Systematic structure control of ammonium iodide salts as feasible UCST-type forward osmosis draw solutes for the treatment of wastewater

Article abstract of DOI:10.1039/c7ta09741g

A variety of UCST-type thermoresponsive ammonium iodide salts is systematically designed and synthesised with desired aqueous solubility, phase-transition temperature, osmolality change, phase-transition concentration range, stability, and toxicity by controlling the ionic interactions, hydrophobicity, and symmetry of the salt structure. Suitable draw solutes can be selected based on the characteristics of the ammonium iodide salts as well as the feed solutions for feasible forward osmosis (FO)-based water purification. In this research, highly concentrated wastewater from a flue gas desulfuriser (FGD), with osmotic pressure three times higher than seawater, is targeted for purification by FO using the draw solutes. Two ammonium iodide salts (HM8I and HM10I) show a remarkable water flux from the wastewater samples and significantly lower salt leakage compared to conventional draw solutes. The osmolality of the phase-separated draw solution drops to less than one tenth of that of the initial feed solution, and reverse osmosis or nanofiltration can be applied to the solution with much lower external pressure for the final purification to produce fresh water. This systematic approach to the design and selection of suitable draw solutes can be an effective strategy for future practical FO-based wastewater treatment and seawater desalination.

Full text of DOI:10.1039/c7ta09741g

View Article Online
View Journal
Journal of
Materials Chemistry A
Accepted Manuscript
This article can be cited before page numbers have been issued, to do this please use: J. Park, H. Joo, M.
Noh, Y. Namkoong, S. Lee, K. H. Jung, H. R. Ahn, S. Kim, J. Lee, J. H. Yoon and Y. Lee, J. Mater. Chem. A,
2017, DOI: 10.1039/C7TA09741G.
This is an Accepted Manuscript, which has been through the
Volume
4 Number 1 7 January 2016 Pages 1–330
Royal Society of Chemistry peer review process and has been
accepted for publication.
Journal of
Materials Chemistry A
Accepted Manuscripts are published online shortly after
Materials for energy and sustainability
www.rsc.org/MaterialsA
acceptance, before technical editing, formatting and proof reading.
Using this free service, authors can make their results available
to the community, in citable form, before we publish the edited
article. We will replace this Accepted Manuscript with the edited
and formatted Advance Article as soon as it is available.
You can find more information about Accepted Manuscripts in the
author guidelines.
Please note that technical editing may introduce minor changes
to the text and/or graphics, which may alter content. The journal’s
standard Terms & Conditions and the ethical guidelines, outlined
in our author and reviewer resource centre, still apply. In no
event shall the Royal Society of Chemistry be held responsible
for any errors or omissions in this Accepted Manuscript or any
consequences arising from the use of any information it contains.
ISSN 2050-7488
PAPER
Kun Chang, Zhaorong Chang et al.
Bubble-template-assisted synthesis of hollow fullerene-like
MoS2 nanocages as a lithium ion battery anode material
rsc.li/materials-a

Please do not adjust margins
Journal of Materials Chemistry A
Page 1 of 14
View Article Online
DOI: 10.1039/C7TA09741G
Journal Name
ARTICLE
Systematic Structure Control of Ammonium Iodide Salts as
Feasible UCST-type Forward Osmosis Draw Solutes for the
Treatment of Highly Concentrated Wastewater
Received 00th January 20xx,
Accepted 00th January 20xx
Jeongseon Park,a Heeyoung Joo,a Minwoo Noh,a Yon Namkoong,a Seonju Lee,a Kyung Hwa Jung,b
Hye Ryun Ahn,c Seulah Kim,a Jong-Chan Lee,b Jae Hoon Yoon,a and Yan Lee*
DOI: 10.1039/x0xx00000x
www.rsc.org/
A variety of UCST-type thermoresponsive ammonium iodide salts is systematically designed and synthesised with desired
aqueous solubility, phase-transition temperature, osmolality change, phase-transition concentration range, stability, and
toxicity by controlling the ionic interactions, hydrophobicity, and symmetry of the salt structure. Suitable draw solutes can
be selected based on the characteristics of the ammonium iodide salts as well as the feed solutions for feasible forward
osmosis (FO)-based water purification. In this research, highly concentrated wastewater from a flue gas desulfuriser (FGD),
with osmotic pressure three times higher than seawater, is targeted for purification by FO using the draw solutes. Two
ammonium iodide salts (HM8I and HM10I) show remarkable water flux from the wastewater samples, and significantly
lower salt leakage compared to conventional draw solutes. The osmolality of the phase-separated draw solution drops to
less than one tenth of that of the initial feed solution, and reverse osmosis or nanofiltration can be applied to the solution
with much lower external pressure for the final purification to produce fresh water. This systematic approach to the design
and selection of suitable draw solutes can be an effective strategy for future practical FO-based wastewater treatment and
seawater desalination.
Introduction
Water shortage has been one of the most serious
environmental issues for sustainable societies due to severe
depletion of freshwater sources as well as increasing water
pollution. Diverse techniques are actively investigated for
future eco-friendly water treatment methods requiring low
energy input.1 Since currently market-dominant reverse
osmosis (RO) method requires large amount of electric energy,
which should be initially transformed from thermal energy
with 30-40% energy conversion efficiency, forward osmosis
(FO) or membrane distillation (MD), which uses mild and
cheap thermal energy sources such as waste heat or residual
heat, are considered to be attractive techniques for future
water treatment, although both technique still remain at the
initial step in the global market yet.2-4 Especially, FO using
Figure 1. Schematic of an FO process from wastewater to fresh water using
ammonium iodide-based UCST draw solutes that are systematically designed and
selected.
This journal is © The Royal Society of Chemistry 20xx
J. Name., 2013, 00, 1-3 | 1
Please do not adjust margins

