Synthesis, silver (I) extraction and silver (I) binding studies of novel N 1,N 3-bis(2-(benzylthio)ethyl)propanediamide derivatives

Article abstract of DOI:10.1007/s11696-020-01307-x

Abstract: Solvent or liquid–liquid extraction represents a highly valuable technique for the selective recovery of metals from the aqueous phase due to the ease of operation and short turnaround times. Ligands bearing soft donor atoms including nitrogen and sulfur are ideal candidates for selective silver recovery due to their preference for silver binding. Herein, novel N¹,N³-bis(2-(benzylthio)ethyl)propanediamide derivatives bearing sulfur and nitrogen donor atoms were prepared in low to high yields and tested for Ag⁺ extraction from ternary aqueous solutions also containing Cu²⁺ and Pb²⁺ following a well-established solvent extraction protocol. It was observed that electronics effects at the 4-aryl position in the propanediamide (or malondiamide) derivatives had a significant effect on the selectivity, but little effect on the efficiency of Ag⁺ extraction with the 4-methoxy analogue proving the most selective. Steric hindrance provided by dimethyl substitutions at the α-positions to the sulfur atoms had negative effects on Ag⁺ extraction efficiency and selectivity, while diethyl steric hindrance at the methylene center lowered selectivity but increased extraction efficiency for Ag⁺. Detailed binding studies reveal that one of the malondiamide derivatives which lacked the electronic and steric hindrance groups studied coordinated Ag⁺ in a 1:1 fashion suggesting a tetrahedral complex geometry. Overall, the results show that simple modification of the electronics and sterics of the N¹,N³-bis(2-(benzylthio)ethyl)propanediamides, can improve their selectivity for Ag⁺ recovery from the aqueous phase. Graphic abstract: [Figure not available: see fulltext.].

Full text of DOI:10.1007/s11696-020-01307-x

Chemical Papers
https://doi.org/10.1007/s11696-020-01307-x
ORIGINAL PAPER
Synthesis, silver (I) extraction and silver (I) binding studies of novel
N1,N3‑bis(2‑(benzylthio)ethyl)propanediamide derivatives
Abiodun D. Aderibigbe1,2 · Andrew J. Clark1
Received: 21 May 2020 / Accepted: 6 August 2020
© Institute of Chemistry, Slovak Academy of Sciences 2020
Abstract
Solvent or liquid–liquid extraction represents a highly valuable technique for the selective recovery of metals from the aque-
ous phase due to the ease of operation and short turnaround times. Ligands bearing soft donor atoms including nitrogen and
sulfur are ideal candidates for selective silver recovery due to their preference for silver binding. Herein, novel N1,N3-bis(2-
(benzylthio)ethyl)propanediamide derivatives bearing sulfur and nitrogen donor atoms were prepared in low to high yields
and tested for Ag+ extraction from ternary aqueous solutions also containing Cu2+ and Pb2+ following a well-established
solvent extraction protocol. It was observed that electronics efects at the 4-aryl position in the propanediamide (or malon-
diamide) derivatives had a signifcant efect on the selectivity, but little efect on the efciency of Ag+ extraction with the
4-methoxy analogue proving the most selective. Steric hindrance provided by dimethyl substitutions at the α-positions to
the sulfur atoms had negative efects on Ag+ extraction efciency and selectivity, while diethyl steric hindrance at the meth-
ylene center lowered selectivity but increased extraction efciency for Ag+. Detailed binding studies reveal that one of the
malondiamide derivatives which lacked the electronic and steric hindrance groups studied coordinated Ag+ in a 1:1 fashion
suggesting a tetrahedral complex geometry. Overall, the results show that simple modifcation of the electronics and sterics of
the N1,N3-bis(2-(benzylthio)ethyl)propanediamides, can improve their selectivity for Ag+ recovery from the aqueous phase.
Graphic abstract
Keywords Malondiamide derivatives · Selective silver(I) extraction · Electronics · Sterics · Job plot · 1H NMR titration
Electronic supplementary material The online version of this
Introduction
article (https://doi.org/10.1007/s11696-020-01307-x) contains
supplementary material, which is available to authorized users.
As the world population increases, the demand for prod-
* Abiodun D. Aderibigbe
ucts like phones, laptops, batteries and jewelry, contain-
ing precious metals like silver is certain to increase as well
abiodunadaderibigbe23@gmail.com
Extended author information available on the last page of the article
Vol.:(0123456789)
1 3

Chemical Papers
(Taillades and Sarradin 2004; Sahan et al. 2019). Unfor-
tunately, silver just like other metals is a non-renewable
resource. Therefore, to meet the increasing demand, the
supply of silver must be guaranteed. Two major options that
have been explored to guarantee the supply of silver and
indeed other metals is the search for new primary sources or
new metal ore deposits (Kolpakova et al. 2019), and the sec-
ondary sources or urban mining, in other words, recycling of
metals (Avarmaa et al. 2019). The success of these options
rests on the development of an efcient, metal selective and
cost-efective Ag+ recovery process.
(Sun et al. 2017). Furthermore, the disulphide extractant—
bis(2,4,4-trimethylpentyl)dithiophosphinic acid-containing
sulfur and phosphorus donors was reported to demonstrate
remarkable selectivity for Ag+ from a chloride solution con-
taining more than 1000-fold excess of other transition met-
als. Extraction efciencies of 98.6, 0.2, 0.6, 0.6 0.5 and 0.7%
were observed for Ag+, Ni2+, Cu2+, Co2+, Zn2+ and Fe3+,
respectively. The loaded extractant was completely stripped
of Ag+ using a 0.5 M thiourea solution containing a sulfur
donor (Grigorieva et al. 2018).
N1,N3-propanediamide or malondiamide derivatives
(MADs) such as 1a (Fig. 1) have attracted attention due to
their versatility for metal recovery in separation science,
having demonstrated outstanding efciency and selectivity
in the recovery of highly valuable metals including lantha-
nides, actinides and precious metals (Daubinet and Kaye
2002; Manchanda and Pathak 2004; Jańczewski et al. 2007;
Patil et al. 2014). The presence of two acyl groups and an
active methylene center allows for the modifcation of the
structural features of the MADs towards accessing deriva-
tives with selective metal recovery and desired solubility
properties (Daubinet and Kaye 2002). Addition of specifc
donors has been shown to improve the efciency and selec-
tivity of the MADs for metal recovery. For example, Daubi-
net and Kaye (2002) prepared a range of MADs (1a–d)
(Fig. 1) with diferent donor groups and varying degrees of
lipohilicities via a microwave-assisted method and observed
that the MAD 1a bearing a benzylthio arm demonstrated the
highest efciency (97%) and selectivity for Ag+ recovery
from aqueous solution also containing Cu2+ and Pb2+.
As part of an ongoing research project exploring the
development of technologies for recovery of Ag+ from
waste repositories (including metallurgical slag heaps,
mine tailings and possibly landfills), we queried the
possibility of accessing a Ag+-selective ligand bearing
a vinyl handle which can be tethered by polymerization
to a magnetic nanoparticle. The MAD 1e, (Fig. 1) was
To date, a considerable number of studies have reported
on the recovery of Ag+ by membrane fltration (Zanain and
Lovitt 2013), ion exchange (Virolainen et al. 2015), elec-
trochemical treatment (Ölmez 2009; Selim et al. 2017),
solid phase extraction/adsorption (Chen and Wang 2008;
El-Shafey and Al-Kindy 2013; Aderibigbe et al. 2020) and
solvent extraction (Saha et al. 2013; Bray et al. 2015). Sol-
vent extraction is a popular technique due to the ease of
operation, low operating costs, short turnaround times and
the ability of extractants to form stable complexes with the
metal ion (Nishihama et al. 2001; Grigorieva et al. 2018).
Extractants containing at least one of phosphorus, nitrogen
or sulfur donor atoms have been reported to be selective for
the extraction of Ag+ from the aqueous phase consistent with
the hard/soft acid/base rule (Pearson 1968; Alam et al. 1997;
Ohto et al. 1997; Shimojo and Goto 2004; Saha et al. 2013).
For example, Bray et al. (2015) reported that the ligand—
1,3,5-tris((pyridin-4-ylthio)methyl)benzene bearing sulfur
donors demonstrated better efciency and selectivity for Ag+
compared with other ligands containing oxygen donors. In a
diferent study, an oxime-based extractant (LIX63) contain-
ing a nitrogen donor was found to selectively extract Ag+
over Zn2+, achieving extraction efciencies of 60% and 0%
for Ag+ and Zn2+, respectively, from the aqueous phase. The
extracted Ag+ was later stripped from the LIX63 extractant
using aqueous ammonia also containing a nitrogen donor
Fig. 1 N1,N3-bis(2-(benzylthio)
ethyl)propanediamide deriva-
tives 1a–e
1 3

