Professional Documents
Culture Documents
The Extraction of Hexamethylenediamine From Aqueous Solution by PH Control and Salt Addition-Full
The Extraction of Hexamethylenediamine From Aqueous Solution by PH Control and Salt Addition-Full
Keywords: This paper details an investigation into the separation of Hexamethylenediamine (HMD) from dilute aqueous
Phase separation solutions by exploiting phase separation. HMD is a monomer used in the production of Nylon-6,6 and as
Amine such this separation is potentially useful in extracting purifying HMD produced from biological sources or
Salt
reclaimed from waste made by the production of nylon. An investigation was undertaken into the effect of
Hexamethylenediamine
using sodium hydroxide (NaOH) to promote the phase separation, and a model was derived to predict this
Nylon
behaviour. Furthermore, the effect of various salts on this phase separation was investigated, and it was found
that sodium sulphate and sodium carbonate promote the phase separation, but are unable to cause phase
separation without the presence of sodium hydroxide.
1. Introduction feedstocks, they are not necessarily applicable to HMD processing from
biological feedstocks.
Hexamethylenediamine, hereafter referred to as HMD, is an un- To this end, an investigation was made into methods for extracting
branched 6 carbon chain, terminated at each end by an amine group. HMD from aqueous solutions at concentrations below 50 g L−1 . Initially
This makes it an important monomer for the production of nylons, chromatographic methods of separation were used, however a chance
particularly nylon-6,6, in which it is reacted with adipic acid. Due to discovery showed that HMD can spontaneously form a secondary phase
the presence of the amine groups, it is highly soluble in water. under some conditions, when the pH was raised high enough. Since this
Historically HMD has been produced from petrochemical sources, secondary phase contained HMD at a concentration above 500 g L−1 , it
including acrylonitrile, butadiene, and cyclohexane [1]. As an alterna- was decided that this was a valid alternative method for the extraction
tive to this, there have been efforts to produce HMD from biological of HMD in its own right.
processes or feedstocks [2–6]. A drawback of these biological methods In this paper, the phase separation behaviour of HMD in aqueous
of HMD production is that the concentration is low [7], necessitating solutions is investigated as a new and separate principle on which to
the separation and purification of dilute HMD solutions. However, some achieve purification. Firstly the effect of only adding sodium hydroxide,
progress has been made towards the biological production of higher hereafter referred to as NaOH, is quantified, and a model for the phase
concentrations of similar molecules [8]. compositions is developed to allow for the design of separation process
Established purification methods for HMD are commonly based equipment. Then, the effect of various salts on the phase separation
on thermal processes such as distillation or crystallisation, in which was looked into. This was chosen as the NaOH addition was observed to
the secondary component may be water [9], ammonia [10], or a have similarities with protein extraction by salting-out [17]. Salting-out
mixture of the two [11,12]. More recent advances in HMD purifica- has also been used to assist liquid–liquid extraction of amines within
tion have tended to focus on removing specific impurities which are the food [18,19] and dye [20,21] industries.
usually byproducts of HMD production, such as imines [13,14] or In this work, contaminants to the HMD were not included, as would
aminocapronitrile [15,16]. Whilst these methods are well tailored to be present in any fermentation broth. This was chosen for two reasons.
the problems of HMD separation and purification from conventional The first is simplification, to reduce the number of species present in
∗ Corresponding author.
E-mail address: mjs44@cam.ac.uk (M.J. Sargent).
URL: https://www.ceb.cam.ac.uk/directory/michael-sargent (M.J. Sargent).
https://doi.org/10.1016/j.cej.2021.129428
Received 18 August 2020; Received in revised form 11 March 2021; Accepted 15 March 2021
Available online 25 March 2021
1385-8947/© 2021 Elsevier B.V. All rights reserved.
M.J. Sargent et al. Chemical Engineering Journal 419 (2021) 129428
2.1. Solution preparation The stock solutions described in Section 2.1 allowed for samples to
be made at a wide range of concentrations for both HMD and NaOH.
Solutions containing carefully controlled concentrations of HMD 66 samples were prepared to map out the concentration-space, and to
and NaOH were prepared by first mixing stock solutions. Different find the region in which spontaneous phase separation occurs. The total
2
M.J. Sargent et al. Chemical Engineering Journal 419 (2021) 129428
Fig. 3. Phase separation region. Preliminary work in assessing the phase separating region Fig. 4. Tie line error analysis. Tie lines, linking aqueous and organic phases in equilibrium
of the concentration-space of the water/HMD/NaOH system. Error bars are given by the size with each other, through the total concentration which underwent phase separation. Raw data
of the symbols. is shown alongside the correction to account for errors in concentration measurements.
volume of the stock solutions used to prepare these samples was 10 mL,
and there was negligible volume change on mixing. This data is shown
in Fig. 3.
