Download as pdf or txt
Download as pdf or txt
You are on page 1of 8

Int J Plast Technol

DOI 10.1007/s12588-015-9129-9
R E S E A R C H A RT I C L E

Reactivity ratio determination of styrene


and 2-ethylhexyl acrylate by least squares methods

Samaneh Ashenagar 1 & Farshid Ziaee 2 &


Iman Shabani 3

Received: 6 January 2015 / Accepted: 9 December 2015


# Central Institute of Plastics Engineering & Technology 2015

Abstract The reactivity ratios of styrene (STY) and 2-ethylhexyl acrylate (EHA)
synthesized by bulk copolymerization method, were determined by ordinary least
square (OLS) and generalized least square (GLS) approaches according to Inverted
Finemann - Ross (IFR), Yezrielev - Brokhina - Roskin (YBR), Kelen - Tds (KT)
methods. Composition results were summarized and the 95 % individual confidence
limits were determined and compared with each other and Error - in - Variables - Model
(EVM) approach. Based on the copolymer compositions, the calculated reactivity ratios
of STY and EHA through the GLS approach show acceptable percentage deviation
with the well-known EVM method, which proves GLS approach, is the best linear
estimator program for determining the monomer reactivity ratios. The reactivity ratios
of EHA and STY are equal to 0.292 and 0.980 in IFR technique through the generalized
least squares (GLS) approach which is closer to the point estimated from Error-in-
variables model (EVM) and also r EHA and r STY are equal to 0.292 and 0.979 in EVM
method, respectively. The present paper shows an estimation integral method for
investigating the reactivity ratios of styrene and 2-ethylhexyl acrylate in Eviews
software and the estimated values are discussed in terms of two regression approaches
and compared with EVM method.

Keywords Reactivity ratios . Generalized least square . Ordinary least square .


Error-in-Variables-Model . Styrene . 2-Ethylhexyl acrylate

* Farshid Ziaee
f.ziaee@ippi.ac.ir

1
Young Researchers and Elite Club, Sarvestan Branch, Islamic Azad University, Sarvestan, Iran
2
Iran Polymer and Petrochemical Institute, P.O. BOX: 14965/115, Tehran, Iran
3
Young Researchers and Elite Club, Ahvaz Branch, Islamic Azad University, Ahvaz, Iran
Int J Plast Technol

Introduction

Copolymers of acrylic esters (e.g., methyl acrylate and butyl acrylate) with styrene have
used for several applications as adhesives, non-woven binders and industrial coatings.
They are principally appropriate for latex paints since they resist UV radiation,
weathering and heat, and give a long-lasting color; for example n-butyl acrylate and
2-ethylhexyl acrylate are essential components of the acrylic latexes [1, 2]. Because the
physico-chemical properties of the copolymer vary with comonomer concentrations, it
is necessary to estimate the copolymer composition and determine the monomer
reactivity ratio which is a challenging area for both the academy and industry [35].
It is well known that monomer reactivity ratios can suggest the message of the relative
reactivity between the two monomers and predict the mechanistic aspects of copoly-
merization, relationship between feed ratio and conversion and sequence-length distri-
butions of copolymer [6]. The Mayo-Lewis [7] equation is a primitive procedure which
leads to different unreliable linear estimation methods for the computation of the
reactivity ratio. The way that reactivity ratio values are determined has been the subject
of a great number of publications [8]. In recent years, some investigations have
accumulated through the monomer reactivity ratios of styrene and 2-ethylhexyl acrylate
copolymer [9]. In our previous research, we estimated the reactivity ratios of the styrene/
2-ethylhexyl acrylate (STY/EHA) copolymer with different techniques (both linear and
nonlinear). We found that the results obtained from Inverted Finemann - Ross (IFR) [10],
Kelen - Tds (KT) [11] and Yezrielev Brokhina- Roskin (YBR) [12] methods were
closer to the point estimate from Error-in-variables model (EVM). The majority of the
methods are linearized techniques that may often afford less accurate estimations and do
not predict the error terms. On the other hand, EVM is a nonlinear approach which
correctly accounts for the error in all variables [2, 13]. So, in the present work we
investigate the monomer reactivity ratios by applying IFR, Tidwell - Mortimer (TM)
[14], KT, and YBR methods through ordinary least square (OLS) and generalized least
square (GLS) approaches in EViews software. The GLS estimator program is a statisti-
cally valid approach for determining the monomer reactivity ratios which increases the
F-statistics and R-square values of regression models [15, 16].
The novelty of this work was to show that the linear estimations obtained by the
generalized least squares (GLS) approach are in a fairly good agreement with most
reliable EVM approach. This analytical technique was then successfully confirmed
through handling regression approaches with error terms that are heteroskedastic or
autocorrelation, or both and clearly proved that the model was able to calculate the
reactivity ratios by accounting response error structure. However, we have tried to
compare the 95 % individual confidence limits of OLS and GLS approaches with each
other and calculate the deviation in reactivity ratio values with EVM method.

