism of contaminant removal in Fe0–H2O 
systems, Environ. Technol. 29 (2008), 909–920. 
[53] C. Noubactep, The suitability of metallic iron for environmental remediation, Environ. 
Progr. (2009), doi: 10.1002/ep.10406. 
[54] C. Noubactep, A. Schöner and M. Sauter, Significance of oxide-film in discussing the 
mechanism of contaminant removal by elemental iron materials. In "Photo-
Electrochemistry & Photo-Biology for the Sustainablity"; S. Kaneco, B. Viswanathan, 
H. Katsumata (Eds.), Bentham Science Publishers 2010, 1, 34–55. 
[55] M. Shafiquzzaman, M.S. Azam, I. Mishima, J. Nakajima, Technical and social 
evaluation of arsenic mitigation in rural Bangladesh, J. Health Popul. Nutr. 27 (2009), 
674–683. 
[56] S. Tuladhar, L.S. Smith, SONO filter: An excellent technology for save water in Nepal. 
SOPHEN 7 (2009), 18–24. 
 27
[57] S. Tuladhar, Er.B. Man Shakya, A study on the performance of SONO filter in reducing 
different drinking water quality parameters of ground water: A case study in Ramgram 
municipality of Nawalparasi District, Nepal. Paper presented at the Regional 
Conference on Appropriate Water Supply, Sanitation and Hygiene (WASH) Solution for 
Informal Settelments and Marginalized Communities, Kathmandu, Nepal, May 19-21, 
(2010), 297–310. 
683 
684 
685 
686 
687 
688 
689 
690 
691 
692 
693 
694 
695 
696 
697 
698 
699 
700 
701 
702 
703 
704 
705 
706 
707 
708 
[58] J.L. Jambor, M. Raudsepp, K. Mountjoy, Mineralogy of permeable reactive barriers for 
the attenuation of subsurface contaminants. Can. Miner. 43 (2005), 2117–2140. 
[59] R.W. Gillham, S.F O’Hannesin, Enhanced degradation of halogenated aliphatics by zero-
valent iron, Ground Water 32 (1994), 958–967. 
[60] L.J. Matheson, P.G. Tratnyek, Reductive dehalogenation of chlorinated methanes by iron 
metal. Environ. Sci. Technol. 28 (1994), 2045–2053. 
[61] C.-B. Wang, W.-X. Zhang, Synthesizing nanoscale iron particles for rapid and complete 
dechlorination of TCE and PCBs. Environ. Sci. Technol. 31 (1997), 2154–2156.
[62] W.-x. Zhang, Nanoscale iron particles for environmental remediation: An overview. J. 
Nanopart. Res. 5 (2003), 323–332. 
[63] R. Muftikian, Q. Fernando, N. Korte, A method for the rapid dechlorination of low 
molecular weight chlorinated hydrocarbons in water, Water Res. 29/10 (1995), 2434–
2439. 
[64] N. Korte, L. Liang, R. Muftikian, C. Grittini, Q. Fernando, The dechlorination of 
hydrocarbons: Palladised iron utilised for groud water purification. Platinum Metals 
Rev. 41 (1997), 2–7. 
[65] A. Ghauch, H.A. Assi, A. Tuqan, Investigating the mechanism of clofibric acid removal 
in Fe0/H2O systems, J. Hazard. Mater. 176 (2010), 48–55. 
[66] C. Noubactep, A. Schöner, Fe0-based alloys for environmental remediation: Thinking 
outside the box, J. Hazard. Mater. 165 (2009), 1210–1214. 
 28
709 
710 
711 
712 
713 
714 
715 
716 
717 
718 
719 
720 
721 
722 
723 
724 
725 
726 
727 
728 
729 
730 
731 
732 
733 
[67] M. Stratmann, J. Müller, The mechanism of the oxygen reduction on rust-covered metal 
substrates, Corros. Sci. 36 (1994), 327–359. 
[68] S. Nesic, Key issues related to modelling of internal corrosion of oil and gas pipelines – 
A review, Corros. Sci. 49 (2007), 4308–4338. 
[69] S. Mitsunobu, Y. Takahashi, Y. Terada, M. Sakata, Antimony(V) incorporation into 
synthetic ferrihydrite, goethite, and natural iron oxyhydroxides. Environ. Sci. Technol. 
44 (2010), 3712–3718. 
[70] D. Talbot and J Talbot, Corrosion Science & Technology. CRC Press LLC, Boca Raton, 
New York, Washington, London. 1998, 406 pp. 
[71] E. Zitrou, J. Nikolaou, P.E. Tsakiridis, G.D. Papadimitriou, Atmospheric corrosion of 
steel reinforcing bars produced by various manufacturing processes. Const. Build. 
Mater. 21 (2007), 1161–1169. 
[72] R.E. Dickerson, H.B. Gray and G.P. Haight Jr., Chemical Principles, 3. Edition, 
Benjamin/Cummings Inc. London, Amsterdam, 1979, 944 pp. 
[73] A.F. White, M.L. Peterson, Reduction of aqueous transition metal species on the surfaces 
of Fe(II)-containing oxides. Geochimica et Cosmochimica Acta 60 (1996), 3799–3814. 
[74] J.S. Alvarado, C. Rose, L. LaFreniere, Degradation of carbon tetrachloride in the 
presence of zero-valent iron. J. Environ. Monit. 12 (2010) 1524–1530. 
[75] S. Vigneswaran, J.-S. Chang Effect of media pore size distribution on deep-bed filtration, 
Sep. Technol. 1 (1991), 259–266. 
[76] A. Nur, G. Mavko, J. Dvorkin, D. Galmudi, Critical porosity; a key to relating physical 
properties to porosity in rocks. The Leading Edge 17 (1998), 357–362. 
[77] S. Caré, Q.T. Nguyen, V. L'Hostis, Y. Berthaud, Mechanical properties of the rust layer 
induced by impressed current method in reinforced mortar, Cement Concrete Res. 38 
(2008), 1079–1091. 
 29
734 
735 
736 
737 
738 
739 
740 
741 
742 
743 
744 
745 
746 
747 
748 
749 
750 
751 
752 
753 
754 
755 
756 
757 
758 
759 
[78] C. Noubactep, S. Caré, F. Togue-Kamga, A. Schöner, P. Woafo, Extending service life 