Each symbol in Fig. 3 represents a sample made to a different
choice of concentrations for HMD and NaOH. Error bars for these
concentrations are given by the sizes of the symbols. 22 of the samples
exhibited phase separation, forming a small quantity of an organic
phase in addition to the much larger volume of aqueous phase. This
data so far gives a rough indication of the conditions required for phase
separation to occur.
3
M.J. Sargent et al. Chemical Engineering Journal 419 (2021) 129428
𝑦 = 𝑎0 + 𝑎1 𝑥 + 𝑎2 𝑥2 (4) Fig. 6. Tie line gradient modelling. Experimental tie lines, compared to a model for
predicting the gradients of the tie lines.
At any point along the length of this curve, (𝑥𝑇 , 𝑦𝑇 ), a tangent line
to the curve can be drawn. This tangent line will have a gradient, 𝑚,
that is the same as the local gradient of the quadratic, given by Eq. (6).
( )
𝑑𝑦
𝑚= = 𝑎1 + 2𝑎2 𝑥𝑇 (5)
𝑑𝑥 𝑥=𝑥
𝑇
From the gradient and the intercept point with the quadratic, the
equation of the line can be determined for each possible tangent point.
4
M.J. Sargent et al. Chemical Engineering Journal 419 (2021) 129428
Fig. 9. Tie line model. Model for the phase separating region of the HMD–NaOH system,
Fig. 8. Organic phase concentration modelling. Model fitting for the concentrations in the allowing prediction of compositions of separated phases and fraction of each phase. Tie lines
organic phase, as a function of the gradient of the model tie line which the total concentration are shown in decrements of 0.02 gradient below the plait point at 𝑚 = −0.2857. Volume
lies on. fraction lines are shown in increments of 10%.
parameters which were used in Fig. 7 are shown in Eq. (12). If the volume fraction lies outside the range (0, 1), it implies that the
( ) ( ) ( )
𝑎2 𝑎1 original composition will not undergo phase separation, as it cannot
1
𝑦𝑎𝑞 = 𝑎0 − 1 + 𝑚− 𝑚2 + 𝑚𝑏0 + 𝑚𝑏1 exp(𝑏2 𝑚) (11) balance mass of two phases in equilibrium. Another case to consider
4𝑎2 2𝑎2 4𝑎2
is that the tie line falls below the Plait point, which is the point at
𝑏0 = 3.050 [g L−1 ] which the tie line separating the two phases has no length, and so the
𝑏1 = 3027 [g L−1 ] (12) two phases would have the same composition. The plait point can be
𝑏2 = 41.86 [−] located by finding the value of 𝑚 at which the organic and aqueous
Whilst the concentrations of the two components within each phase concentration are predicted to have the same values, and for this fitting
are linked to one another, the fitting for the organic phase is entirely corresponds to 𝑚 = −0.2857. Therefore, if the value of 𝑚 calculated in
separate from that of the aqueous phase. However, the aqueous con- step 1 above is above this value, phase separation cannot occur.
centrations of HMD and NaOH are similarly linked to one another in This model is illustrated in Fig. 9. The diagonal lines are the tie
the organic phase. The concentrations of both HMD and NaOH in the lines, truncated to the concentrations predicted for the organic and
organic phase are shown in Fig. 8. aqueous phases. The curved lines denote the volume fraction of organic
The fitting lines shown in Fig. 8 were arrived at by a similar phase from 0%, which corresponds to only the aqueous phase, in
method to that used for Fig. 7, however the order of the components increments of 10% up to 100%, which corresponds to only the organic
was reversed. For the organic phase, the first component to have phase.
a fitting equation assigned to it was the NaOH concentration, 𝑦𝑜𝑟𝑔 ,
where the aqueous phase first fitted to the HMD concentration. This 4. Salt addition
dependence was again observed to follow an exponential relationship,
given by Eq. (13), for the same reasons as before. With the phase separation behaviour of the three component system
fully quantified, the next step is to investigate other compounds that
𝑦𝑜𝑟𝑔 = 𝑐0 + 𝑐1 exp(𝑐2 𝑚) (13)
could be added to promote the phase separation. The spontaneous
Once more, the tie line was used to convert this NaOH concentration generation of a second phase rich in HMD by the addition of NaOH was
into the equivalent HMD concentration, given by Eq. (14). Both of these identified as being similar to the separation of proteins from aqueous
equations were used to fit the parameters in Eq. (13), which are shown solutions using salts. The solubility of proteins in aqueous solutions can
in (15). be increased or decreased by the presence of salts, termed salting-in or
( ) ( ) ( ) salting-out respectively [17,24].