Please do not adjust margins
Journal of Materials Chemistry A
Page 2 of 14
View Article Online
DOI: 10.1039/C7TA09741G
Journal Name
ARTICLE
osmotic flow between two solutions is usually mentioned as a freshwater production step. However, highly soluble or high-
potential alternative of RO due to lower energy requirement osmotic-pressure thermoresponsive solutes are generally
and reduced membrane fouling in the FO process.5
difficult to phase-separate via a small temperature change,
and considerable amounts remain in the water-rich phase
after phase separation.12 Moreover, the salinity and
temperature of the feed solutions vary according to local
circumstances, for example, from 3.5% and 20°C for surface
seawater in middle latitudes to 20% and 50°C for wastewater
from flue gas desulfurisers (FGD).14 Suitable thermoresponsive
draw solutes are even more difficult to select due to the
discrepancy between their drawing activity and separability as
well as the heterogeneous properties of feed solutions.
In general, an FO system comprises a semipermeable
membrane, a feed solution, and a draw solution. The detailed
principle is that the draw solution pulls out water from the
feed solution using the osmotic pressure difference between
two solutions in the first step, and the drawn water is
separated from the draw solutes with energy input in the
second step.
The draw solute plays a critical role in the FO system, and
various prototypes have been suggested. High osmotic
pressure for effective drawing, easy separation with minimal
energy input, negligible solute leakage through the membrane,
and efficient recovery and reusability are all essential
properties for an ideal draw solute. Several draw solutes,
including ammonium bicarbonate (NH4HCO3),7 ionic liquids,8
hydrogels,9 and nanoparticles,10 meet some of the criteria.11
However, most hydrogels and nanoparticles generate only
limited osmotic pressure, and these are used for desalination
of water with low salt concentrations.12 With NH4HCO3, the
current FO draw solute, difficult gas/liquid separation after the
drawing process, and relatively high solute leakage hinder
energy-efficient recovery and reuse of the draw solute.7
There is little similarity in the structures of previously
reported thermoresponsive draw solutes, so that the essential
characteristics for the selection of draw solutes, including
osmotic pressure, Tp range, and residual amounts after phase
separation, are difficult to estimate from the structures. In this
research, we designed novel low-molar-mass UCST materials
whose physicochemical characteristics could be systematically
controlled by modulating the symmetry, hydrophobicity, and
ionic interactions in the molecular structure based on
ammonium iodide salts (Figure 1). We can observe significant
variation in the maximum osmolality from 700 mOsm kg-1 to 9,000
mOsm kg-1, Tp from 2°C to 67°C, and the residual osmolality in
the water-rich phase from 100 mOsm kg-1 to 3,000 mOsm kg-1
by tuning the structure of ammonium iodides (Table 1). We
believe that this approach is important not only for the
structure-based design of thermoresponsive materials with
tailored physicochemical characteristics but also for the
selection of an appropriate draw solute that is optimised for
each FO system and conditions. As a proof of our concept, we
suggested an FO system based on selected ammonium iodides
for the treatment of FGD wastewater (Figure 1). Generally,
wastewater has an osmotic pressure that is about three times
higher than that of seawater, so a significant amount of energy
is inevitably required to purify it by conventional water
treatment methods.15 After active drawing from highly
concentrated wastewater at a temperature above Tp, raise the
osmotic pressure of the drawing solution above that of the
wastewater, phase separation is induced by lowering the
temperature below Tp, and fresh water is obtained by RO
under a much reduced external pressure due to the minimal
residual osmotic pressure in the water-rich phase.
Thermoresponsive materials in an aqueous environment
were proposed as an alternative FO draw solute to overcome
such drawbacks.12 Lower critical solution temperature (LCST)
and upper critical solution temperature (UCST) materials show
sudden solubility changes at certain phase-transition
temperatures (Tp).13 LCST materials exist as one phase with
solvents below Tp but are phase-separated above Tp, whereas
UCST materials are miscible with solvents above Tp but
immiscible below Tp. In several previous studies,
thermoresponsive materials were applied as draw solutes for
FO.9,10,12,14 When thermoresponsive materials are dissolved in
water to prepare a draw solution at a high concentration, the
draw solution can draw water from a feed solution. After the
drawing process, the draw solution is phase-separated into a
water-rich phase and a solute-rich phase by changing the
temperature. The water-rich phase is further purified to
generate fresh water with minimal energy input, and the
solute-rich phase is recycled for the next drawing process. The
simple temperature-driven separation and easy recovery and
reuse of the draw solute are advantages of using
thermoresponsive materials as potential draw solutes.12
Results and discussion
Systematic design and synthesis of draw solutes
In the first drawing step, thermoresponsive materials that
can generate high osmotic pressure under one-phase
conditions are useful for drawing water with a high flux from
highly concentrated saline or polluted water. In the separation
step, thermoresponsive materials should be phase-separated
within a tolerable temperature range, and minimal solute
materials should be retained in the water-rich phase to
achieve low osmotic pressure for the final low-energy
The aim was to systematically design thermoresponsive draw
solutes with desired osmotic pressures, Tp ranges, and other
phase separation properties. In aqueous solutions, LCST phase
transition is highly dependent upon entropy factors, generally
the hydrophobicity of solute molecules.16 Although some LCST
materials with phase transitions in the range 20°C–60°C have
This journal is © The Royal Society of Chemistry 20xx
J. Name., 2013, 00, 1-3 | 2
Please do not adjust margins

Please do not adjust margins
Journal of Materials Chemistry A
Page 3 of 14
View Article Online
DOI: 10.1039/C7TA09741G
Journal Name
ARTICLE
been proposed as FO draw solutes, the solubility or osmotic
pressure was limited, and they drew limited amounts of seawater
This journal is © The Royal Society of Chemistry 20xx
J. Name., 2013, 00, 1-3 | 3
Please do not adjust margins

Journal of Materials Chemistry A
Page 4 of 14
Please do not adjust margins
Journal Name
ARTICLE
UCST
Formula
UCST
temperature
changeb)
Maximum
osmolality
at 60°Cc)
Alkyl
group
Name
(Abbreviation)
Residual osmolality
concentration
weight
after phase separationc)
rangea)
R1=CH3,
R2=CH3
Tetramethylammonium iodide
(4MAI)
201.05
243.13
271.19
299.24
400.09
408.24
456.19
484.25
512.30
484.25
3ꢀ15
60ꢀ80
≥ 80
2ꢀ60
46ꢀ61
≥ 27
700
8100d)
5700
110
R1=CH3,
R2=C4H9
N,N,N,NꢀTrimethylbuthylammonium iodide
1700
400
(3MBAI)
R1=CH3,
N,N,N,NꢀTrimethyloctylammonium iodide
R2=C8H17
(3MOAI)
R1=C3H7
R2=C2H7
N,N,N,NꢀTripropylethylammonium iodide
60ꢀ80
60ꢀ70
60ꢀ70
40ꢀ70
20ꢀ70
20ꢀ70
20ꢀ40
17ꢀ67
4ꢀ64
3700d),e)
8800d)
8800
3000
2200
2000
800
(3PEAI)
N,N,N,N’,N’,N’ꢀhexamethylethanediammonium iodide
n=2,
R3=CH3
(HM2I)
N,N,N,N’,N’,N’ꢀhexamethylbutanediammonium iodide
n=4,
7ꢀ28
R3=CH3
(HM4I)
N,N,N,N’,N’,N’ꢀhexamethylhexanediammonium iodide
n=6,
13ꢀ62
8ꢀ50
5900f)
6200
R3=CH3
(HM6I)
N,N,N,N’,N’,N’ꢀhexamethyloctanediammonium iodide
n=8,
310
R3=CH3
(HM8I)
N,N,N,N’,N’,N’ꢀhexamethyldecanediammonium iodide
n=10,
26ꢀ51
25ꢀ58
5400d)
1200
180
R3=CH3
(HM10I)
N,N,N,N’,N’,N’ꢀhexaethylethanediammonium iodide
n=2,
140
R3=C2H5
(HE2I)
a) The unit for UCST concentration range: wt%
b) The unit for temperature change: °C
c) The units for maximum osmolality and residual osmolality after phase separation: mOsm kg-1
d) The osmolality was measured at 65°C due to the phase separation at 60°C
e) The osmolality of 70 wt% solution was measured.
.
f) The osmolality of 60 wt% solution was measured.
Table 1. Structures and characteristics of ammonium iodide-based UCST draw solutes.
This journal is © The Royal Society of Chemistry 20xx
J. Name., 2013, 00, 1-3 | 4
Please do not adjust margins