Chemical Papers
chosen as our initial target after we had judged that it
could be easily prepared and that it has high stability in
low pH environments. Unfortunately, previously reported
conditions for accessing the related ligand 1a involve high
temperatures which we feared could occasion the polym-
erization of the important vinyl group in 1e (Daubinet
and Kaye 2002). Therefore, initial study focused on the
development of a room temperature synthetic route to
accessing derivatives of 1a; 1e–i (Fig. 2). Subsequently
structure–activity relationship of the synthesized ligands
(1a, 1e–i), in particular the effect of electronics (using
1a and 1e–h) and sterics (using 1a, 1i and 3a) (Fig. 2)
on Ag+ extraction was investigated to understand how
modification affects the selectivity and efficiency of the
ligand. Finally, the binding stoichiometry of the Ag+-1a
complex, as an example, was thoroughly investigated to
gain an insight into the Ag+ binding nature of the N1,N3-
bis(2-(benzylthio)ethyl)propanediamide derivatives pre-
pared in this study.
Methods
General synthesis procedure for the 2‑(benzylthio)
ethan‑1‑amine derivatives (2a–e)
The 2-(benzylthio)ethan-1-amine derivatives 2a–e were
synthesized following the method reported by Ghosh and
Tochtrop (2009). Briefy, to a stirred H2O/ethanol (1:3) solu-
tion of LiOH (2 eq.) and 2-aminoethanethiol hydrochloride
(1 eq.) was added dropwise for 5 min, the appropriate benzyl
chlorides (1 eq.) and the reaction was left to stir at 35 °C for
20 min (for 2a) or 40 min (for 2b–e) after which the solvent
was removed in vacuo. The crude mixture was extracted
with dichloromethane (DCM) after it had been solubilized
with H2O. The DCM extract was then dried with anhydrous
Na2SO4 or MgSO4, fltered and concentrated in vacuo. All
amines except for 2d (the fuoro analogue) were obtained as
oils needing no purifcation. The amine 2d was subsequently
purifed by column chromatography using a mobile phase
gradient of 100% EtOAc to 50% v/v EtOAc/MeOH.
2‑(Benzylthio)ethan‑1‑amine (2a) (Ghosh and Tochtrop
2009) Viscous colorless oil. Yield: 5.76 g (86%), ʋ(cm−1)
Experimental details
1
3363, 3279 (N–H stretch), 1600 (N–H bend), H NMR
Chemicals and instrumentation
(400 MHz, CDCl3) δ 7.39 – 7.18 (m, 5H, ArH), 3.70 (s,
2H, PhCH2S), 2.81 (t, J=6.5 Hz, 2H, SCH2CH2), 2.51 (t,
J = 6.5 Hz, 2H, SCH2CH2), 1.29 (s, 2H, CH2CH2NH2),
13C NMR (100 MHz, CDCl3) 138.5, 128.9, 128.6 and
127.1 (ArC), 40.9 (CH2CH2NH2), 36.0 (PhCH2S), 35.7
(SCH2CH2); m/z (ESI): 168 [M+H]+.
All reagents were purchased from commercial sources
(Sigma Aldrich, VWR, Alfa Aesar and Acros Organ-
ics) and used as received. Proton (1H) and Carbon-13
(13C) NMR spectra were recorded at room temperature
on Bruker Advance spectrometers and all chemical shift
values were referenced to an internal standard of tetra-
methylsilane (δ = 0.0 ppm). Fourier transform infra-red
(FTIR) spectra were recorded on Bruker Alpha Platinum-
Attenuated Total Reflectance IR spectrometer. All mass
spectra were run on a Bruker MaXis mass spectrometer.
Metal concentrations were measured by means of a Perki-
nElmer 5300DV Inductively Coupled Plasma Optical
Emission spectrophotometer (ICP-OES).
2‑((4‑Vinylbenzyl)thio)ethan‑1‑amine (2b) Viscous yel-
low oil. Yield: 5.87 g (77%),, ʋ(cm−1) 3365, (N–H stretch),
1
1627, 1567 (N–H bend), H NMR (300 MHz, CDCl3)
δ 7.36 (d, J = 8.0 Hz, 2H, ArH), 7.27 (d, J = 8.0 Hz, 2H,
ArH), 6.70 (dd, J=17.5, 11.0 Hz, 1H, CH2CHPh), 5.73 (d,
J = 17.5 Hz, 1H, CHaHbCHPh), 5.23 (d, J = 11.0 Hz, 1H,
CHaHbCHPh), 3.69 (s, 2H, PhCH2S), 2.82 (t, J = 6.0 Hz,
2H, CH2CH2NH2), 2.51 (t, J=6.0 Hz, 2H, SCH2CH2), 1.29
(s, 2H, CH2CH2NH2), 13C NMR (75 MHz, CDCl3) δ 138.1
Fig. 2 N1,N3-bis(2-(benzylthio)ethyl)propanediamide derivatives synthesized in this study
1 3

Chemical Papers
and 136.5 (ArC), 136.4 (CH2CHPh), 129.1 and 126.5 (ArC),
113.9 (CH2CHPh),41.0 (CH2CH2NH2), 35.8 (PhCH2S), 35.7
(SCH2CH2), m/z (ESI): [M + H]+ for [C11H16NS]+, calcu-
lated; 194.1003, found 194.0998.
concentrated in vacuo and purifed by fash chromatography
using n-hexane as eluent to yield 1-(benzylthio)-1,1-dimeth-
ylnitroethane 2f (2.60 g, 23%). In the second step, a dry
ether solution of 1-(benzylthio)-1,1-dimethylnitroethane
2f (1 g, 4.4 mmol) was added to a cold ether solution of
LiAlH4 (4 M, 3.25 mL) inside a two-necked fask. The mix-
ture was left to stir for another 3 min after which it was
transferred to a heating block and heated under refux for
100 min. The crude mixture was transferred to a 250 mL
fask and a stir bar added. H2O and subsequently potas-
sium sodium tartarate (25 mL, 20% w/w) was added. The
mixture was left to stir until all solids dissolved. The crude
product was then extracted with ether (20 mL × 3) and
purifed by fash chromatography (16 – 50% ethyl acetate:
n-hexane, ethyl acetate and fnally methanol) to furnish
2-(benzylthio)-2,2-dimethylmethylamine 2g as a yellow oil.
2‑((4‑Methoxybenzyl)thio)ethan‑1‑amine (2c) (Ghosh and
Tochtrop 2009) Viscous colorless oil. Yield: 4.37 g (88%),
1
ʋ(cm−1): 3261 (N–H stretch), 1663 (N–H bend), H NMR
(400 MHz, CDCl3) δ 7.23 (d, J = 8.5 Hz, 2H and 6.84 (d,
J=8.5 Hz, 2H) (ArH), 3.79 (s, 3H, CH3OPh), 3.67 (s, 2H,
PhCH2S), 2.85 (t, J=6.5 Hz, 2H, CH2CH2NH2), 2.71 (s, 2H,
CH2CH2NH2), 2.55 (t, J=6.5 Hz, 2H, SCH2CH2). 13C NMR
(100 MHz, CDCl3) δ 159.0, 130.2, 130.0 and 114.1 (ArC),
55.4 (CH3OPh), 40.5 (CH2CH2NH2), 35.4 (PhCH2,S), 34.5
(SCH2CH2), m/z (ESI): 198 [M+H]+.
1
2‑((4‑Fluorobenzyl)thio)ethan‑1‑amine (2d) (Ghosh and
Tochtrop 2009) Viscous colorless oil. Yield: 4.27 g (85%),
Yield: (0.60 g, 70%), ʋ (cm−1) 3376 (NH), H (300 MHz
CDCl3) 7.40–7.19 (m, 5H, ArH), 3.68 (s, 2H, PhCH2S), 2.61
(s, 2H, CH2NH2), 1.44 (s, 2H, NH2), 1.28 (s, 6H, SC(CH3)2);
13C (75 MHz, CDCl3) 138.5, 129.0, 128.7 and 127.1
(ArC), 51.8 (C(CH3)2CH2NH2), 48.9 (C(CH3)2CH2NH2),
32.8 (PhCH2S), 26.6 (C(CH3)2CH2NO2), m/z (ESI): 196
[M+H]+.
1
ʋ(cm−1): 3365 (N–H stretch), 1599 (N–H bend), H NMR
(400 MHz, CDCl3) δ 7.35–7.23 (dd, 2H, JHCCH = 9 Hz,
JHCCCH = 3 Hz, ArH), 7.06–6.96 (t, 2H, JHCCF = 15 Hz,
ArH), 3.68 (s, 2H, PhCH2S), 2.82 (t, J = 6.5 Hz, 2H,
CH2CH2NH2), 2.51 (t, J = 6.5 Hz, 2H, SCH2CH2), 1.33
(s, 2H, CH2CH2NH2), 13C NMR (100 MHz, CDCl3) δ
162 (d, JCF = 244 Hz), 134.2 (d, JCCCCF = 3 Hz), 130.4
(d, JCCF = 8 Hz) and 115.4 (d, JCCF = 21 Hz) (ArC), 40.9
(CH2CH2NH2), 35.6 (PhCH2S), 35.2 (SCH2CH2), m/z (ESI):
186 [M+H]+.
General synthesis procedure
for the N1,N3‑bis(2‑(benzylthio)ethyl)propanediamide
derivatives (1a and 1e‑i)
To a stirred solution of the amines 2a–e, 2g (1 eq.) in dry
THF, malonyl chloride (0.25 eq.) solution in dry THF was
added dropwise for 90 min. The reaction mixture was left
to stir overnight and the THF solvent was removed in vacuo
after which H2O was added and the crude product was
extracted with EtOAc. The EtOAc extracts were washed
successively with HCl (2 M), saturated NaHCO3, H2O and
fnally brine. It was then dried over Na2SO4 and concentrated
in vacuo and purifed by fash chromatography (20% EtOAc/
Hexane to 100% EtOAc) to deliver the MADs 1a and 1e–i.
2‑((4‑Nitrobenzyl)thio)ethan‑1‑amine (2e) (Ghosh and
Tochtrop 2009) Yellow oil. Yield: 3.23 g (85%), ʋ(cm−1):
3365 (N–H stretch), 1599 (N–H bend), 1367 (N–O stretch),
1H NMR (300 MHz, CDCl3) δ 8.18 (d, J = 8.5 Hz, 2H,
ArH), 7.50 (d, J=8.5 Hz, 2H, ArH), 3.79 (s, 2H, PhCH2S),
2.85 (t, J= 6.5 Hz, 2H, CH2CH2NH2), 2.53 (t, J= 6.5 Hz,
2H, SCH2CH2), 1.38 (s, 2H, CH2CH2NH2), 13C NMR
(75 MHz, CDCl3) δ 147.1 (ArC), 146.4 (ArC), 129.7 (ArC),
123.9 (ArC), 40.9 (CH2CH2NH2), 35.8 (PhCH2S), 35.6
(SCH2CH2), m/z (ESI): 213 [M+H]+.
N1,N3‑bis(2‑(benzylthio)ethyl)propanediamide (1a) Yellow
powder. Yield: 0.21 g (33%), m.p.: 108–112 °C ((Daubinet
and Kaye 2002): 105–106 °C),, ʋ(cm−1): 3296 (amide N–H
Synthesis of 2,2‑dimethyl‑2‑benzylthioethylamine (2g)
1
Access to 2,2-dimethyl-2-benzylthioethylamine 2g was
gained following the method reported by Carroll et al.
(1963). In the frst step, a mixture of acetone (3.67 mL,
0.05 mol), benzylmercaptan (5.90 mL, 0.05 mol), nitrometh-
ane (2.71 mL, 0.05 mol) and benzene (18.75 mL) were
heated at 100 °C inside a fask ftted with a Dean Stark
apparatus flled with benzene. After 22 h, the crude mix-
ture still in the benzene solvent was left to cool to room
temperature, transferred into a separatory funnel where it
was washed with 2.0 M HCl (20 mL × 2) and then with
H2O (20 mL×2). It was then dried with anhydrous MgSO4,
stretch), 1647 (amide C = O bend), H NMR (400 MHz,
CDCl3) δ 7.40–7.19 (m, 10H, ArH), 7.05 (s, 2H, CON-
HCH2), 3.71 (s, 4H, PhCH2S), 3.39 (m, 4H, NHCH2CH2),
3.10 (s, 2H, COCH2CO), 2.56 (t, J=6.5 Hz, 4H, CH2CH2S),
13C NMR (100 MHz, CDCl3) δ 167.3 (COCH2CO), 138.2,
129.0, 128.8 and 127.3 (ArC), 43.1 (COCH2CO), 38.5
(NHCH2CH2), 36.0 (PhCH2S), 30.9 (CH2CH2S), m/z (ESI):
425 [M+Na]+.
N1,N3‑bis(2‑((4‑vinylbenzyl)thio)ethyl)propanediamide
(1e) Yellow powder. Yield: 0.09 g (33%), m.p.: 151–156 °C,
1 3