Experimental

Materials

Inhibitors were eliminated from the monomers by washing with a 5 % NaOH solution
and then washing with excess distilled water. The monomers were distilled under
Int J Plast Technol

decreased pressure and the median fractions were kept for the reactions. Benzoyl
peroxide was used as the initiator and was purified two times via a crystallization
process from chloroform by using methanol. Solvents (chloroform and methanol) were
used as received without further purification. All chemicals were obtained from Merck
Chemical Co. Inc. [2].

Copolymerization

The copolymerization reactions were performed with different mole ratios of


styrene and EHA (1: 9 through 9: 1) in the presence of benzoyl peroxide as
initiator. The reactions were done in sealed ampoules that were degassed by
several vacuum freeze-thaw cycles at 0.1 mmHg. The reactions were carried out
at 80.0 0.1 C to low conversions. For these low conversion experiments, the
resulting polymer remain dissolved in the reaction mixture. The polymer was
further dissolved with minimal amounts of chloroform and precipitated with
methanol. The procedure was performed three times. The resulting mass of
polymer was dried in a vacuum oven under 0.1 mmHg and 50 C until
constant weights were achieved. The conversion levels were measured via
gravimetric procedures [2].

Characterization

Compositional analysis of the copolymers was determined by 1 H-NMR spectra using a


Bruker FT-NMR AC 80 spectrometer with tetramethylsilane (TMS) as internal
reference.

Results and Discussion

It is well known that monomer reactivity ratios play the most essential role in
copolymerization, provide great incentives for revealing the copolymerization mecha-
nism, copolymer composition, distribution of comonomers units, relationship between
feed ratio and conversion and sequence-length distributions of copolymer [6, 17]. So, it
is quite necessary to obtain the values of the reactivity ratios in copolymerization
processes [1820]. To calculate the reactivity ratios of styrene (rSTY) and 2-
ethylhexyl acrylate (rEHA), copolymerization in low conversion (<10 %) was carried
out and the monomer feed ratio (f) and the resultant copolymer compositions (F), are
shown in Table 1.
From the molar ratios of the monomers in the feed (f) and the copolymer (F) the
reactivity ratios of comonomers were determined by the methods Inverted Finemann -
Ross (IFR) [10], Kelen - Tds (KT) [11] and Yezrielev - Brokhina - Roskin (YBR)
[12].
The inverted Fineman-Ross equation is defined as:

 
F1 F
r1 r2 2 1
f f
Int J Plast Technol

Table 1 Monomer and copoly-


Feed fraction styrene Copolymer fraction styrene Conversion (%)
mer composition for low
conversion
0.198 0.361 13.96
0.305 0.454 12.95
0.398 0.560 11.88
0.505 0.601 9.37
0.599 0.673 8.70
0.702 0.728 8.84
0.801 0.832 8.80
0.895 0.896 7.09