Please do not adjust margins
Journal of Materials Chemistry A
Page 5 of 14
View Article Online
DOI: 10.1039/C7TA09741G
Journal Name
ARTICLE
or diluted saline mainly due to their intrinsic structural abbreviated as 3MOAI. Similarly, the names of the
hydrophobicity.12 On the other hand, since UCST phase diammonium
transition is more dependent upon enthalpy factors, many
ionic structures with higher aqueous solubility have been
reported.17 We decided to design UCST draw solutes with ionic
structures for effective drawing from concentrated
wastewater with high osmotic pressure. In our previous
research, halide salts of branched polyethylenimine (
alkylated -PEI showed UCST phase transitions in an aqueous
system.17,18 As the alkylammonium unit of alkylated
-PEI can be
readily synthesised by one-step reaction, and the
b-PEI) and
b
b
a
hydrophobicity and ionic interactions can be easily modulated
by changing the alkyl groups, we selected the alkyl ammonium
halide as the basic structure for the UCST draw solute.
Figure 2. UCST phase transition of aqueous solutions of a) HM8I at
concentration of 20 wt% (●, solid line), 40 wt% (○, long dashed line), 60 wt% (
short dashed line), and 70 wt% (
of 20 wt% (●, solid line), 40 wt% (○, long dashed line), 60 wt% (
line), and 70 wt% , dotted line).
a
,
The UCST phase transition of a mixture can be predicted by
▼
△
, dotted line) and b) HM10I at a concentration
, short dashed
Gibbs free energy equation (∆Gmix = ∆Hmix - T∆Smix) for a certain
composition, although this is a simplified prediction.19 When a
UCST material dissolves in water, the enthalpy of mixing
(∆Hmix) is generally positive because solute-solute interactions
are stronger than solute-solvent interactions. The sum of the
combinatorial entropy and interaction entropy determines the
entropy of mixing (∆Smix), which is also generally positive. At a
lower temperature, the positive enthalpy term is greater than
▼
(
△
iodide salts are abbreviated by representing the number of
methylene groups between two ammonium ions:
N,N,N,N’,N’,N’-hexamethylhexanediammonium iodide (HM6I).
Ammonium iodide salts were prepared by alkylation of the
amine precursors except for commercially available 4MAI and
HM10I. The synthetic scheme is shown in Figure S1 and Table
S1, ESI†, and the detailed procedure is described in the
Experimental Section of the ESI†. The alkylating agent was 1-
iodomethane, and bases were occasionally added as proton
quenchers during the alkylation. All the salts could be
synthesised by a one-step reaction, which is an important
factor for future low-cost synthesis of draw solutes. The
molecular structures were confirmed by NMR, which is shown
in Figure S1, ESI†. Successful and quantitative quaternarisation
of amines is clearly shown as the peak shift in the NMR spectra.
the entropy term (-
T
∆
Smix), and ∆Gmix becomes positive to
Smix
induce phase separation. As the temperature increases, -
T
∆
is greater than ∆Hmix, resulting in a negative ∆Gmix, so that the
mixture becomes miscible. The UCST phase transition occurs at
Tp = ∆Hmix/∆Smix, where the Gibbs free energy becomes zero.
Thus, the UCST phase transition properties are expected to be
controlled by systematic structural changes in the molecular
size, ionic interactions, hydration, and the hydrophobicity of
solutes.
As the first step toward systematic structure control, we
considered the molecular sizes. Since osmotic pressure is
inversely related to molar mass, low-molar-mass molecules
with one or two ammonium moieties would be preferred to UCST phase-transition behaviour of ammonium iodide salts
achieve high osmotic pressure. However, low-molar-mass
molecules generally produce a high positive combinatorial We investigated the effect of the structure of ammonium
∆
Smix to decrease Tp far below the tolerable temperature iodide salts on the UCST phase transition behaviour. Amongst
range. To elevate Tp, solute-solute interactions might be various techniques for the measurement of the UCST phase
strengthened to produce a more positive ∆Hmix, or the transition, we used the simplest technique, turbidimetry with a
hydrophobicity might be increased to achieve a less positive UV-Vis spectrophotometer. An abrupt change in transmittance
∆
Smix. Amongst the halide anions, iodide has the lowest occurred when the solute was phase-separated from water.
negative free energy of hydration as it has the strongest The UCST phase-transition temperature was defined as that
interaction with alkylated ammoniums.20 Therefore, we for which the transmittance is 50%. The important UCST
selected iodide as the counter anion for the ammonium cation phase-transition properties of ammonium iodide salts are
in the designed structure. Moreover, fine-tuning of summarised in Table 1. Aqueous solutions of all the draw
hydrophobicity and symmetry of the backbone could be solutes exhibited UCST phase transitions within certain
achieved by changing the alkyl chains around and between the concentration ranges. Below the minimum concentration, the
ammonium ions. According to the rationale above, we draw solutes were miscible with water, whereas they were
designed four monoammonium iodide salts and six immiscible with water regardless of the temperature change
diammonium iodide salts as potential UCST draw solutes above the maximum concentration.
(Table 1). The names of the monoammonium iodide salts are
Draw solute 4MAI revealed relatively poor aqueous
solubility (15 wt%) compared to other salts although it
abbreviated by indicating the number of alkyl groups. For
example, N,N,N,N-trimethyloctylammonium iodide is
This journal is © The Royal Society of Chemistry 20xx
J. Name., 2013, 00, 1-3 | 5
Please do not adjust margins

Please do not adjust margins
Journal of Materials Chemistry A
Page 6 of 14
View Article Online
DOI: 10.1039/C7TA09741G
Journal Name
ARTICLE
possesses the least hydrophobic methyl groups on the with the concentration. At the same concentration, HM10I
ammonium. It was supposed that the poor solubility was due showed a higher UCST temperature than that of HM8I.
to the symmetric structure of the tetramethyl ammonium ion, Hydrophobicity generally has a negative influence on ∆Smix
which could facilitate regular stacking of ions to elevate the because of the ordering of water molecules around the solute
lattice energy.21 Poor aqueous solubility is clearly a weak point molecule.22 Therefore, the longer hydrophobic chain may
in a draw solute requiring high osmotic pressure. Moreover, elevate Tp (=∆Hmix/∆Smix) by decreasing ∆Smix, assuming that all
slight dilution from 15 wt% to 3 wt% resulted in an abrupt the other thermodynamic factors are not significantly
change in Tp from 60°C to 2°C. This means that the diluted changed. In addition, HM8I and HM10I exhibited UCST phase
draw solution, after drawing a small amount of water at 60°C, transitions in the temperature ranges 8°C–50°C and 26°C–
could be phase-separable only below 2°C, requiring high 51°C, respectively. Clearly, the longer hydrophobic chain
energy costs for cooling. Therefore, asymmetric ammonium narrowed the Tp range. As mentioned above, a minimal change
structures are preferred for practical draw solutes, with high in Tp is essential to reduce the energy cost for the separation
aqueous solubility and a wide concentration range as well as a and recovery of the draw solute. For example, both HM8I and
minimal change in Tp
.
HM10I form one phase with water at 70 wt% and at 55°C,
showing high osmotic pressure for drawing water. When both
solutions are diluted to 40 wt%, the diluted HM8I and HM10I
solutions can be phase-separated at approximately 15°C and
30°C, respectively. Therefore, cooling by more than 40°C
would be required for the separation of HM8I, but cooling by
only 25°C would be sufficient for HM10I.
The introduction of asymmetric ammonium to the structure
clearly enhanced the aqueous solubility. Solute 3MBAI with
one butyl and three methyl groups showed a maximum
solubility of 80 wt% and a UCST phase transition at 60–80 wt%.
The corresponding Tp values were 46°C and 61°C at 60 wt%
and 80 wt%, respectively. Solute 3MOAI with a longer octyl
group showed even higher aqueous solubility of more than 80
wt%, although it has a more hydrophobic structure. At this
concentration, 3MOAI exhibited a UCST transition at 27°C. On
the other hand, 3PEAI, with bulkier alkyl groups and a more
symmetric structure, was less soluble in water than 3MOAI
and showed a large change in Tp (17°C–67°C) upon dilution.
Overall, the maximum aqueous solubility is highly dependent
upon the symmetry of the ammonium iodide salts, and the
concentration and Tp ranges for which UCST phase transitions
occur can be controlled by changing the hydrocarbon chain
between the two ammonium ions.
Change in osmolality through the UCST phase transition
Although we found that an asymmetric structure was The abrupt change in the solubility of thermoresponsive
important in obtaining the desired UCST phase transition in a materials at Tp induces a dramatic change in osmotic pressure.
draw solute, the concentration range of monoammonium iodide At higher temperatures above Tp, draw solutions should have
salts was too narrow to be applied to a practical FO system. osmotic pressures as high as possible for effective drawing
Therefore, we investigated diammonium iodide salts for more from feed solutions. On the other hand, phase-separated draw
precise control of the UCST phase-transition behaviour, solutions at temperatures below Tp should have osmotic
because they possess an alkyl chain between two ammonium pressures as low as possible to reduce the energy cost to
ions and exhibit different hydrophobicity according to its produce fresh water. A large difference in the osmotic
length. All diammonium iodide salts with six methyl groups pressure through the UCST phase transition is preferable for
(HM series) had a maximum solubility above 70 wt%. There is a an FO system based on thermoresponsive draw solutes. Since
tendency for diammonium iodide salts with longer alkyl chains osmotic pressure is approximately proportional to
to show the UCST phase transitions over a wider range of osmolality,23 we measured the osmolality before and after the
concentrations. HM2I and HM4I showed UCST phase UCST phase transition were measured by osmometry. A
transitions in the range 60–70 wt%, and HM6I in the range 40– frequently used method for the measurement of osmolality is
70 wt%. HM8I and HM10I exhibited a UCST phase transition in
the range 20–70 wt%. However, HE2I with six ethyl groups
showed a maximum solubility of 40 wt%, which was much
lower than that of HM8I, the structural isomer of HE2I. In
addition, HE2I exhibited a UCST phase transition in a narrower
concentration range of 20–40 wt%. The lower solubility and
the narrower UCST concentration range were presumably due
to the more symmetric ammonium structure of HE2I with all
ethyl or ethylene groups, which is similar to that of 4MAI.
f
r
e
e
z
i
n
g
p
o
i
n
t
Of all the ammonium iodide salts in this study, HM8I and
HM10I showed the widest concentration ranges for UCST
phase transitions (20–70 wt%). The concentration-UCST
relationships for HM8I and HM10I are examined in more detail
in Figure 2a and b, respectively. It is clear that Tp increased
Figure 3. a) Concentration-osmolality relationship of HM8I solutions (○) and
HM10I solutions (
▲
) at 60°C with the exception of 70% HM10I(*) at 65°C. b)
This journal is © The Royal Society of Chemistry 20xx
J. Name., 2013, 00, 1-3 | 6
Please do not adjust margins