Chemical Papers
ʋ(cm−1): 3292 (amide N–H stretch), 1648 (amide C = O
bend), 1H NMR (400 MHz, CDCl3) δ 7.36 (d, J= 7.5 Hz,
4H) and 7.27 (d, J=7.5 Hz, 4H) (ArH), 6.94 (s, 2H, CON-
HCH2), 6.69 (dd, J=17.5, 11.0 Hz, 2H, PhCHcCHaHb), 5.73
(d, J=17.5 Hz, 2H, PhCHCHaHb), 5.24 (d, J=11.0 Hz, 2H,
PhCHcCHaHb), 3.70 (s, 4H, PhCH2S), 3.40 (m, CH2CH2NH),
3.11 (s, 2H, COCH2CO), 2.56 (t, J=6.5 Hz, 4H, SCH2CH2),
13C NMR (100 MHz, CDCl3) δ 167.2 (COCH2CO), 137.7
and 136.8 (ArC), 136.5 (CH2CHPh), 129.2 and 126.6 (ArC),
114.1 (PhCHCH2), 43.1 (COCH2CO), 38.4 (CH2CH2NH),
35.7 (CH2SCH2), 30.8 (SCH2CH2), m/z (ESI): [M+Na]+ for
[C25H27N2NaO2S2]+ calculated: 477.1646, found: 477.1643.
(PhCH2S), 31.0 (SCH2CH2), m/z (ESI): [M + Na]+ for
[C21H24N4NaO6S2]+ calculated: 515.1035, found: 515.1029.
N1,N3‑bis(2‑(benzylthio)‑2‑methylpropyl)propanediamide
(1i) Cream solid. Yield: 0.68 g (89%), m.p.: 71–74 °C,,
ʋ(cm−1): 3260 (amide N–H stretch), 1663 (amide C = O
bend), 1H NMR (300 MHz, CDCl3) δ 7.40–7.20 (m, 10H,
ArH), 7.01 (s, 2H, CH2NHCO), 3.71 (s, 4H, PhCH2S),
3.32 (d, J = 6.0 Hz, 4H, C(CH3)2CH2NH), 3.11 (s, 2H,
COCH2CO), 1.29 (s, 12H, SC(CH3)2CH2), 13C NMR
(75 MHz, CDCl3) δ 167.3 (NHCOCH2), 138.3 129.0, 128.8
and 127.2 (ArC), 48.0 (CH2NHCO), 46.6 (PhCH2S), 43.2
(COCH2CO), 33.1 (SC(CH3)2), 26.8 (SC(CH3)2), m/z (ESI):
[M + Na]+ for [C25H34N2NaO2S2]+ calculated: 481.1959
found: 481.1959.
N1,N3‑bis(2((4‑methoxybenzyl)thio)ethyl)propanediamide
(1f) Yellow solid. Yield: 0.43 g (12%), m.p.: 146–149 °C,,
ʋ(cm−1): 3337 (NH), 1650 (amide C = O bend), 1240
1
(C–O; aromatic ether), H NMR (300 MHz, CDCl3) δ
General synthesis procedure
7.24 (d, J = 8.5 Hz, 4H, ArH), 6.97 (s, 2H, CH2CH2NH),
6.85 (d, J=8.5 Hz, 4H, ArH), 3.79 (s, 6H, CH3OPh), 3.67
(s, 4H, PhCH2S), 3.40 (m, 4H, CH2CH2NH), 3.11 (s, 2H,
COCH2CO), 2.55 (t, J=6.5 Hz, 4H, SCH2CH2), 13C NMR
(75 MHz, CDCl3) δ 167.2 (COCH2CO), 158.6 (ArC),
130.1 (ArC), 130.0 (ArC), 114.2 (ArC), 55.4 (CH3O),
43.1 (COCH2), 38.5 (CH2CH2NH), 35.4 (PhCH2S), 30.8
(SCH2CH2), m/z (ESI): [M+Na]+ for [C23H30N2NaO4S2]+
calculated: 485.1545, found: 485.1539.
for the N1,N3‑bis(2‑(benzylthio)ethyl)
diethylpropanediamide derivatives (3a–e)
To a stirred solution of amine 2a–e (1 eq.) and triethylamine
(1 eq.) in dry diethyl ether was added a dry diethyl ether
solution of diethyl malonyl chloride (0.5 eq.) dropwise for
90 min. The reaction mixture was left to stir overnight after
which H2O was added, and the crude product was extracted
using diethyl ether. The diethyl extract was washed succes-
sively with HCl (2 M), NaHCO3, H2O and brine after which
it was dried over Na2SO4 and concentrated in vacuo. The
crude product was purifed by fash chromatography (20%
EtOAc/Hexane to 100% EtOAc) to give the diethylpropan-
ediamide derivatives 3a–e.
N1,N3‑bis(2‑((4‑fluorobenzyl)thio)ethyl)propanedi‑
amide (1g) Yellow powder. Yield: 0.55 g (31%), m.p.:
121–123 °C,, ʋ(cm−1): 3296 (amide N–H), 1647 (amide
1
C = O bend), H NMR (400 MHz, CDCl3) δ 7.28 (d, J
HCCF =13.5 Hz, 4H, ArH), 7.18 (s, 2H, CH2NHCO), 6.99 (t,
J=8.5 Hz, 4H, ArH), 3.69 (s, 4H, PhCH2S), 3.41 (m, 4H,
CH2CH2NH), 3.15 (s, 2H, COCH2CO), 2.55 (t, J=6.5 Hz,
4H, SCH2CH2), 13C NMR (100 MHz, CDCl3) 167.3
(COCH2CO), 162.1 (JCF =242, FArC), 133.8 (JCCCCF =3 Hz,
ArC), 130.6 and 130.5 (d, JCCCF =8 Hz), 115.7 and 115.5
(d, JCCF = 21 Hz), 43.1 (COCH2CO), 38.5 (CH2CH2NH),
35.2 (FPhCH2S), 30.8 (CH2CH2S), m/z (ESI): [M + Na]+
for [C21H24F2N2NaO2S2]+ calculated: 461.1145 found:
461.1143.
N1,N3‑Bis(2‑(benzylthio)ethyl)‑2,2‑diethylpropanedi‑
amide (3a) Viscous colorless oil. Yield: 1.1 g (54%), ʋ
(cm−1): 3326 (amide N–H stretch), 1630 (amide C = O
1
bend), H NMR (300 MHz, CDCl3) δ 7.42 (s, 2H, CON-
HCH2), 7.38–7.12 (m, 10H, ArH), 3.72 (s, 4H, SCH2Ph),
3.41 (m, 4H, NHCH2CH2), 2.57 (t, J = 6.5 Hz, 4H,
CH2CH2S), 1.85 (q, J = 7.0 Hz, 4H, CCH2CH3), 0.82 (t,
J = 7.0 Hz, 6H, CCH2CH3), 13C NMR (75 MHz, CDCl3)
δ 173.2 (COC(CH2CH3)2CO), 138.2, 129.0, 128.7 and
127.3 (ArC), 58.3 (COC(C2H5)CO), 38.2 (NHCH2CH2S),
35.9 (CH2CH2S), 31.1 (SCH2Ph), 30.3 CCH2CH3), 9.6
(CCH2CH3). m/z (ESI): [M+H]+ for [C29H35N2O2S2]+ cal-
culated: 459.2140 found: 459.2138.
N1,N3‑bis(2‑((4‑nitrobenzyl)thio)ethyl)propanediamide
(1h) Yellow solid. Yield: 1.22 g (34%), m.p.: 82–85 °C,
ʋ(cm−1): 3294 (N–H stretch), 1647 (amide C = O bend),
1544 and 1367 (N–O stretch), 1H NMR (300 MHz, CDCl3)
δ 8.17 (d, J = 8.5 Hz, 4H, ArH), 7.50 (d, J = 8.5 Hz, 4H,
ArH), 7.19 (s, 2H, CH2CH2NH), 3.80 (s, 4H, PhCH2S),
3.43 (m, 4H, CH2CH2NH), 3.16 (s, 2H, COCH2CO), 2.58
(t, J=6.5 Hz, 4H, SCH2CH2).13C NMR (75 MHz, CDCl3)
δ 167.3, (COCH2), 147.2 (ArC), 145.9 (ArC), 129.9 (ArC),
124.0 (ArC), 42.9 (COCH2), 38.4 (CH2CH2NH), 35.4
N1,N3‑Bis(2‑((4‑vinylbenzyl)thio)ethyl)‑2,2‑diethylpropane‑
diamide (3b) Colorless gel. Yield: 0.12 g, (83%), ʋ (cm−1):
1
3323 (amide N–H stretch), 1655 (amide C = O bend), H
NMR (300 MHz, CDCl3) δ 7.36 (d, J = 7.5 Hz, 4H, ArH
and CH2NHCO), 7.27 (d, J = 7.5 Hz, 4H, ArH), 6.69 (dd,
J=17.5, 11 Hz, 2H, CH2CHcPh), 5.73 (d, J=17.5 Hz, 2H,
CHaHbCHPh), 5.23 (d, J = 11.0 Hz, 2H, CHaHbCHPh),
1 3