Where, f is the molar ratio of the monomers in the feed and F in the copolymer. To
gain a relative, reliable and accurate reactivity ratio in STY/EHA copolymerization,
computer programming method has been performed in EViews software methods
through ordinary least square (OLS) and generalized least square (GLS) approaches.
The comparison between 95 % (joint confidence limit) JCL of the reactivity ratio
values from GLS and OLS for IFR method are shown in Fig. 1.
The reactivity ratio can also be estimated with Kelen-Tudos equation which vari-
ables are expressed as:

  2
r2  r2 f F1 f G H
r1 ; G ;H ; ; 2
F F H H

Where and are the functions of the parameters G and H, and is a constant
equal to (HmaxHmin) 1/2, Hmax and Hmin being the lowest and highest H values,

Fig. 1 Comparison between 95 % JCL from OLS (line) and (dotted line) GLS for IFR method
Int J Plast Technol

respectively. The intercepts at = 0 and = 1 of the versus plot gives - r1/ and r2
[17]. Joint confidence limits for this method also has been plotted in Fig. 2.
Furthermore, the YBR method is proper for analytical evaluation of reactivity ratios
which is based on the following equation:
r s
p 1 f k
kf p r1 r2 3
kf k f

Where: k Ff .
The 95 % joint confidence limit obtained for the reactivity ratios is depicted in Fig. 3.
The values of monomer reactivity ratios and the statistics in regression analysis
which are evaluated according to the standard procedures given by IFR, KT and YBR
methods through OLS and GLS approaches are shown in Table 2.
From the obtained data, we can compare 95 % JCL with GLS and OLS approaches.
The Figures depict that the JCL decreased considerably through GLS approach for all
methods with respect to OLS and there is a significant difference between the JCL of
OLS and GLS. In addition, the results summarized in Table 2 show the statistics of
regression for OLS and GLS approaches and it is evident that the GLS approach
increases the values of the R-square of regression models and the order of magnitude of
F-statistics and decreases the variance of the monomer reactivity ratios. Moreover,
because existing statistical Eviews software can be functional, the publication of the
proposed approach could be easy and fast [17]. We also estimated the reactivity ratios
by EVM [2] method in our previous article and the deviation in reactivity ratio values
of GLS and OLS approaches with respect to the well-known EVM method is presented
in Table 3. Also, the comparison between the data points including the 95 % Joint

Fig. 2 Comparison between 95 % JCL from OLS (line) and (dotted line) GLS for KT method
Int J Plast Technol

Fig. 3 Comparison between 95 % JCL from OLS (line) and (dotted line) GLS for YBR method

confidence limit of the reactivity ratio values from GLS and OLS for IFR, KT, YBR,
and EVM methods have been summarized in Fig. 4.
It can be seen that all of the points which are obtained by applying GLS approach
through are certainly closer to the point estimated from EVM than OLS.

Conclusion

The point estimate values of monomer reactivity ratios and corresponding JCLs for
IFR, YBR and KT methods are obtained based on OLS and GLS approach. From the
above results, it is obvious that the estimated values of GLS, were near to the values
obtained by OLS, but it can be seen that the values obtained by GLS show a fairly good

Table 2 Results of GLS and OLS analysis forms of Mayo-Lewis terminal model at low conversion

OLS-GLS Methods rEHA rSTY S(rBA)a S(rSTY)a R2 Log(L)b Sum Standard.error


approaches squared of regression
residual

OLS IFR 0.2929 0.9813 0.0664 0.0174 0.97921 5.3539 0.1228 0.1430
KT 0.2775 0.9537 0.0403 0.0444 0.97790 12.5520 0.0203 0.0581
YBR 0.2807 0.9516 0.0330 0.0364 0.98624 6.02945 0.1037 0.1314
GLS IFR 0.2926 0.9804 0.0069 0.0153 0.99666 6.4739 0.1273 0.1964
KT 0.2818 0.9645 0.0288 0.0412 0.98728 13.3170 0.0167 0.0528
YBR 0.2930 0.9822 0.0173 0.0362 0.98980 7.6265 0.0695 0.1076

a
standard deviation
b
likelihood function
Int J Plast Technol

Table 3 The estimated values of monomer reactivity ratios for three copolymerization approach