Please do not adjust margins
Journal of Materials Chemistry A
Page 7 of 14
View Article Online
DOI: 10.1039/C7TA09741G
Journal Name
ARTICLE
Comparison of the osmolality at the maximum solubility of HM8I and HM10I
before the phase separation (grey) and the osmolality of the water-rich layer
after phase separation (black). The osmolality of HM8I and HM10I were
measured at 60°C and 65°C, respectively. The osmolality of average seawater
(dotted line) and a sample of FGD wastewater (TDS = 258,000 ppm) (solid line)
are indicated as references in the graph.
HM10I showed the largest (30 fold) drop in the osmolality to
180 mOsm kg-1. Again, the longer chain contained in HM10I
induced the larger drop in the osmolality.
Selection of suitable draw solutes for FO
Since the salinity and the temperature of feed solutions are
heterogeneous, as mentioned previously, a suitable draw
solute should be selected according to local conditions. For
seawater with an osmolality of 1,200 mOsm kg-1, most of the
ammonium iodide salts except for 4MAI and HE2I could be
selected as draw solutes considering the maximum osmolality
only. Consideration of the UCST concentration range gives
priority to HM8I and HM10I. Both HM8I and HM10I exhibited
sufficient osmolality well above that of seawater at
concentrations 30–40 wt% (Figure 3). For a higher osmotic
flux, HM8I with a higher osmolality would be preferable. On
the other hand, a minimal change in Tp upon dilution (Figure 2)
favours HM10I. Furthermore, the residual osmolality of HM10I
was 180 mOsm kg-1, which is lower than the 310 mOsm kg-1 of
HM8I. Fresh water could be obtained by post-treatment such
as nanofiltration (NF) or RO with external pressures of over 3.8
atm (HM10I) or 6.5 atm (HM8I) at 25°C, which are far lower
than the requirement for seawater desalination (~25 atm). To
achieve lower energy costs during the recovery of draw solutes
and the production of fresh water, HM10I would be a better
choice.
osmometry (FPO). However, it is difficult to measure
osmolality of concentrated UCST draw solutions because the
draw solutes are phase-separated from water below Tp
.
Therefore, we instead used vapour pressure osmometry (VPO)
for such measurement. The osmolality at the maximum
solubility of each draw solution is represented as the
maximum osmolality in Table 1. On the other hand, we could
use FPO for the measurement of the osmolality of the water-
rich layer of a phase-separated draw solution below Tp. The
osmolality of the water-rich layer is also represented as the
residual osmolality after phase separation in Table 1.
As expected from the low aqueous solubility of 4MAI and
HE2I, their maximum osmolality were quite low at 700 and
1,200 mOsm kg-1, respectively. Considering that the osmolality
of seawater is approximately 1,200 mOsm kg-1, 4MAI and HE2I
could not induce osmotic flux from seawater due to their low
osmolality. Two monoammonium iodide salts with high
aqueous solubility, 3MBAI and 3MOAI, showed maximum
osmolality of 8,100 and 5,700 mOsm kg-1, respectively, at a
concentration of 80 wt%. After phase separation, the residual
osmolality of the water-rich layer dropped to 1,700 and 400
mOsm kg-1. Solute 3MOAI with a longer hydrophobic chain
showed a 14 fold larger drop in osmolality through the UCST
phase transition compared to 3MBAI with a 4.7 fold drop.
Interestingly, 3PEAI with hydrocarbon chains shorter than
propyl groups showed almost no drop in osmolality (3,700
mOsm kg-1 vs. 3,000 mOsm kg-1). A low residual osmolality
indicates that the water-rich layer contains a small amount of
solute due to effective phase separation. Since hydrophobic
chains tend to form aggregates to phase-separate from water
due to hydrophobic interactions, draw solutes with longer
hydrocarbon chains can be more effectively phase-separated into
a water-rich layer with a lower residual solute content and a
solute-rich layer with a higher solute content below Tp. In the
same manner, diammonium iodide salts with a longer chain
showed a larger drop in osmolality though the UCST phase
transition: HM2I, HM4I, and HM6I showed 4 fold (from 8,800
to 2,200 mOsm kg-1), 4.5 fold (from 8,800 to 2,000 mOsm kg-1),
and 7.4 fold (from 5,900 to 800 mOsm kg-1) drops in
osmolality.
In this research, we intended to develop draw solutes for FO
purification of highly concentrated wastewater. The osmolality
of a wastewater sample from an FGD system with total
dissolved solids (TDS) of 258,000 ppm was 3,500 mOsm kg-1 at
60°C. HM8I and HM10I could be selected through the
rationalization above. The final candidate for the FO system
may be selected by comparing the cost of cooling and the gain
from the fast osmotic flux. In the following FO experiments, we
compared the drawing flux and solute leakage of HM8I and
HM10I.
FO water flux of HM8I and HM10I solutions
The water flux of HM8I and HM10I was measured using two
systems. A pair of handmade glass tubes was used for small-
scale analysis (dead-end type) (Figure S2a, ESI†), whereas a
cross-flow instrument consisting of reservoirs, pumps, and an
osmosis cell was utilised for large-scale analysis (Figure S2b,
ESI†). A commercially available thin film composite (TFC)
membrane was used, and the active layer faced a feed solution
(FO mode). NaCl solutions with various concentrations were
used as feed solutions. In addition, as actual feed solutions in
the field, two samples of wastewater were taken from an FGD
factory of Doosan Heavy Industries and Construction (Korea).
The TDS in each wastewater sample were measured as
164,000 and 258,000 ppm. The water flux was expressed as
the volume of of drawn water (L) per unit area of
The osmolality of HM8I and HM10I, which showed the
widest concentration ranges for the UCST phase transition, are
shown in Figure 3a. The osmolality increased non-linearly with
the concentration of the draw solution. The maximum
osmolality of HM8I and HM10I were 6,200 and 5,400 mOsm
kg-1, respectively. After phase separation, both draw solutes
showed a dramatic change in osmolality, which is clearly
shown in Figure 3b. The residual osmolality of HM8I was 310
mOsm kg-1, which was 1/21 of the maximum osmolality.
semipermeable membrane (m2) per unit time (
h) (LMH).
This journal is © The Royal Society of Chemistry 20xx
J. Name., 2013, 00, 1-3 | 7
Please do not adjust margins