Chemical Papers
3.71 (s, 4H, CH2CHPhCH2S), 3.43 (m, 4H, CH2CH2NH),
2.57 (t, J=6.5 Hz, 4H, SCH2CH2), 1.85 (q, J=7.5 Hz, 4H,
COC(CH2CH3)2), 0.82 (t, J=7.5 Hz, 6H, COC(CH2CH3)2),
13C NMR (75 MHz, CDCl3) δ 173.2 (COC(CH2CH3)2CO),
137.8 and 136.7 (ArC), 136.5 (CH2CHPh), 129.2 and 126.6
(ArC), 114.0 (CH2CHPh), 58.3 (COC(CH2CH3)2), 38.2
(CONHCH2), 35.6 (CH2CHPhCH2S), 31.1 (C(CH2CH3)2),
30.3 (SCH2CH2), 9.6 (C(CH2CH3)2), m/z (ESI): [M+H]+
for [C29H36N2O2S2]+ calculated: 511.2453 found: 511.2450.
Synthesis of silver complexes of N1,N3‑bis(2‑(benzylthio)
ethyl)propanediamide ([Ag(1a)]OSO2CF3)
The Ag+-1a complexes (4a–c) were accessed following the
method reported by Daubinet (2001). Briefy, 1a (1 eq.)
was dissolved in EtOAc (5 mL) and added to AgOSO2CF3
(1, 2 or 3 eq.) also in EtOAc (5 mL) and the mixture was
reacted under refuxing conditions for 15 min. The product
was concentrated in vacuo and samples suitable for mass
spectrometry analysis were prepared.
N1,N3‑Bis(2‑((4‑methoxybenzyl)thio)ethyl)‑2,2‑diethyl‑
propanediamide (3c) Viscous colorless oil. Yield: 1.27 g
(45%), υ (cm−1): 3337 (NH), 1650 (amide C=O bend), 1240
(C–O; aromatic ether), 1H NMR (300 MHz, CDCl3) δ 7.44
(s, 2H, CH2CH2NH), 7.23 (d, J = 8.5 Hz, 4H, ArH), 6.84
(d, J=8.5 Hz, 4H, ArH), 3.79 (s, 6H, CH3O), 3.68 (s, 4H,
PhCH2S), 3.42 (m, 4H, CH2CH2NH), 2.56 (t, J = 6.5 Hz,
4H, SCH2CH2), 1.86 (q, J = 7.5 Hz, 4H, CCH2CH3),
0.83 (t, J = 7.5 Hz, 6H, CCH2CH3), 13C NMR (75 MHz,
CDCl3) δ 173.2 (NHCOC), 158.8 (ArC), 130.1 (ArC),
130.0 (ArC), 114.1 (ArC), 58.3 (COCCO), 55.4 (OCH3),
38.2 (NHCH2CH2), 35.3 (SCH2Ph), 31.0 (CCH2CH3), 30.2
(CH2CH2S), 9.5 (CCH2CH3), m/z (ESI): [M + Na]+ for
[C27H38N2NaO4S2]+ calculated: 518.2171, found: 541.2165.
Silver N1,N3‑bis(2‑(benzylthio)ethyl)propanediamide tri‑
fate (4a) Brown paste. Yield: 30 mg (91%), m/z (ESI):
[M + Ag]+ calculated for [C21H26AgN2O2S2]+; 509.0487
found; 509.0481.
Silver N1,N3‑bis(2‑(benzylthio)ethyl)propanediamide tri‑
fate (4b) Brown solid. Yield: 30 mg (65%), m/z (ESI):
[M + Ag]+ calculated for [C21H26AgN2O2S2]+; 509.0487
found; 509.0481.
Silver (N1,N3‑bis(2‑(benzylthio)ethyl)propanediamide tri‑
fate (4c) Yellow paste. Yield: 50 mg (85%). m/z (ESI):
[M + Ag]+ calculated for [C21H26AgN2O2S2]+; 509.0487
found; 509.0481.
N1,N3‑Bis(2‑((4‑fuorobenzyl)thio)ethyl)‑2,2‑diethylpropan‑
ediamide (3d) Viscous yellow oil. Yield: 0.91 g (46%),
ʋ(cm−1): 3296 (amide N–H), 1647 (amide C=O bend), 1H
NMR (300 MHz, CDCl3) δ 7.42 (s, 2H, CH2NHCO), 7.29 (t,
J=8.0 Hz, 4H, ArH), 7.00 (t, J=8.0 Hz, 4H, ArH), 3.70 (s,
4H, SCH2Ph), 3.44 (m, 4H, NHCH2CH2), 2.56 (t, J=6.0 Hz,
4H, CH2CH2S), 1.87 (q, J = 7.0 Hz, 4H, CCH2CH3), 0.83
(t, J=7.0 Hz, 6H, CCH2CH3). 13C NMR (75 MHz, CDCl3)
δ 173.2 (COCCO), 162.1 (d, JCF = 242, ArC), 133.9 (d,
JCCCCF = 3 Hz, ArC), 130.5 (d, JCCCF = 8 Hz), 115.6 (d,
JCCF = 21 Hz), 58.3 (COCCO), 38.2 (NHCH2CH2), 35.1
(SCH2ArC), 31.1 (CCH2CH3), 30.3 (CH2CH2S), 9.6
(CCH2CH3). m/z (ESI): [M+Na]+ for [C25H32F2N2NaO2S2]+
calculated; 517.1771 found; 517.1765.
Selective Ag+ extraction studies
The selective Ag+ extraction experiments were carried out
following the method reported by Sole and Hiskey (1992).
Briefy, a solution of equal concentrations of Cu2+, Ag+ and
Pb2+ (4 mg/L each) in 0.023 M Na2SO4 and 0.476 M HNO3
was prepared from a 500 mg/L stock solution represented the
aqueous phase. The organic phase was chloroform solution
of the ligands. The chloroform solvent was presaturated with
twice its volume of 0.5 M H2SO4 by constant shaking in a
separatory funnel for 5 min prior to been used to prepare the
MAD ligand solutions. The concentration of the ligand solu-
tions for each batch extraction was 1.15 M, which is equal
to 250 times that of Ag+ (4.6 mM) used for each experi-
ment. For each metal extraction experiment, equal volumes
(10 mL) of the metal solution and the ligand solution were
contacted by rapid and vigorous stirring (using a stir bar)
for 15 min in a capped plastic vial partially immersed in
an oil bath set at 25 °C. Each experiment was undertaken
in duplicates. After stirring, the immiscible solutions were
transferred to a separatory funnel, allowed to separate and
collected separately. The aqueous phase was collected into
a beaker and the residual chloroform was removed over a
steam bath in about 15 min. Then the aqueous phase was
made back up to 10 mL and prepared for ICP-OES analysis.
The metal extraction efciencies were determined using the
equation:
N1,N3‑2,2‑Diethyl‑bis(2‑((4‑nitrobenzyl)thio)ethyl)propan‑
ediamide (3e) Viscous yellow oil. Yield: 2.84 g (76%),
ʋ(cm−1): 3294 (N–H stretch), 1647 (amide C=O bend),
1544 and 1367 (N–O stretch), 1H NMR (300 MHz, CDCl3) δ
8.18 (d, J=8.5 Hz, 4H, ArH), 7.51 (d, J=8.5 Hz, 4H, ArH),
7.45 (s, 2H, CH2CH2NH), 3.81 (s, 4H, PhCH2S), 3.46 (m,
4H, SCH2CH2), 2.57 (t, J=6.5 Hz, 4H, SCH2CH2), 1.87 (q,
J=7.5 Hz, 4H, CCH2CH3), 0.83 (t, J=7.5 Hz, 6H, CCH2CH3),
13C NMR (75 MHz, CDCl3) δ 173.3 (NHCOC), 147.2 (ArC),
146.0 (ArC), 130.0 (ArC), 124.0 (ArC), 58.3 (COCCO),
38.2 (CH2CH2NH), 35.3 (PhCH2S), 31.3 (CCH2CH3), 30.5
(SCH2CH2), 9.6 (CCH2CH3), m/z (ESI): [M + Na]+ for
[C25H32N4NaO6S2]+ calculated: 571.1661, found 571.1605.
1 3