of household water filters by mixing metallic iron with sand. Claen (2010) 
doi:10.1002/clen.201000177. 
[79] U.K. Manandhar, S. Vigneswaran, Effect of media size gradation and varying influent 
concentration in deep-bed filtration: Mathematical models and experiments. Sep. 
Technol. 1 (1991), 178–183. 
[80] A. Santos, P.H.L. Barros, Multiple particle retention mechanisms during filtration in 
porous media. Environ. Sci. Technol. 44 (2010), 2515–2521. 
[81] O.X. Leupin, S.J. Hug, Oxidation and removal of arsenic (III) from aerated groundwater 
by filtration through sand and zero-valent iron, Wat. Res. 39 (2005) 1729–740. 
[82] E. Sikora, D.D. Macdonald, The Passivity of Iron in the Presence of 
Ethylenediaminetetraacetic Acid I. General Electrochemical Behavior, J. Electrochem. 
Soc. 147 (2000), 4087–4092. 
[83] K. Hanna, Sorption of two aromatic acids onto iron oxides: Experimental study and 
modeling, J. Colloid Interf. Sci. 309 (2007), 419–428. 
[84] A.H. Khan, S.B. Rasul, A.K.M. Munir, M. Habibuddowla, M. Alauddin, S.S. Newaz, A. 
Hussam, Appraisal of a simple arsenic removal method for groundwater of Bangladesh. 
J. Environ. Sci. Health A35 (2000), 1021–1041. 
[85] A.K.M. Munir, M. Habibuddowla, M. Alauddin, A. Hussam and A.H. Khan, Evaluation 
of perfomance of Sono 3-Kolshi filter for arsenic removal from groundwater using zero 
valent iron through laboratory and field studies. In: Ahmed, F.M.A., Adeel, A.M.Z. 
(Eds.), Technologies for Arsenic Removal from Drinking Water, vol. 1.1. BUET, UNU, 
Dhaka, 2001, p. 177–189. 
[86] O.X. Leupin, S.J. Hug, A.B.M. Badruzzaman, Arsenic removal from Bangladesh tube 
well water with filter columns containing zerovalent iron filings and sand. Environ. Sci. 
Technol. 39 (2005), 8032–8037. 
 30
760 
761 
762 
763 
764 
765 
766 
767 
768 
769 
770 
771 
772 
773 
774 
775 
776 
777 
778 
779 
780 
781 
782 
783 
784 
785 
[87] S.J. Morrison, D.R. Metzler, B.P. Dwyer, Removal of As, Mn, Mo, Se, U, V and Zn 
from groundwater by zero-valent iron in a passive treatment cell: reaction progress 
modeling. J. Contam. Hydrol. 56 (2002), 99–116. 
[88] S.J. Morrison, P.S. Mushovic, P.L. Niesen, Early Breakthrough of Molybdenum and 
Uranium in a Permeable Reactive Barrier. Environ. Sci. Technol. 40 (2006), 2018–2024.
[89] J. Peña, E. Torres, M.J. Turrero, A. Escribano, P.L. Martín, Kinetic modelling of the 
attenuation of carbon steel canister corrosion due to diffusive transport through 
corrosion product layers. Corros. Sci. 50 (2008) 2197–2204. 
[90] G.A. Waychunas, C.S. Kim, J.F. Banfield, Nanoparticulate iron oxide minerals in soils 
and sediments: unique properties and contaminant scavenging mechanisms. J. Nanopart. 
Res. 7 (2005) 409–433.  
[91] N. Moraci, P.S. Calabrò, Heavy metals removal and hydraulic performance in zero-
valent iron/pumice permeable reactive barriers. J. Environ. Manag. 91 (2010) 2336–
2341. 
[92] C.F. Baes, R.E. Mesmer, The hydrolysis of cations, John Wiley & Sons, New York 
(1976) 489 pp. 
[93] J.M. Combes, A. Manceau, G. Calas, J.Y. Bottero, Formation of ferric oxides from 
aqueous solutions: A polyhedral approach by X-ray absorption spectroscdpy: I. 
Hydrolysis and formation of ferric gels. Geochim. Cosmochim. Acta 53 (1989) 583–
594. 
[94] A. Hussam, Contending with a development disaster: SONO filters remove arsenic from 
well water in Bangladesh. Innovations 4 (2009), 89–102. 
[95] S. Li, S.G.J. Heijman, J.Q.J.C. Verberk, J.C. van Dijk, An innovative treatment concept 
for future drinking water production: fluidized ion exchange – ultrafiltration – 
nanofiltration – granular activated carbon filtration. Drink. Water Eng. Sci. 2 (2009), 
41–47. 
 31
786 
787 
788 
789 
790 
791 
792 
793 
794 
795 
796 
797 
798 
799 
800 
801 
802 
803 
804 
[96] J.J. Rook, Formation of haloforms during chlorination of natural waters. Water 
Treatment Exam. 23 (1974), 234–243. 
[97] J. Rook, Chlorination reactions of fulvic acids in natural waters. Environ. Sci. Technol. 
11 (1977), 478–482. 
[98] M.A. El-Dib, R.K. Ali, THMs formation during chlorination of raw Nile River water, 
Water Res. 29 (1995), 375–378. 
[99] J.W. Li, Z. Yu, X. Cai, M. Gao, F. Chao, Trihalomethanes formation in water treated 
with chlorine dioxide, Water Res. 30 (1996), 2371–2376. 
[100] L.M. Pole, Decline or survival? Iron production in West Africa from the seventeenth to 
the twentieth centuries. J. African Hist. 23 (1982), 503–513. 
[101] B. Prakash, Metallurgy of iron and steel making and blacksmithy in ancient India. 
Indian J. History Sci. 261 (1991), 351–371. 
[102] S.P. Dix, Onsite wastewater treatment: a technological and management revolution. 
Water Eng. Manag. 148 (2001) 24–29. 
[103] B. Tansel, New technologies for water and wastewater treatment: A survey of recent 
patents. Recent Pat. Chem. Eng. 1 (2008) 17–26. 
[104] S. Slaughter, Improving the sustainability of water treatment systems: Opportunities for 
innovation. Solutions 1 (2010) 42–49. 
 32
Figure 1 804 
805  
 806 
807 
808 
 