Please do not adjust margins
Journal of Materials Chemistry A
Page 8 of 14
View Article Online
DOI: 10.1039/C7TA09741G
Journal Name
ARTICLE
The FO water flux from NaCl solutions or wastewater NaCl, which is the equivalent of seawater, the water flux was
samples to a 60 wt% HM8I draw solution was measured with 4.6 L m-2 -1(LMH). This indicates that effective FO desalination
h
the dead-end type apparatus at 50°C, where the applied of seawater is feasible using the HM8I draw solution.
temperature was 10°C higher than Tp for a stable and effective Furthermore, the HM8I
FO system (Figure 4a). When the feed solution was 0.60 m
Figure 4. Measurement of FO drawing flux using HM8I and HM10I draw solutions. FO drawing water fluxes using a) 60 wt% HM8I at 50°C and b) 60 wt% HM10I at
55°C measured with glass U-shaped tubes (dead-end type). Aqueous NaCl solutions (0.30 m, 0.60 m, 1.80 m, and 2.40 m) and FGD wastewater samples (wastewater
1: TDS = 164,000 ppm, wastewater 2: TDS = 258,000 ppm) were used as feed solutions. c) The FO water flux between the 60 wt% HM8I draw solution and a 0.60 m
NaCl solution (▲, solid line) or wastewater 1 (●, doꢀed line) at 50°C measured with the cross-flow FO instrument.
draw solution could draw water from high salt saline solutions,
In most practical water treatment systems using
1.80 m and 2.40 m NaCl. Using wastewater samples as feed semipermeable membranes, processes based on a cross-flow
solutions, the water flux were observed as 0.90 and 0.70 LMH, system are dominant for a higher water flux compared to the
respectively. The results suggest that the HM8I draw solution dead-end type system.24 We measured the drawing flux using
can draw water from highly concentrated wastewater as well a large-scale cross-flow system with an HM8I draw solution. A
as seawater.
60 wt% HM8I solution was placed in the draw solution (DS)
reservoir and 0.6 m NaCl or the wastewater sample (TDS =
164,000 ppm) was placed in the feed solution (FS) reservoir
(Figure S2b, ESI†). The volumetric change along with the
movement of water was observed at 50°C. The water flux
measured with the cross-flow system was 3–4-fold higher than
the values measured with the dead-end type system (Figure
4c). The rapid regeneration of the osmotic gradient across the
membrane by the cross flow might be the main reason for the
significant enhancement of the water flux. The water flux from
the 0.60 m NaCl was observed to be around 15 LMH after
fluctuation during the initial 30 min. With the wastewater as
the FS, the osmotic water flux was maintained at 3.3 LMH
without much fluctuation. Both results indicate that effective
FO can be operated with the 60 wt% HM8I DS for the
The FO water flux using a 60 wt% HM10I draw solution was
measured in a similar way at 55°C (Figure 4b). A water flux of
3.0 LMH was observed by using 0.60 m NaCl. The water flux
using HM10I was slightly lower than when using HM8I against
the same feed solution. The main reason for the lower flux is
probably that the osmolality of HM10I was lower than that of
HM8I at 60 wt% (Figure 3a). Nevertheless, the HM10I draw
solution could draw water from both wastewater samples
(0.90 LMH and 0.40 LMH). The osmolality of HM10I at 60 wt%
was almost equivalent to that of the wastewater sample with
TDS of 258,000 ppm. We supposed that the drawing from the
wastewater sample might be close to the limit of the HM10I
drawing activity.
This journal is © The Royal Society of Chemistry 20xx
J. Name., 2013, 00, 1-3 | 8
Please do not adjust margins

Please do not adjust margins
Journal of Materials Chemistry A
Page 9 of 14
View Article Online
DOI: 10.1039/C7TA09741G
Journal Name
ARTICLE
treatment of highly concentrated wastewater as well as Post-FO water treatment
seawater desalination.
After the FO drawing and phase separation, a water-rich layer
with a small amount of draw solutes is obtained. The water-
rich layer could be post-treated to produce fresh water by NF
or RO with external pressures that are much lower than the
initial osmotic pressures of FSs. As a test for the potential
production of fresh water, we treated the water-rich layer of
the phase-separated HM8I solution by secondary RO with a
dead-end setup and an RO membrane (ACM4, TriSep). The
rejection efficiency, which indicates the percentage of rejected
solutes during RO,27 was calculated from the osmotic pressure
of the produced water. The water flux and the rejection
efficiency of
Reverse solute flux of draw solutes
Possible leakage of draw solutes could not only cause a
gradual decline in drawing activity but also raise
environmental problems. The solute flux into the FS during FO
was expressed as the reverse solute flux, the amount of
permeated draw solutes into the FS (mole) per unit area of
semipermeable membrane (m2) per unit time (
h) (MMH). We
analysed the reverse solute flux of HM8I, HM10I, and other
c
o
n
t
r
o
l
d
r
a
w
Feed solution
RO fluxa)
Rejectionb)
NaCl
NH4HCO3
HM8I
1.54
1.69
0.82
99.0±0.2
98.9±0.2
99.6±0.0
a) RO flux unit: LMH bar-1
b) Rejection unit: %
Figure 5. Comparison of the reverse solute flux from NaCl, NH4HCO3, HM8I, and HM10I
solutions at 2.0 m through CTA membrane (black) and TFC membrane (grey) into a feed
solution (deionised water). The flux was measured at 50°C except for NH4HCO3 (at
40°C).
Table 2. Water flux and rejection (%) data for an RO process under 15.5 bar at
RT. The rejection percent was calculated by comparing the osmolalities of the
initial feed solution (0.125 m) and the product water.
the RO process using NaCl, NH4HCO3, and HM8I FSs at the
same concentration of 0.125 m are summarised in Table 2,
ESI†. HM8I showed the highest rejection of 99.6% compared to
NaCl with 99.0% and NH4HCO3 with 98.9%. The higher
rejection of HM8I over the other draw solutes was due to its
larger size and the divalency of the diammonium cation, as for
the lower reverse solute flux in FO (Figure 5). The loss of draw
solutes was negligible in the secondary RO process. Main loss
arose from the leakage in the first FO process. In our dead-end
type FO system using HM8I, the solute leakage rate was
approximately 0.5%/hr through the TFC membrane, and
0.06%/hr through the CTA membrane. In order to maximise
the recovery efficiency of the draw solutes, a careful choice of
a membrane is necessary.
solutes (NaCl and NH4HCO3) through representative FO
membranes, cellulose triacetate (CTA), and TFC membranes,
by measuring changes in the conductivity of the FS (Figure 5).
The FO system was operated at 50°C, except for NH4HCO3 at
40°C, because NH4HCO3 tends to decompose into ammonia
and carbon dioxide at temperatures near 60°C. NH4HCO3, the
standard FO draw solute, showed the highest reverse solute
flux for both membranes: 0.61 MMH for the CTA membrane
and 0.55 MMH for the TFC membrane. Our potential draw
solutes, HM8I and HM10I, exhibited significantly lower reverse
solute flux values than the control draw solutes: HM8I with 0.027
MMH (CTA), and 0.19 MMH (TFC), HM10I with 0.025 MMH
(CTA) and 0.15 MMH (TFC). The lower leakage of HM8I and
HM10I could be rationalised by the fact that the hydrated sizes
-
of ions are in the order: diammonium ions >> Na+ ≈ HCO3 > Cl-
This result supported the potential of wastewater
purification by an FO-RO combined system using our
thermoresponsive ammonium iodides. Without the primary FO
process, the RO-based purification was almost impossible due
to requirement of exceeding external pressure (> 150 atm).
Most RO membranes cannot resist the high pressure, and the
energy consumption also elevates rapidly.1,7 However, using the FO
process with HM8I, the operating pressure can be effectively
reduced to about 15 atm, representing a 10 fold decrease from
the initial FS. The energy consumption for the secondary RO
process can be estimated to approximately 0.8 kWh m-3.28
Considering the NH4HCO3-based FO generally consumes 0.8
+ 21
≈ I- > NH4 . HM8I and HM10I are composed of one large
cation and two anions, which may have a small probability of
permeating through the membrane compared to other draw
solutes composed of two small ion pairs. As shown in Figure 5,
there seems to be a big difference of solute leakages between
membranes. Apart from NH4HCO3, higher solute leakage was
observed when the TFC membrane was applied. This
phenomenon was frequently observed in FO processes using
TFC membranes with higher permeation performances for
both water and ions.25 The high leakage of NH4HCO3 through
the CTA membrane indicated possible damage to the
vulnerable CTA membrane by the draw solutes.26
This journal is © The Royal Society of Chemistry 20xx
J. Name., 2013, 00, 1-3 | 9
Please do not adjust margins