Chemical Papers
1
For the H NMR titration study, the ligand 1a (13 mg,
Ci − Cf
%EE =
× 100,
0.03 mmol) was dissolved in 500 µL DMSO-d6 and trans-
ferred into an amber-colored NMR tube. Following this,
standard solution of AgClO4 (539 mM) was made up in
DMSO-d6 and constant volume (30 µL) of the AgClO4
solution was added to 1a in the NMR tube to target vary-
ing Ag+/1a mole ratios (nAg+/n1a = 0.5 to 4.0). For exam-
ple, to target a nAg+/n1a =0.5, a 30 µL, 539 mM solution of
the AgClO4 solution was added using a glass syringe to the
neat 1a solution and the 1H NMR spectrum was recorded.
Also, to target a nAg+/n1a =1.0, a 30 µL, 539 mM solution
AgClO4 solution was added to the AgClO4-1a solution with
nAg+/n1a =0.5 and the 1H NMR spectrum was recorded. The
constant addition of the AgClO4 solution was continued
until a nAg+/n1a =4.0 was achieved. The plot of Δδ (where
Δδ=chemical shift of Ag+-1a complex minus chemical shift
of free 1a) against nAg+/n1a was constructed using the data
generated.
Ci
where % EE is the percentage extraction efficiency, Ci
and Cf are the initial and fnal metal ion concentrations,
respectively.
The selectivity of the ligands for Ag+ relative to Cu2+ and
Pb2+ was determined as follows (Toure et al. 2018);
+
DAg
+
KAg ∕M
n+
=
,
n+
DM
+
+
n +
where KAg ∕M represents the selectivity coefcient of Ag
relative to Mn+ (Mn+ =Cu2+ or Pb2+), DMn+ is the distribu-
tion coefcient of the metal ions between the aqueous and
the organic phases defned as;
Ci − Cf
Vaq
DMn+
=
×
,
Cf
Vorg
Results and discussion
Ci and Cf are as defned above, Vaq and Vorg are the vol-
Syntheses of the 2‑(benzylthio)
umes of the aqueous and the organic solutions, respectively.
ethan‑1‑amine derivatives (2a–e)
and 2,2‑dimethyl‑2‑benzylthioethylamine (2g)
Job plot and 1H NMR titration studies of the interaction
of Ag+and N1,N3‑bis(2‑(benzylthio)ethyl)propanediamide
(1a) (Fielding 2000)
The S-benzylcysteamines 2a–e, which are starting materials
for the synthesis of MADs 1a, 1e–h were accessed following
a modifed protocol to that reported by Ghosh and Tochtrop
(2009). For the amine 2a, it was necessary to shorten the
reaction time from 40 to 20 min to minimize the formation
of the dibenzylated product 2h (Scheme 1). However, for the
other amines 2b–e it was necessary to extend the reaction
time to 40 min (Scheme 2) to give good selectivity as less
than 3% dibenzylated products were observed as evidenced
from their 1H NMR spectra. Using these reaction durations
(20 min for 2a and 40 min for 2b–e), the desired compounds
apart from 2d were prepared without the need for further
purifcation as evidenced by their 1H NMR spectra. These
amines served as starting material for the synthesis of 1a
and 1e–h.
Standard solutions of equal concentrations (5 mM) of ligand
1a and AgClO4 were prepared separately in DMSO-d6.
Varying ratios of each solution (1a and AgClO4) were then
taken and mixed (with the overall concentration remaining
the same) to target diferent Ag+ mole fractions (χAg+). For
+
example, to target a χAg =0.1, a 450 µL, 5 mM 1a solution
was mixed with a 50 µL, 5 mM AgClO4 solution inside an
+
amber-colored NMR tube. Also, to target a χAg = 0.2, a
400 µL, 5 mM 1a solution was mixed with a 100 µL, 5 mM
+
AgClO4 solution and so on until a χAg =0.9 was prepared
by mixing a 50 µL, 5 mM 1a solution with a 450 µL, 5 mM
AgClO4 solution. The 1H NMR experiments of all mixtures
and the free ligand 1a were recorded. The plot of χ1a·Δδ
(where χ1a =mole fraction of 1a and Δδ=chemical shift of
Ag+-1a complex minus chemical shift of free 1a) against
χAg+ was constructed from the data generated.
To assess the efect of sterics on Ag+ efciency and selec-
tivity by the novel MAD ligands, the substituted S-benzylcy-
steamine 2g was synthesized by a modifcation of a method
reported by Carroll, White, and Wall (1963) as schematized in
Scheme 3. The poor yield (23%) obtained for 2g was attributed
Scheme 1 Illustration of the
synthesis of amine 2a showing
the dibenzylated byproduct 2h
1 3