 33
Figure 2 808 
809  
 810 
811 
812 
 
 34
Figure 3 812 
813  
Fe
0
Fe
2(
O
H
)6
g-
Fe
O
O
H
a-
Fe
O
O
H
Fe
2O
3
Fe
3O
4
3
4
5
6
7
8
 
 density
 BET area
ρ /
 [g
/c
m
3 ]
0
100
200
300
400
(a)
B
ET area / [m
2/g]
 814 
815  
Fe0 Fe2(OH)6 g-FeOOH a-FeOOH Fe2O3 Fe3O4
0
1
2
3
4
5
6
7
(b)
211
5059
1
327
ex
pa
ns
io
n 
/ [
-]
 816 
817 
818 
 
 35
Figure 4 818 
819 
820 
 
 
 821 
822 
823 
824 
 
 
 36
Figure 5 824 
825  
 826 
827 
828 
829 
 
 
 37
Figure captions829 
830 
831 
832 
833 
834 
835 
836 
837 
838 
839 
840 
841 
842 
843 
844 
845 
846 
847 
848 
849 
850 
851 
852 
 
Figure 1: Flow scheme of treatment concept. Potential materials are enumerated without care 
on their relative proportions. In the reactive filtration bed, sand and iron are mixed. The 
volumetric proportion of Fe0 should not exceed 60%. 
 
Figure 2: Comparison of the evolution of contaminant loading in granular activated carbon 
(GAC - up) and Fe0 (dawn) filters. The evolution of the GAC filters is virgin - preloaded 
(reaction front) and saturated carbon. For the Fe0 filters a reaction front may exist due to 
increased O2 in the influent but iron corrosion by H2O (or H+) occurs uniformly in the whole 
column. 
 
Figure 3: Relative variation of density, specific surface area (SSA), and volume of Fe species 
during the process of iron corrosion. The values in (b) represent the SSA from Hanna [83]. 
Strictly any crystallization goes through dissolution, nucleation and aggregation. Intermediate 
species are of high specific area and even more voluminous than Fe2(OH)6. 
 
Figure 4: Schematic diagram of a treatment plant. Media filtration can be performed before 
storage (raw water). Raw water is then filtered in Fe0 filters and the filtrates are collected and 
stored in a drinking water tank for distribution. 
 
Figure 5: Comparison of the processes of groundwater treatment in a conventional treatment 
plant and by Fe0 beds. The number of sand and iron columns is arbitrary and does not reflect 
the actual configuration (parallel or series). Modified after refs. [41,103]. 
 38