Please do not adjust margins
Journal of Materials Chemistry A
Page 10 of 14
View Article Online
DOI: 10.1039/C7TA09741G
Journal Name
ARTICLE
kWh m-3 for the seawater desalination,6,7 we believe that our depending on the properties of FSs might have significant
FO-RO combined system can have competitiveness in the potential for the development of feasible FO draw solutes for
purification of highly-concentrated wastewater.
various field conditions.
Stability and toxicity of draw solutes
Experimental
Potential thermal decomposition of a draw solute during
repetitive heating-cooling cycles could be a serious problem
with respect to the durability of the FO process. HM10I DS was
conditioned by sealing under 60 repetitive heating (60°C)-
Materials
Tetramethylammonium iodide (4MAI), tripropylamine,
N,N,N’,N’-tetramethylethylenediamine, N,N,N’,N’-tetramethyl-
1,4-diaminobutane, N,N,N’,N’-tetramethyl-1,6-diaminohexane,
cooling (10°C) cycles for 120
h and then heated at 80°C for a
further 60 days. Neither degradation of the structure nor
decomposition was observed after the harsh thermal
decamethonium
diisopropylethylamine (DIPEA) were purchased from TCI
(Japan); -butylamine, -octylamine, 1,8-diaminooctane,
iodide
(HM10I),
and
N,N-
1
treatment, and this was confirmed by H NMR and LC/MS
(Figure S3a and b, ESI†). In addition, it was found that the
phase-transition behaviour was almost unchanged after the
treatment (Figure S3c, ESI†). Overall, the ammonium structure
of the HM10I draw solute was highly stable under aqueous
conditions under the temperature swing and prolonged
exposure at high temperatures.
n
n
ethylenediamine, 1-iodomethane, 1-iodoethane, dimethyl
sulfoxide (DMSO), and 3-(4,5-dimethylthiazol-2-yl)-2,5-
diphenyl tetrazolium bromide (MTT) were purchased from
Sigma-Aldrich (USA). Acetonitrile, methanol, ethanol, diethyl
ether, dichloromethane and potassium bicarbonate were
purchased from Daejung (Korea). Dulbecco’s Modified Eagle’s
Medium (DMEM), Dulbecco’s phosphate buffered saline
(DPBS), and fetal bovine serum (FBS) were purchased from
WelGENE (USA). All reagents were used without further
purification. The wastewater samples (total dissolved solids;
TDS, wastewater 1: TDS 164,000 ppm and wastewater 2: TDS
258,000 ppm) were supplied by Doosan Heavy Industries and
Construction (Korea) and were obtained from the desulfuriser
apparatus. TFC membranes were supplied by Toray Industries,
Inc. (Japan) and CTA membranes were purchased from
Hydration Technology Innovations (USA). Reverse osmosis
membranes (ACM4 membranes) were purchased from TriSep
(USA).
Imperfect rejection of draw solutes by semipermeable
membranes produces concentrated FSs as well as product
water containing a small amount of leaked draw solutes.
Although the leakage of HM8I and HM10I was much less than
that of NH4HCO3, their potential biological toxicity was
checked by conventional cytotoxicity assay using 3-(4,5-
dimethylthiazol-2-yl)-2,5-diphenyl tetrazolium bromide (MTT).
They showed almost no cytotoxicity at a concentration of 1 mg
mL-1, and only marginal cytotoxicity above this concentration
(Figure S3d, ESI†). Of course, other toxicity tests using animal
models including acute oral toxicity, long-term toxicity, and
mutagenicity test would be further required for practical use
of our draw solutes in the future.
Synthetic draw solutes
Synthesis of monoammonium iodide salts (3MBAI, 3MOAI and
3PEAI)
Conclusions
In summary,
a variety of UCST-type thermoresponsive
ammonium iodide salts were systematically designed and
synthesised considering osmolalities, Tp and other phase
separation properties, which were highly dependent upon
ionic interactions, hydrophobicity, and the molecular
symmetry of the salt structure. Two diammonium salts with a
long hydrophobic chain, HM8I and HM10I, exhibited high
maximum osmolalities, appropriate Tp ranges, broad
concentration ranges for the phase transition, and minimal
residual osmolality. HM8I and HM10I solutions provided
sufficient osmotic pressure to draw water from highly
concentrated FGD wastewater with minimal solute leakage
through the membrane. As the osmolality of the phase-
separated DS dropped to less than one tenth that of the initial
FS, a considerably reduced pressure was required for the final
production of fresh water. This is the first study to design
thermoresponsive materials with tailored properties as draw
solutes, and to show their feasibility in FO purification of highly
concentrated FGD wastewater. We propose that our approach
to the systematic design of thermoresponsive materials
The synthetic scheme is shown in Figure S1. Each starting
material (n-butylamine, n-octylamine or tripropylamine for
3MBAI, 3MOAI, and 3PEAI, respectively) was dissolved in
acetonitrile or methanol (0.25 M). For a starting material
based on a primary amine (n-butylamine or n-octylamine), a
base (DIPEA or KHCO3) was added as a proton quencher. For
the tertiary amine-based starting material (tripropylamine), no
additional base was required. Excess 1-iodomethane was then
slowly added to the amine solution. The mixture was stirred
for 48
h at 40°C (ambient temperature for 3MOAI). The
specific amount of each reagent and the reaction conditions
are summarised in Table S1 and Figure S1. To purify 3MBAI
and 3PEAI, the reaction mixture was evaporated and the
remaining products were diluted with a small amount of
deionised water. The diluted aqueous solution was added to a
diethyl ether/ethanol mixture (7:3) to precipitate the product.
After iterative precipitation (×8), the precipitate was collected
and dried overnight at 70°C under vacuum (yields > 80%). To
purify 3MOAI, the reaction mixture was filtered and the filtrate
This journal is © The Royal Society of Chemistry 20xx
J. Name., 2013, 00, 1-3 | 10
Please do not adjust margins