Chemical Papers
ligands 1e-i were accessed (in low to high yields) using the
same protocol (Scheme 4).
The gem-diethyl MADs 3a-e were obtained in much bet-
ter yields (46–75%) than those obtained for the unsubsti-
tuted MADs (1a and 1e-i), presumably due to removal of
competing enolization pathways that would lower the yields
(Scheme 5).
Scheme 2 Syntheses of the 2-(benzylthio)ethan-1-amine derivatives
(2a–e)
Ag+ extraction studies
Extraction studies in chloroform, following the method
reported by Sole and Hiskey (1992), were carried out to
assess the selectivity and efciency of Ag+ extraction by the
MAD ligands 1a, 1e–h and 3a prepared. Prior to its use in
preparation of the ligand solutions, the chloroform solvent
was presaturated with twice its volume of an acidifed deion-
ized H2O to remove any H2O-soluble components to mini-
mize volume changes during the solvent metal extraction
process. We decided to prepare mixed metal aqueous solu-
tions with very low concentrations (4 mg/L) of Ag+, Cu2+
and Pb2+ because Ag+ typically exists in very low concen-
trations in Ag+ repositories where our novel MAD ligands
may be applied (Crane et al. 2017). The choice of Cu2+ and
Pb2+ as competing ions for the efciency and selectivity
to the loss of acetone during the reaction. This amine 2g will
be employed for the synthesis of a novel MAD 1i (Fig. 2) with
steric hinderance at the α-position of the sulfur atom.
Syntheses of the N1,N3‑bis(2‑(benzylthio)ethyl)
propanediamide derivatives (1a, 1e‑i and 3a‑e)
In the original published procedure by Daubinet and Kaye
(2002), access to the ligand 1a required a reaction duration
of 7 days, at the end of which a yield of 24% was obtained.
We prepared 1a in less than 1 day with an improved yield
of 33% through an alternative pathway involving the drop-
wise addition of one equivalent of malonyl chloride to four
equivalents of the amine 2a in dry THF (Scheme 4). Related
Scheme 3 Synthesis of
2,2-dimethyl-2-benzylthioethyl-
amine (2g)
Scheme 4 Syntheses of the
malondiamide derivatives (1a,
1e–i)
Scheme 5 Syntheses of the
gem-diethyl malondiamide
derivatives (3a–e)
1 3

Chemical Papers
extractions is borne out of the knowledge that these met-
als usually coexist with Ag+ in ores and mine tailings, for
instance (Crane et al. 2017). For sensible comparisons to be
made, each solvent extraction experiment was undertaken
in duplicate and with vigorous stirring (using a stir bar with
equal volumes of ligand and metal solutions for 15 min). The
concentration of the metals in the rafnate was measured by
means of ICP-OES and each metal concentration value was
the average of three analytic runs.
100
80
60
40
20
0
Cu(II)
Ag(I)
Pb(II)
A control experiment was undertaken with the metal solu-
tion using the chloroform solvent without any ligand present.
This would determine any extraction efciency of the solvent
itself (the background extraction) and any possible extraction
due to the plastic vials used for the experiment. Interestingly,
we observed that the neat chloroform itself (the control) was
able to extract some metals with a slight preference for Ag+
(Fig. 3). This is likely due to either the diferential partial
solubility of these metals in chloroform or Ag+ adsorbed
onto the inner walls of the plastic vial. Notwithstanding,
the extraction efciency of the 1a for Ag+ was signifcantly
higher than that in the control experiment (Fig. 3), underlin-
ing the importance of 1a in efcient Ag+ extraction.
1a
1e
1f
1g
1h
Extractant
Fig. 4 ICP-OES data showing metal extraction efciencies by the
malondiamide-derived ligands 1a, 1e-h ([Mn+]=4 mg/L each,
volume of organic and aqueous phases=10 mL each, contact
time=15 min)
The highest extraction efciency observed for 1e could
be because it would give the least polar Ag+ complex and
thus have the highest solubility in chloroform (Ocak et al.
2006). Correspondingly, the least extraction efciencies
were observed for 1g (X=F) and 1h (X=NO2) because (a)
the Ag+ complexes are the most polar and so less soluble
in chloroform (Ocak et al. 2006) and (b) the electron with-
drawing nature of the aryl groups may have made the sulfur
donor atom lone pairs less available for binding Ag+ (the
soft sulfur atom being crucial in efciency and selectivity
following the HSAB rule) (Pearson 1968). For the methoxy
analogue 1f, a combination of the polar nature of the Ag+-1f
complex lowering extraction efciency with the electron
donation activity of the methoxy group making the sulfur
atom lone pairs readily available and increasing the bind-
ing to Ag+ leads to an intermediate extraction efcacy. The
latter explanation (b), if true, would also explain the poorer
selectivity’s of the ligands with electron withdrawing groups
(1e, 1g, 1h), noting that in the styryl analogue 1e the induc-
tive electron withdrawing nature of the vinyl group will be
the dominant efect. A greater selectivity would be predicted
and is observed for the strongly electron donating methoxy
substituent in 1f, where selectivity for Ag+ versus Cu2+ is
approximately nine times better than was observed by 1a.
Electronics efect
The efect of electronic diferences in the aryl grouping of
the MAD ligands on the efciency and selectivity of Ag+
recovery was studied by comparing extraction efciencies
and selectivity’s exhibited by the ligands 1a and 1e–h. The
observed extraction efciencies are presented in Fig. 4.
While the efciencies for Ag+ extraction by these ligands
difer only slightly; 1a=(96.3 1.2)%, 1e=(98.4 0.4)%,
1f = (94.1 1.6)%, 1g = (91.1 1.0)%, 1h = (91.1 2.7)%
(Fig. 4) extraction efficiencies for those with electron
withdrawing substituents 1g and 1h are the lowest (extrac-
tion efficiency order: 1h = 1g < 1f < 1a < 1e) (Table 1).
Cu(II)
Ag(I)
Pb(II)
100
80
60
40
20
0
Table 1 Efciency and selectivity of malondiamide-derived ligands
1a, 1e-h for Ag+ relative to Cu2+ and Pb2+
+∕Cu2+
+∕Pb2+
Ligand
Efciency (%)
KAg
KAg
Control
1a
Extractant
1a
1e
1f
1g
1h
96.3 1.2
98.4 0.4
94.1 1.6
91.1 1.0
91.1 2.7
11.6
5.2
85.5
5.2
321
5.2
235.3
5.2
5.4
Fig. 3 ICP-OES data showing metal extraction efciencies in the
control and the malondiamide-derived ligand 1a ([Mn+]=4 mg/L
each, volume of organic and aqueous phases=10 mL each, contact
time=15 min)
5.0
1 3

Chemical Papers
The results observed here suggest that electronics have only
a small infuence on the efciency of Ag+ extraction by these
ligands but a signifcant efect on selectivity. The implication
of this result for large scale Ag+ recovery is that purer Ag+
may be recovered from a mixed metal solution by a ligand
bearing electron donating groups like the methoxy group.
than that of 1a in chloroform due to the gem dialkyl group-
ings (Ocak et al. 2006). Overall, steric hindrance imparted
by dimethyl substitution at α-positions to the crucial sulfur
donor reduced efciency of Ag+ extracted compared to
a ligand without such hindrance. The ligand with steric
hindrance occasioned by diethyl substitution at the acyl
position (not in the neighborhood of the sulfur donor), was
observed to demonstrate higher efciency for Ag+ than a
ligand without such steric hindrance. Selectivity for Ag+
against Pb2+ reduced in ligands with steric hindrances
aforded by either dimethyl or diethyl substitutions when
compared to the ligand without any steric hindrance.
Steric efect
The efect of steric hindrance at various sites in the MAD
ligands was investigated. Steric hindrance around the cru-
cial sulfur atom required for binding in ligand 1i might be
predicted to lead to lower efciencies in Ag+ extraction
while placing steric hindrance further away at the central
carbon represented by ligand 3a, might be predicted to
have less efect on the extraction of Ag+ (Fig. 5). Ligand
1i, sterically hindered near the sulfur atom, extracted with
lesser efciency (~ 4%) for Ag+ (92.2 0.2)% than the
analogue without such steric hindrance 1a (96.3 1.2)%,
(Fig. 5) as predicted. This is attributed to the slightly
greater difculty of 1i in assuming the right conforma-
tion needed to bind Ag+ because of the presence of the
dimethyl substitution at the α-position to the crucial sul-
fur donor center (Hiruta et al. 2014). On the other hand,
MAD 3a, sterically hindered at the acyl region extracted
slightly higher amounts of Ag+ compared to 1a attributed
to the higher solubility of the Ag+-3a complex in chloro-
form than the Ag+-1a complex [1a = (96.3 1.2)% versus
3a = (98.9 0.2)%]. It is worth noting that the ligand 3a
extracted slightly higher amounts of Ag+ than the ligand
1i, attributed to the location of steric hindrance away from
the crucial sulfur donor center as earlier explained. Inter-
estingly both 1i and 3a were less selective than 1a for
Investigation of Ag+‑1a binding stoichiometry
Mass spectrometry studies To gain insight into the nature
of the binding of Ag+ by the MAD ligands, the MAD 1a
was chosen as a case study. This is because 1a contains only
the basic motif reported to be responsible for Ag+ binding
(Daubinet 2001). Therefore, Ag+ complexes of 1a, were pre-
pared by varying the equivalents of the silver salt (1–3 eq.)
relative to the ligand 1a (Table 2). Consequently, the Ag+-1a
complexes (4a–c) were obtained in moderate to high yields
(Table 2) and characterized by means of low- and high-res-
olution mass spectrometry. In the wide scan (m/z 0–2400)
low resolution mass spectrometry (LRMS) of 4a, the only
pair of peaks with ratio of~1:1 (indicative of a Ag+ com-
plex) were observed at m/z 509 and 511 (Fig. 6). Indeed, the
m/z 509 was confrmed by high resolution mass spectrome-
try (HRMS) analyses as the complex—[107Ag(1a)]+ (found:
509.0481, calculated: 509.0487). Interestingly, reacting the
ligand 1a with two equivalents and three equivalents of sil-
ver trifate gave 4b and 4c, respectively (Table 2, entries 2
and 3) which were also both found, after LRMS analyses,
to be Ag+-1a complexes with a 1:1 stoichiometry. These
results indicate that the binding stoichiometry of the Ag+
complex with MAD 1a is perhaps 1:1.
+
2+
+
2+
binding Ag+ over Pb2+ (1i KAg
= 4.5, 3a KAg
=
/Pb
+
2+
11.1, 1a KAg
= 321) and/Ptbhis could be attributed to
/Pb
the higher solubility of the Pb2+ complexes of 1i and 3a
100
Cu(II)
Ag(I)
Pb(II)
80
60
40
20
0
Table 2 Efect of Ag+ equivalent on stoichiometry of AgOSO2CF3-1a
AgOSO2CF3
1 − 3 eq.
⟶
complexes ꢀꢁ
[Agx(ꢀꢁ)y]n+
EtOAC
Reflux, 15 min
Ag+:1a
Yield (%)
1a
1i
3a
Entry
Ligand
AgO-
Product
SO2CF3 eq
Extractant
1
2
3
1a
1a
1a
1
2
3
4a
4b
4c
1:1
1:1
1:1
91
65
85
Fig. 5 ICP-OES data showing metal extraction efciencies by malon-
diamide-derived ligands 1a, 1i and 3a ([Mn+]=4 mg/L each, volume
of organic and aqueous phases=10 mL each, contact time=15 min)
1 3