𝑎21 1 𝑎1 1 𝑐 𝑐
𝑥𝑜𝑟𝑔 = − 𝑎0 − − + 𝑚 + 0 + 1 exp(𝑐2 𝑚) (14) The first salt to be tested was sodium sulphate, Na2 SO4 . This was
4𝑎2 𝑚 2𝑎2 4𝑎2 𝑚 𝑚
chosen because it is the result of neutralising sodium hydroxide with
𝑐0 = 2.742 × 10−7 [g L−1 ] sulphuric acid, both of which are used during the nylon-6,6 production
𝑐1 = 3601 [g L−1 ] (15) process [22,25]. From this sodium sulphate a range of other salts were
𝑐2 = 13.12 [−] tested, by changing either the anion or cation.
The model is now complete, in that it can completely predict all
the properties of a system which undergoes phase separation. The 4.1. Solution preparation
procedure for using the model for a given theoretical mixture of HMD
and NaOH is as follows: Stock solutions for these salt addition tests were produced by the
same formulations as given in Section 2.1., with a few modifications.
1. Using the concentrations of HMD and NaOH, calculate the gra- Firstly, the HMD stock solutions were made up to 500 g L−1 , to allow
dient of the tie line which that mixture lies on (Eq. (8)). for samples to be made in closer concentrations to the organic phases
2. Using the tie line gradient, calculate the concentrations of each produced by the phase separation. Secondly, the salts were added to
component in each phase (Eqs. (10), (11), (13), and (14)). all three stock solutions in the same concentration. The result of this is
3. Using the compositions of each phase, calculate the volume that any combination of these salted stock solutions will maintain the
fraction of the total which is occupied by the organic phase (Eq. same salt concentration, so that only the effect of the HMD and NaOH
(1)). concentrations can be observed.
5
M.J. Sargent et al. Chemical Engineering Journal 419 (2021) 129428
Table 1
Salts tested for effect on phase separation behaviour.
Salt name Cation Anion Concentration
[g L−1 ]
Sodium sulphate Na+ SO2−
4
100
Sodium chloride Na+ Cl− 82.3
Sodium iodide Na+ I− 211
Sodium nitrate Na+ NO−3 120
Sodium carbonate Na+ CO2−
3
74.6
Potassium sulphate K+ SO2−
4
123
Ammonium sulphate NH+4 2−
SO4 93.0
Zinc sulphate Zn2+ SO2−
4
114
• An HMD/Na2 SO4 stock solution, containing 500±14 g L−1 of HMD Fig. 10. Sodium series. Effect of different sodium salts on the location of the phase
and 100 ± 2 g L−1 of Na2 SO4 . boundary for the HMD–NaOH system. Salts may increase or decrease the solubility of HMD,
• An NaOH/Na2 SO4 stock solution, containing 500 ± 12 g L−1 of which raises or lowers the required amount of NaOH to cause phase separation. All salts
NaOH and 100 ± 2 g L−1 of Na2 SO4 . are present at the same sodium ion concentration as 100 g L−1 Na2 SO4 .
This process was repeated for all of the different salts, which are
shown in Table 1. The mass concentrations of each salt were chosen
to ensure that each salt has the same sodium ion or sulphate ion
concentration as sodium sulphate, whichever was appropriate.
Instead of repeating the same tie line analysis and modelling that
was used for the three component system, a simplified analysis of the
phase separation behaviour was used to investigate the effect of the
addition of these salts.
In the first stage, solutions at different concentrations of HMD
were titrated with the NaOH stock solution until the presence of a
secondary phase was observed. Then in the second stage, solutions at
different concentrations of NaOH were titrated with the HMD stock
solution, until the secondary phase was again observed. By recording
Fig. 11. Sulphate series. Effect of different sulphate salts on the location of the phase
the volumes of each solution in each titration, the location of the phase boundary for the HMD–NaOH system. Salts may increase or decrease the solubility of HMD,
boundary in concentration-space can be determined without having to which raises or lowers the required amount of NaOH to cause phase separation. All salts
perform the full tie line analysis. are present at the same sulphate ion concentration as 100 g L−1 Na2 SO4 .
6
M.J. Sargent et al. Chemical Engineering Journal 419 (2021) 129428
5. Conclusions
7
M.J. Sargent et al. Chemical Engineering Journal 419 (2021) 129428
Acknowledgement [12] Albert J. (Jr) Isacks, Process for the purification of Hexamethylenediamine, 1965.