Please do not adjust margins
Journal of Materials Chemistry A
Page 11 of 14
View Article Online
DOI: 10.1039/C7TA09741G
Journal Name
ARTICLE
was evaporated. The slurry was extracted with Measurement of the FO water flux
dichloromethane, and the dichloromethane fraction was
The FO water flux was measured with two instrumental
systems. One system, consisting of handmade U-shaped glass
tubes, was used for small-scale analysis (dead-end type)
(Figure S2a), and the other system, based on a cross-flow
instrument, was used for larger-scale analysis (Figure S2b). All
flux values were measured under temperature-controlled
collected and evaporated. The remaining product was
dissolved in a small amount of ethanol and precipitated into
diethyl ether. The precipitate was collected and dried
overnight at 70°C under vacuum (yield > 50%). Figure S1a-c
shows 1H NMR spectra of 3MBAI, 3MOAI, and 3PEAI.
Synthesis of diammonium iodide salts (HM2I, HM4I, HM6I, conditions.12 A semipermeable membrane (TFC membrane,
HM8I, and HE2I)
Toray Industries, Inc., Japan) with a diameter of 3.80 cm (dead-
end) or 2.50 cm (cross-flow) was placed between a feed
The synthetic scheme is shown in Figure S1. Each starting
solution and
a draw solution. In the cross-flow-type
material
(
N,N,N’,N’-tetramethylethylenediamine, N,N,N’,N’
-
measurement, the flow rates of the feed solution and the draw
solution were controlled at 0.5 L min-1 with two micropumps
(Sterlitech Corporation, USA). The active layer of the
membrane faced the feed solution (FO mode). The FO flux
from the feed solution to the draw solution was measured
using the volumetric change in each solution after 30 min
(dead-end) or 150 min (cross-flow) flux stabilization.
tetramethyl-1,4-diaminobutane,
N,N,N’,N’-tetramethyl-1,6-
diaminohexane, 1,8-diaminooctane, and ethylenediamine for
HM2I, HM4I, HM6I, HM8I, and HE2I, respectively) was
dissolved in acetonitrile or methanol (0.25 M). For a starting
material based on a primary amine, DIPEA was added to the
solution as a proton quencher. For the tertiary amine-based
starting material (tripropylamine), no additional base was
required. Excess 1-iodomethane was then slowly added to the Measurement of reverse solute flux
diamine solution. The specific amount of each reagent is
The reverse solute flux was measured using the system based
on handmade glass tubes used for the FO water flux under
temperature-controlled conditions.12 The reverse solute flux
was compared through two different semipermeable
membranes (TFC membrane (Toray Industries, Inc., Japan) and
CTA membrane (Hydration Technology Innovations, USA)). The
membrane (diameter: 3.80 cm) was placed between a feed
solution (deionised water) and the draw solution. The active
layer faced the feed solution. After 30 min FO, the volume and
the conductivity of the feed solution were measured to
calculate the reverse solute flux. The conductivity was
measured with an S470 SevenExcellence™ pH/conductivity
meter (Mettler Toledo, Switzerland). The conductivity was
converted into a concentration based on the concentration-
conductivity standard curves of the draw solutions.
described in Table S1. The mixture was stirred for 48
h at 40°C.
The reaction mixture was evaporated and the remaining
products were diluted with a small amount of deionised water.
The diluted aqueous solution was added to a diethyl
ether/ethanol mixture (7:3) to precipitate the product. After
iterative precipitation (×8), the precipitate was collected and
1
dried overnight at 70°C under vacuum (yields > 80%). The H
NMR spectra of HM2I, HM4I, HM6I, HM8I and HE2I are shown
in Figure S1d-f, h and j, ESI† respectively. The 13C NMR spectra
of HM6I and HM8I are shown in Figure S1g and i of ESI†
respectively.
Measurement of UCST phase transition
Changes in light transmittance were measured at a wavelength
of 600 nm to observe the UCST phase transition of aqueous
solutions of ammonium iodide salts. The solutions were Thermal stability of draw solutes
prepared by dissolving the salts in deionised water at 60°C.
Aqueous solutions of HM8I and HM10I at 60 wt% were
The transmittance change was measured using a V-650 UV-Vis
spectrophotometer (Jasco, Japan) with temperature change at
a cooling rate of 1.5°C min-1. The phase-transition temperature
was defined as the point when the transmittance is 50% during
the cooling process.
exposed to 60 cooling-heating cycles mimicking the
temperature-controlled FO process. A cycle consisted of two
incubation steps at 10°C and 60°C for 1 h. After the 60 cycles,
the draw solutions were further incubated at 80°C for 60 days.
The structural stability of the draw solutes was confirmed by
both 1H NMR and LC/MS (Agilent 1260 Infinity (Agilent, USA)).
Measurement of osmolality
The osmolality of the ammonium iodide salt solutions at a Cytotoxicity assay
temperature above the phase-transition temperature was
An MTT assay was conducted to measure the cytotoxicity of
measured by vapour pressure osmometry (VPO), K-7000
(Knauer, Germany), at 60°C or 65°C. After phase separation of
the draw solution at a temperature below the phase-transition
temperature, the upper layer (water-rich layer) of the phase-
separated draw solution was carefully collected to measure
the osmolality by freezing point osmometry (FPO), Micro-
Osmometer Model 210 (Fiske, Germany).
HM8I and HM10I. HeLa (human cervical cancer) cells were
seeded at 5.0 × 103 cells per well in a 96 well plate in 90 mL of
DMEM containing 10% FBS, and further incubated at 37°C for
24
sample was added at various concentrations to the cells with
subsequent incubation at 37°C for 24 . For the assay, the cells
h. To determine the cytotoxicity of HM8I and HM10I, each
h
were washed with DPBS, followed by the addition of 20 μL of
This journal is © The Royal Society of Chemistry 20xx
J. Name., 2013, 00, 1-3 | 11
Please do not adjust margins