Chemical Papers
Fig. 6 Partial low-resolution ESI mass spectra of Ag+-1a complexes from treatment of 1a with (a) one (b) two and (c) three equivalents of Ag+
Job plot and 1H NMR titration studies
uncomplexed states. The gentle curvature of the curves in the
Job plot (Fig. 7b) indicates the binding between the ligand
1a and Ag+ is perhaps weak (Swiegers and Malefetse 2000;
Renny et al. 2013). Furthermore, it can be observed that the
coefcient on the x-axis of the maxima of all curves in the
The Job plot study for the Ag+-1a complex was carried out
using equal concentrations (5 mM) of ligand 1a and AgClO4
in DMSO-d6 after initial solvent screening experiments
revealed the unsuitability of other common NMR solvents
tested (due to poor solubility of the ligand 1a). The Job plot
experiment did not lead to the appearance of new peaks
(Fig. 7a) but caused distinct changes in chemical shifts, indi-
cating a fast rate of Ag+ exchange between the complexed and
+
Job plot is at χAg = 0.5, indicating that 1a binds Ag+ in any
of the 1:1, 2:2 or any other n:n fashion. To confrm the actual
stoichiometry, 1H NMR titration study of the interaction of 1a
with Ag+ was carried out. Therefore, a constant amount of a
539 mM Ag+ solution was successively added to the ligand
1 3

Chemical Papers
(a)
(b)
0.010
a,b,c
d
e
f
g
h
0.008
0.006
0.004
0.002
0.000
0.0
0.2
0.4
0.6
0.8
1.0
Ag+
(c)
(d)
0.25
0.20
0.15
0.10
0.05
0.00
a,b,c
d
e
g
h
0
1
2
3
4
nAg(I)/n(2.10)
1
Fig. 7 (a, c) H NMR (300 MHz) spectra of the interaction of Ag+
solvent=DMSO-d6. For 1H NMR titration, [1a]=60 mM,
1
with 1a. (b) Job plot and (d) H NMR titration plot of the interac-
[Ag+]=539 mM, solvent=DMSO-d6)
tion of Ag+ with ligand 1a (For job plot, [1a]=[Ag+]=5 mM,
1a solution to prepare Ag+/1a mole ratios ranging from 0.5
to 4.0. Proton NMR spectra after successive addition of the
Ag+ solution was recorded (Fig. 7c), and the 1H NMR titration
plot was constructed from the values obtained (Fig. 7d). In a
typical 1H NMR titration plot, the stoichiometry of a complex
is the coefcient on the x-axis of the point of infection. In the
case of the Ag+-1a complex (Fig. 7d), it was very difcult to
get a distinct point of infection for any of the curves, indicat-
ing a weak binding between the Ag+ and 1a and confrm-
ing the observation from the Job plot (Renny et al. 2013). It
also suggests that the Ag+-1a complex may be adopting more
than one type of stoichiometry in solution including the 1:1
observed from mass spectrometry and the Job plot.
In the Job plot (Fig. 7b), the highest Δδ values were
observed for protons ‘e’ and ‘d’ in 1a (the protons on
α-positions to the sulfur donor atoms). The next highest
Δδ values were observed for protons ‘g’ in 1a (the amide
protons). The observed high Δδs may be due to deshield-
ing efects experienced by these protons as the sulfur and
nitrogen donors participate in binding Ag+. Unsurpris-
ingly, lower Δδ values were observed for phenyl protons
(‘a’, ‘b’ and ‘c’ in 1a) since they are not in the neighbour-
hood of the sulfur and nitrogen donors (Fig. 7b). Based on
these observations, it was hypothesized that the ligand 1a
binds Ag+ using its two sulfur and nitrogen donor centers
to form a tetrahedral complex (Fig. 8).
1 3

Chemical Papers
Fig. 9 Idealized structure of a highly efcient and selective N1,N3-
bis(2-(benzylthio)ethyl)malondiamide derivative suitable for attach-
ment by polymerization to a magnetic nanoparticle
Fig. 8 Proposed optimized structure for [Ag(1a)]+ based on stoichi-
ometry results observed from mass spectrometry, Job plot and 1H
NMR titration results studies
the 4-positions of diferent aromatic groups and gem-diethyl
substitution at the methylene center as shown in Fig. 9.
Author contributions Conceptualization, experimental design, exper-
imental work, frst draft (ADA). Conceptualization, experimental
design, paper review (AJC).
Conclusion
Herein, a range of novel N1,N3-bis(2-(benzylthio)ethyl)pro-
panediamide derivatives including the divinyl derivative (1e)
which could be attached by polymerization to a magnetic
nanoparticle have been synthesized under mild conditions
and in low to high yields. The efect of electronics (using 1a
and 1e—h) and sterics (using 1a, 1i and 3a) on efciency
and selectivity for Ag+ extraction was investigated using
extraction efciencies measured following solvent extrac-
tion experiments. It was observed that electronic efects at
the 4-position of the aromatic groups in the malondiamide
derivatives (1e–h) had little efect on Ag+ extraction ef-
ciency but a great efect on metal selectivity, with the elec-
tron rich (4-methoxy)malondiamide derivative 1f being the
most selective and the (4-vinyl) derivative 1e showing the
best efciency for Ag+ over Cu2+ and Pb2+. Increased steric
hindrance at the α-positions to the sulfur donors conferred
by dimethyl substitutions lowered efciency and selectiv-
ity for Ag+ when compared to the derivative 1a lacking
such hindrance. The derivative 3a with steric hindrance at
the methylene center provided by gem-diethyl substitution
showed better efciency but lower selectivity for Ag+ than
1a devoid of such steric hindrance. Binding studies carried
out by means of mass spectrometry, Job plot and 1H NMR
titration reveal that the derivative 1a binds Ag+ in a 1:1 fash-
ion leading to the inference that 1a forms a tetrahedral com-
plex with Ag+. Binding studies suggest that 1e, our target
for attachment by polymerization to a nanoparticle may also
bind Ag+ in the same way as 1a since both ligands contain
essentially the same donor atoms crucial for Ag+ binding.
In which case it is presumed that the polymerized version
of 1e—poly(1e), may lose its Ag+ binding efciency poten-
tially due to the difculty in wrapping round Ag+ especially
in the case where poly(1e) crosslinks between nanoparticles.
In addition, since the dimethoxy derivative 1f and the gem-
diethyl derivative 3a showed the best selectivity and ef-
ciency respectively for Ag+, an ideal malondiamide-derived
ligand for attachment to a magnetic nanoparticle would be
one containing both the vinyl and methoxy groups on each of
Funding The authors would like to thank The Commonwealth Scholar-
ship Commission in the United Kingdom and The University of War-
wick for funding the research under grant number NGCS-2015-448.
Data availability Not applicable.
Code availability Not applicable.
Compliance with ethical standards
Conflict of interest The authors declare no confict of interest.
References
Aderibigbe AD, Crane RA, Lees MR, Clark AJ (2020) Selective uptake
of Ag (I) from aqueous solutions using ionic liquid-modifed iron
oxide nanoparticles. J Nanoparticle Res 216:1–14. https://doi.
org/10.1007/s11051-020-04944-1
Alam MS, Inoue K, Yoshizuka K, Dong Y, Zhang P (1997) Solvent
extraction of silver from chloride media with some commercial
sulfur-containing extractants. Hydrometallurgy 44:245–254. https
://doi.org/10.1016/S0304-386X(96)00053-9
Avarmaa K, Klemettinen L, O’Brien H, Taskinen P (2019) Urban min-
ing of precious metals via oxidizing copper smelting. Miner Eng
133:95–102. https://doi.org/10.1016/j.mineng.2019.01.006
Bray DJ, Clegg JK, Wenzel M, Gloe K, McMurtrie JC, Jollife KA,
Gloe K, Lindoy LF (2015) Selective solvent extraction of silver(I)
by tris-pyridyl tripodal ligands and X-ray structure of a silver(I)
coordination polymer incorporating one such ligand. Aust J Chem
68:549–554. https://doi.org/10.1071/CH14540
Carroll FI, White JD, Wall ME (1963) Organic sulfur compounds.
I. Synthesis of sec-Mercaptoalkylamine Hydrochlorides1a, b. J
Org Chem 28:1236–1239. https://doi.org/10.1021/jo01040a017
Chen C, Wang J (2008) Removal of Pb(II), Ag(I), Cs(I) and Sr(II) from
aqueous solution by brewery’s waste biomass. J Hazard Mater
151:65–70. https://doi.org/10.1016/j.jhazmat.2007.05.046
Crane RA, Sinnett DE, Cleall PJ, Sapsford DJ (2017) Physicochemi-
cal composition of wastes and co-located environmental designa-
tions at legacy mine sites in the south west of England and Wales:
Implications for their resource potential. Resour Conserv Recycl
123:117–134. https://doi.org/10.1016/j.resconrec.2016.08.009
1 3