[13] Claudia Merk, Peter Bassler, Guido Voit, Hermann Luyken, Method for pu-
rifying Hexamethylenediamine in mixtures of Hexamethylenediamine and an
The authors would like to thank INVISTA Textiles (U.K.) Limited
unsaturated cyclic imine, 2002.
funding for this work. [14] Maria Ignez Broglio, Daniel Amoros, Jean Vannier, Dider Letourneur, Purification
of impure Hexamethylenediamines, 2013.
References [15] John J. Ostermaier, Leon S. Scott, Recovery of Hexamethylenediamine (HMD)
with low polarographically reducible impurities (PRI) from mixtures of HMD,
[1] Park L. Morse, Process Economics Program Report 31: Hexamethylenediamine, aminocapronitrile, and PRI, 2001.
Stanford Research Institute, 1967, pp. 17–19. [16] John J. Ostermaier, Distillative method for separating Hexamethylenediamine
[2] Adriana L. Botes, Alex Van Eck Conradic, Methods of producing 6-carbon from a mixture comprising Hexamethylenediamine, 6-aminocapronitrile and
chemicals via COA-dependent carbon chain elongation associated with carbon tetrahydroazepine, 2003.
storage, 2015. [17] C. Dennison, A Guide to Protein Isolation, Springer Netherlands, 2003, pp. 77–83.
[3] Adriana L. Botes, Alex Van Eck Conradic, Methods of producing 6-carbon [18] Archana Jain, Manju Gupta, Krishna K. Verma, Salting-out assisted liquid-
chemicals via methyl-ester shielded carbon chain elongation, 2017. liquid extraction for the determination of biogenic amines in fruit juices and
[4] Anthony P. Burgard, Priti Pharkya, Robin E. Osterhout, Microorganisms for the alcoholic beverages after derivatization with 1-naphthylisothiocyanate and high
production of adipic acid and other compounds, 2012. performance liquid chromatography, J. Chromatogr. A 1422 (2015) 60–72.
[5] Mark J. Burk, Anthony P. Burgard, Robin E. Osterhout, Priti Pharkya, Microor- [19] Rui Miguel Ramos, Pedro Francisco Brandão, José António Rodrigues, De-
ganisms and methods fro the biosynthesis of adipate, hexamethylenediamine and velopment of a SALLE-HPLC-FLD analytical method for the simultaneous
6-aminocaproic acid, 2018. determination of ten biogenic amines in cheese, Food Anal. Methods 13 (5)
[6] Olan S. Fruchey, E. Manzer, Leo, Dilum Dunuwila, Brian T. Keen, Brooke A. (2020) 1088–1098.
Albin, Nye A. Clinton, Bernard D. Dombek, Processes for producing heaxane- [20] Miguel Del Nogal Sánchez, Patricia Martín Santos, Cristina Pérez Sappó, José
diol (HDO), hexamethylenediamine (HMD) from fermentation broths containing Luis Pérez Pavón, Bernardo Moreno Cordero, Microextraction by packed sorbent
adipic acid, 2012. and salting-out-assisted liquid-liquid extraction for the determination of aromatic
[7] A.B. Dros, O. Larue, A. Reimond, F. De Campo, M. Pera-Titus, Hexamethylene- amines formed from azo dyes in textiles, Talanta 119 (2014) 375–384.
diamine (HMDA) from fossil- vs. bio-based routes: an economic and life cycle [21] V.F. Sergeeva, Salting-out and salting-in of non-electrolytes, Russ. Chem. Rev.
assessment comparative study, Green Chem. 17 (10) (2015) 4760–4772. 34 (1965) 309.
[8] Harshal Chokhawala, High yield route for the production of compounds from [22] Intratec Solutions LLC, Nylon 6, 6 Production - Cost Analysis - Nylon 66 E12A,
renewable sources, 2017. Intratec, 2019.
[9] Kenji Nishimura, Kanenobu Matsui, Koichi Hirai, Purification of Hexamethylene- [23] F.C. Campbell, Phase Diagrams: Understanding the Basics, ASM International,
diamine, 1973. 2012, pp. 191–199.
[10] Oscar R. Buehler, Harold F. Porter, Process for purifying Hexamethylenediamine, [24] Reginald H. Garrett, Charles M. Grisham, Biochemistry, fourth ed., Cengage
1981. Learning, University of Virginia, 2010, p. 488.
[11] Charles R. Campbell, Richard D. Chapman, Robert Johnson, Method for the [25] V.B. Gupta, V.K. Kothari, Manufactured Fibre Technology, Springer Netherlands,
purification of Hexamethylenediamine, 1962. 1997, pp. 319–324.