Please do not adjust margins
Journal of Materials Chemistry A
Page 12 of 14
View Article Online
DOI: 10.1039/C7TA09741G
Journal Name
ARTICLE
filtered MTT solution (2 mg mL-1 in DPBS). After incubation at
5
D. Zhao, P. Wang, Q. Zhao, N. Chen and X. Lu, Desalination,
2014, 348, 26; M. Xie, L. D. Nghiem, W. E. Price and M.
Elimelech, Environ. Sci. Technol., 2013, 47, 13486; Q. Ge, P.
Wang, C. Wan and T.-S. Chung, Environ. Sci. Technol. Lett.,
2012, 46, 6236; N. C. Nguyen, S.-S. Chen, S. Jain, H. T.
Nguyen, S. S. Ray, H. H. Ngo, W. Guo, N. T. Lam and H. C.
Duong, Environ. Sci. Pollut. Res., 2017 (in press).
37°C for 2 h, the medium was removed from the well and 150
μL of DMSO was added to dissolve insoluble formazan
particles. The absorbance was measured at 570 nm using a
microplate reader (Molecular Devices Co., Menlo Park, CA).
The relative cell viability (%) was defined as the relative
absorbance compared to the cells treated with DPBS buffer
only, which were used as the control.
6
7
Q. Ge, M. Ling and T.-S. Chung, J. Membr. Sci., 2013, 442,
225; R. L. McGinnis, N. T. Hancock, M. S. Nowosielski-
Slepowron and G. D. Mcgurgan, Desalination, 2013, 312, 67.
J. R. McCutcheon, R. L. McGinnis and M. Elimelech,
Desalination, 2005, 174, 1; J. R. McCutcheon, R. L. McGinnis
and M. Elimelech, J. Membr. Sci., 2006, 278, 114; R. L.
McGinnis and M. Elimelech, Desalination, 2007, 207, 370.
Y. Zhong, X. Feng, W. Chen, X. Wang, K.-W. Huang, Y. Gnanou
and Z. Lai, Environ. Sci. Technol., 2016, 50, 1039; X. Fan, H.
Liu, Y. Gao, Z. Zou, V. S. J. Craig, G. Zhang and G. Liu, Adv.
Mater., 2016, 28, 4156; Y. Cai, W. Shen, J. Wei, T. H. Chong, R.
Wang, W. B. Krantz, A. G. Fane and X. Hu, Environ. Sci.:
Post-treatment of water-rich layer
Reverse osmosis (RO) water treatment was performed using a
lab-scale dead-end RO filtration setup at room temperature.27
The effective membrane area was (2.16)2·π cm-2 and the
operating pressure was 15.5 bar. The rejection percentage (%),
the percentage of rejected solute in the permeated feed
solution across the membrane during the RO, was calculated
from the osmolality, which was measured by FPO, Micro-
Osmometer Model 210 (Fiske, Germany), to estimate the
efficiency of the RO-based rejection on the post-water
treatment.
8
9
Water Res. Technol., 2015,
Guo, D. Zhao, J. Ding and X. Lu, RSC Adv., 2016,
1
, 341; Q. Zhao, T. S. Herng, C. X.
, 15731.
6
D. Li, X. Zhang, J. Yao, G. P. Simon and H. Wang, Chem.
Commun., 2011, 47, 1710; A. Razmjou, Q. Liu, G. P. Simon
and H. Wang, Environ. Sci. Technol., 2013, 47, 13160; D. Li, X.
Zhang, G. P. Simon and H. Wang, Water Res., 2013, 47, 209;
H. Zhang, J. Li, H. Cui, H. Li and F. Yang, Chem. Eng. J., 2015,
259, 814; R. Ou, H. Zhang, G. P. Simon and H. Wang, J.
Membr. Sci., 2016, 510, 426.
Conflicts of interest
10 Q. Zhao, N. Chen, D. Zhao and X. Lu, ACS Appl. Mater.
Interfaces, 2013, , 11453; C. X. Guo, D. Zhao, Q. Zhao, P.
There are no conflicts to declare.
5
Wang and X. Lu, Chem. Commun., 2014, 50, 7318; M. M. Ling
and T.-S. Chung, Desalination, 2011, 278, 194; D. Zhao, S.
Chen, C. X. Guo, Q. Zhao and X. Lu, Chin. J. Chem. Eng., 2016,
24, 23; M. M. Ling, T.-S. Chung and X. Lu, Chem. Commun.,
2011, 47, 10788.
Acknowledgements
This work was supported by Doosan Heavy Industries and
Construction, Korea (305-20160003) and by the National
Research Foundation of Korea Grant funded by the Korean
Government (NRF-2015R1C1A1A01052185).
11 Y. Cai and X. ‘M.’ Hu, Desalination, 2016, 391, 16.
12 M. Noh, Y. Mok, S. Lee, H. Kim, S. H. Lee, G.-w. Jin, J.-H. Seo,
H. Koo, T. H. Park and Y. Lee, Chem. Commun., 2012, 48
3845; Y. Mok, D. Nakayama, M. Noh, S. Jang, T. Kim and Y.
Lee, Phys. Chem. Chem. Phys., 2013, 15 19510; D.
,
,
Nakayama, Y. Mok, M. Noh, J. Park, S. Kang and Y. Lee, Phys.
Chem. Chem. Phys., 2014, 16, 5319; J. Park, M. Noh, M. K.
Notes and references
Chey, Y. Mok, J.-H. Seo and Y. Lee, RSC Adv., 2016,
13 R. B. Griffiths and J. C. Wheeler, Phys. Rev. A, 1970,
6
2
, 26526.
, 1047.
14 F. J. Millero, R. Feistel, D. G. Wright and T. J. McDougall,
1
F.-R. Ahmadun, A. Pendashteh, L. C. Abdullah, D. R. A. Biak, S.
S. Madaeni and Z. Z. Abidin, J. Hazard. Mater., 2009, 170
530; V.K. Gupta, I. Ali, T. A. Saleh, A. Nayak and S. Agarwal,
RSC Adv., 2012, , 6380; G. K. Pearce, Desalination, 2008,
,
Deep-Sea Res. Part I Oceanogr. Res. Pap., 2008, 55, 50; R. D.
Norris and P. A. Wilson, Geology, 1998, 26, 823; N.
2
222, 66; B. Nicolaisen, Desalination, 2002, 153, 355; A.
Subramani and J. G. Jacangelo, Water Res., 2015, 75, 164.
T.-S. Chung, X. Li, R. C. Ong, Q. Ge, H. Wang and G. Han, Curr.
Voutchkov, Desalination, 2011, 273 205; K.-C. Emeis,
,
Palaeogeogr. Palaeoclimatol. Palaeoecol., 2000, 158, 259; M.
Shuangchen, C. Jin, C. Gongda, Y. Weijing and Z. Sijie, Renew.
Sust. Energ. Rev., 2016, 58, 1143.
2
Opin. Chem. Eng., 2012,
1, 246; K. Lutchmiah, A.R.D.
Verliefde, K. Roest, L.C. Rietveld and E.R. Cornelissen, Water
Res., 2014, 58, 179; S. Zhao, L. Zou, C. Y. Tang and D.
Mulcahy, J. Membr. Sci., 2012, 396, 1; T. Y. Cath, A. E.
Childress and M. Elemelech, J. Membr. Sci., 2006, 281, 70; E.
Mathioulakis, V. Belessiotis and E. Delyannis, Desalination,
2007, 203, 346.
15 C. S. Slater, R. C. Ahlert and C. G. Uchrin, Desalination, 1983,
48, 171; R. J. Ray, J. Kucera-Gienger and S. Retzlaff, J. Membr.
Sci., 1986, 28, 87; M. S. Mohsen, Desalination, 2004, 167, 75.
16 H. G. Schild, Prog. Polym. Sci., 1992, 17, 163; S. Furyk, Y.
Zhang, D. Ortiz-Acosta, P. S. Cremer and D. E. Bergbreiter, J.
Polym. Sci. Part A: Polym. Chem., 2006, 44, 1492; H. Kim, S.
Lee, M. Noh, S. H. Lee, Y. Mok, G.-w. Jin, J.-H. Seo and Y. Lee,
Polymer, 2011, 52, 1367.
17 M. Noh, Y. Mok, D. Nakayama, S. Jang, S. Lee, T. Kim and Y.
Lee, Polymer, 2013, 54, 5338.
18 M. Noh, S. Kang, Y. Mok, S. J. Choi, J. Park, J. Kingma, J.-H.
Seo and Y. Lee, Chem. Commun., 2016, 52, 509.
19 H. Yoshimitsu, A. Kanazawa, S. Kanaoka and S. Aoshima,
Macromolecules, 2012, 45, 9427.
20 Y. Marcus, J. Chem. Soc. Faraday Trans., 1991, 87, 2995; J.
Heyda, M. Lund, M. Oncak, P. Slavicek and P. Jungwirth, J.
Phys. Chem. B, 2010, 114, 10843.
3
4
M. S. El-Bourawi, Z. Ding, R. Ma and M. Khayet, J. Membr.
Sci., 2006, 285, 4; A. Alkhudhiri, N. Darwish and N. Hilal,
Desalination, 2012, 287, 2; S. Adham, A. Hussain, J. M.
Matar, R. Dores and A. Janson, Desalination, 2013, 314, 101.
M. A. Shannon, P. W. Bohn, M. Elimelech, J. G. Georgiadis, B.
J. Mariñas and A. M. Mayes, Nature, 2008, 452, 301; J. Koo,
S. Lee, Y. Shin and T. Yun, Trans. Korean Soc. Mech. Eng. B,
2015, 39, 279; H. Luo, Q. Wang, T. C. Zhang, T. Tao and A.
Zhou, J. Water, Process. Eng., 2014, 4, 212; T. S. Chung, X. Li,
R. C. Ong, Q. Ge, H. L. Wang, G. Han, Curr. Opinion Chem.
Eng., 2012, 1, 1246.
This journal is © The Royal Society of Chemistry 20xx
J. Name., 2013, 00, 1-3 | 12
Please do not adjust margins

Please do not adjust margins
Journal of Materials Chemistry A
Page 13 of 14
View Article Online
DOI: 10.1039/C7TA09741G
Journal Name
ARTICLE
21 I. Krossing, J. M. Slattery, C. Daguenet, P. J. Dyson, A.
Oleinikova and H. Weingatner, J. Am. Chem. Soc., 2006, 128
13427; A. F. Kapustinskii, Q. Rev. Chem. Soc., 1956, 10, 283; R.
Pinal, Org. Biomol. Chem., 2004, , 2692.
,
2
22 K. A. Dil, T. M. Truskett, V. Vlachy and B. Hribar-Lee, Annu.
Rev. Biophys. Biomol. Struct., 2005, 34, 173.
23 M. Caon, Contemp. Nurse, 2008, 29, 92.
24 Y. H. Cho, J. Han, S. Han, M. D. Guiver and H. B. Park, J.
Membr. Sci., 2013, 445, 220.
25 J. Wei, C. Qiu, C. Y. Tang, R. Wang and A. G. Fane, J. Membr.
Sci., 2011, 372, 292; N. Y. Yip, A. Tiraferri, W. A. Phillip, J. D.
Schiffman and M. Elimelech, Environ. Sci. Technol., 2010, 44
3812.
,
26 K. D. Vos, F. O. Burris, JR. and R. L. Riley, J. Appl. Polym. Sci.,
1966, 10, 825.
27 H. J. Kim, K. Choi, Y. Baek, D.-G. Kim, J. Shim, J. Yoon and J.-C.
Lee, ACS Appl. Mater. Interfaces, 2014, 6, 2819.
28 Y. Zhong, X. Feng, W. Chen, X. Wang, K.-W. Huang, Y. Gnanou
and Z. Lai, Environ. Sci. Technol., 2016, 50, 1039.
This journal is © The Royal Society of Chemistry 20xx
J. Name., 2013, 00, 1-3 | 13
Please do not adjust margins

Journal of Materials Chemistry A
Page 14 of 14
View Article Online
DOI: 10.1039/C7TA09741G
Systematic structure control of ammonium iodide salts is designed as novel
UCST-type forward osmosis draw solutes for wastewater treatment.