Chemical Papers
(DMDCMA) as an extractant for actinides. Sep Sci Technol
49:2927–2932. https://doi.org/10.1080/01496395.2014.943772
Pearson RG (1968) Hard and soft acids and bases, HSAB, part I:
Fundamental principles. J Chem Educ 45:581–587. https://doi.
org/10.1021/ed045p581
Daubinet A (2001) Design, synthesis and evaluation of silver-specifc
ligands. Rhodes University, Grahamstown
Daubinet A, Kaye PT (2002) Designer ligands. VIII. Thermal and
microwave-assisted synthesis of silver(I)-selective ligands. Synth
Commun 32:3207–3217. https://doi.org/10.1081/SCC-120013745
El-Shafey ESI, Al-Kindy SMZ (2013) Removal of Cu(II) and Ag(I)
from aqueous solution on a chemically-carbonized sorbent from
date palm leafets. Environ Technol (UK) 34:395–406. https://doi.
org/10.1080/09593330.2012.698647
Renny JS, Tomasevich LL, Tallmadge EH, Collum DB (2013) Method
of continuous variations: applications of job plots to the study
of molecular associations in organometallic chemistry. Angew
Chem-Int Ed 52:11998–12013. https://doi.org/10.1002/anie.20130
4157
Fielding L (2000) Determination of association constants (Ka) from
solution NMR data. Tetrahedron 56:6151–6170. https://doi.
org/10.1016/S0040-4020(00)00492-0
Saha R, Guha S, Sahana A, Das D (2013) N, N-bis-(2-benzimidazolyl-
methyl)L-methionine: an efcient Ag(I) extractant. J Anal Chem
68:398–402. https://doi.org/10.1134/S1061934813050146
Sahan M, Kucuker MA, Demirel B, Kuchta K, Hursthouse A (2019)
Determination of metal content of waste mobile phones and esti-
mation of their recovery potential in Turkey. Int J Environ Res
Public Health 16:887. https://doi.org/10.3390/ijerph16050887
Selim KA, El HFI, Khalek MAA, Osama I (2017) Kinetics and thermo-
dynamics of some heavy metals removal from industrial efuents
through electro-fotation process. Colloid Surf Sci 2:47–53. https
://doi.org/10.11648/j.css.20170202.11
Ghosh S, Tochtrop GP (2009) A new strategy for the synthesis of
β-benzylmercaptoethylamine derivatives. Tetrahedron Lett
50:1723–1726. https://doi.org/10.1016/j.tetlet.2009.01.133
Grigorieva NA, Fleitlikh IY, Logutenko OA (2018) Silver extraction
from hydrochloric acid solutions with the disulfde of bis(2,4,4-
trimethylpentyl)dithiophosphinic acid. Solvent Extr Ion Exch
36:162–174. https://doi.org/10.1080/07366299.2018.1446675
Hiruta Y, Watanabe T, Nakamura E, Iwasawa N, Sato H, Hamada K,
Citterio D, Suzuki K (2014) Steric hindrance efects in tripodal
ligands for extraction and back-extraction of Ag(I). RSC Adv
4:9791–9798. https://doi.org/10.1039/c3ra45700a
Shimojo K, Goto M (2004) Solvent extraction and stripping of silver
ions in room-temperature ionic liquids containing calixarenes.
Anal Chem 76:5039–5044. https://doi.org/10.1021/ac049549x
Sole KC, Hiskey JB (1992) Solvent extraction characteristics of thio-
substituted organophosphinic acid extractants. Hydrometallurgy
30:345–365. https://doi.org/10.1016/0304-386X(92)90093-F
Sun PP, Min BJ, Kim STA, Cho SY (2017) Separation of silver (I)
and zinc(II) from nitrate solutions by solvent extraction with
LIX63. Mater Trans 58:287–290. https://doi.org/10.2320/mater
trans.M2016316
Jańczewski D, Reinhoudt DN, Verboom W, Malinowska E, Pietrzak
M, Hill C, Allignol C (2007) Tripodal (N-alkylated) CMP(O)
and malonamide ligands: synthesis, extraction of metal ions, and
potentiometric studies. New J Chem 31:109–120. https://doi.
org/10.1039/b613254e
Kolpakova N, Sabitova Z, Sachkov V, Medvedev R, Nefedov R, Orlov
V (2019) Determination of Au(III) and Ag(I) in carbonaceous
shales and pyrites by stripping voltammetry. Minerals 9:78. https
://doi.org/10.3390/min9020078
Swiegers GF, Malefetse TJ (2000) New self-assembled structural
motifs in coordination chemistry. Chem Rev 100:3483–3537.
https://doi.org/10.1021/cr990110s
Manchanda VK, Pathak PN (2004) Amides and diamides as promising
extractants in the back end of the nuclear fuel cycle: an overview.
Sep Purif Technol 35:85–103. https://doi.org/10.1016/j.seppu
r.2003.09.005
Taillades G, Sarradin J (2004) Silver: High performance anode for thin
flm lithium ion batteries. J Power Sources 125:199–205. https://
doi.org/10.1016/j.jpowsour.2003.07.004
Nishihama S, Hirai T, Komasawa I (2001) Review of advanced
liquid-liquid extraction systems for the separation of metal
ions by a combination of conversion of the metal species with
chemical reaction. Ind Eng Chem Res 40:3085–3091. https://doi.
org/10.1021/ie010022
Toure M, Arrachart G, Duhamet J, Pellet-Rostaing S (2018) Tantalum
and niobium selective extraction by alkyl-acetophenone. Metals
(Basel) 8:1–11. https://doi.org/10.3390/met8090654
Virolainen S, Tyster M, Haapalainen M, Sainio T (2015) Ion exchange
recovery of silver from concentrated base metal-chloride solu-
tions. Hydrometallurgy 152:100–106. https://doi.org/10.1016/j.
hydromet.2014.12.011
Ocak Ü, Alp H, Gökçe P, Ocak M (2006) The synthesis of new N2
S2-Macrocyclic schif base ligands and investigation of their
ion extraction capability from aqueous media. Sep Sci Technol
41:391–401. https://doi.org/10.1080/01496390500496942
Ohto K, Yamaga H, Murakami E, Inoue K (1997) Specifc extrac-
tion behavior of amide derivative of calix[4]arene for silver (I)
and gold (III) ions from highly acidic chloride media. Talanta
44:1123–1130. https://doi.org/10.1016/S0039-9140(97)00019-2
Ölmez T (2009) The optimization of Cr(VI) reduction and removal
by electrocoagulation using response surface methodology. J
Hazard Mater 162:1371–1378. https://doi.org/10.1016/j.jhazm
at.2008.06.017
Zanain M, Lovitt R (2013) Removal of silver from wastewater using
cross fow microfltration. E3S Web Conf 1:25005. https://doi.
org/10.1051/e3sconf/20130125005
Publisher’s Note Springer Nature remains neutral with regard to
jurisdictional claims in published maps and institutional afliations.
Patil AB, Pathak PN, Shinde VS, Mohapatra PK (2014) Synthesis and
evaluation of N, N′-dimethyl-N, N′-dicyclohexyl-malonamide
Afliations
Abiodun D. Aderibigbe1,2 · Andrew J. Clark1
1
2
Department of Chemistry, University of Warwick,
Present Address: Department of Chemistry, The Federal
Coventry CV4 7AL, UK
University of Technology Akure, P.M.B. 704, Akure,
Ondo State, Nigeria
1 3