Analysis Methods rEHA rSTY Deviation with respect to EVM (%)

rEHA rSTY

OLS IFR 0.2929 0.9813 0.09 0.23


KT 0.2775 0.9537 1.45 2.53
YBR 0.2807 0.9516 1.13 2.74
GLS IFR 0.2926 0.9804 0.06 0.14
KT 0.2818 0.9645 1.02 1.45
YBR 0.2930 0.9822 0.10 0.32
a
NLLS EVM 0.2920 0.9790 - -

a
Nonlinear least square

agreement with EVM method which is a nonlinear analysis based on the accurate
structure of the Mayo - Lewis model. All of the above mentioned results confirm that

Fig. 4 Comparison between the estimated values of monomer reactivity ratios for three copolymerization
approach (The EVM values are in the beginning point of the arrow)
Int J Plast Technol

GLS method is the best linear unbiased estimation for determination of monomer
reactivity ratios in copolymerization systems. So, based on this approach we can
estimate reactivity ratio values mostly through more reliable linear estimation methods.

References

1. Kammona O, Pladis P, Frantzikinakis CE, Kiparissides C (2003) A comprehensive experimental and


theoretical investigation of the styrene/2-ethylhexyl acrylate emulsion copolymerization. Macromol Chem
Phys 204:983999
2. Kavousian A, Ziaee F, Nekoomanesh MH, Leamen MJ, Penlidis A (2004) Determination of monomer
reactivity ratios in styrene/2-ethylhexylacrylate copolymer. J Appl Polym Sci 92:33683370
3. Bakhshi H, Zohuriaan-Mehr MJ, Bouhendi H, Kabiri K (2010) Emulsion copolymerization of butyl
acrylate and glycidyl methacrylate: determination of monomer reactivity ratios. Iran Polym J 19:781789
4. Siolek M, Matlengiewicz M (2014) Reactivity ratios of butyl acrylates in radical copolymerization with
methacrylates. Int J Polym Anal Ch 19:222233
5. Kazemi N, Lessard BH, Maric M, Duever TA, Penlidis A (2014) Reactivity ratio estimation in radical
copolymerization: from preliminary estimates to optimal design of experiments. Ind Eng Chem Res. doi:
10.1021/ie402765k
6. Kaya E (2011) Copolymers of 4-methoxybenzyl methacrylate and isobornyl methacrylate: synthesis,
characterization and determination of monomer reactivity ratios. J Chem Soc Pak 33:555561
7. Mayo FR, Lewis FM (1944) Copolymerization. I. A basis for comparing the behavior of monomers in
copolymerization; the copolymerization of styrene and methyl methacrylate. J Am Chem Soc 66:1594
1601
8. Ashenagar S, Ziaee F, Jalilian SM (2013) Calculation of reactivity ratios of methacrylic acid-ethyl acrylate
copolymer by on-line quantitative 1H NMR spectroscopy. Iran Polym J 22:635639
9. Deb PC (2005) Monomer reactivity ratios in styrene/2-ethylhexylacrylate copolymer. J Appl Polym Sci
97:17531754
10. Finemann M, Ross SD (1950) Linear method for determining monomer reactivity ratios in copolymer-
ization. J Polym Sci 5:259262
11. Kelen T, Tds F (1974) A new improved linear graphical method for determining copolymerization
reactivity ratios. React Kinet Catal L 1:487492
12. Ezrielev A, Brokhina E, Roskin E (1969) Analytical method for calculating copolymerization constants.
Vysokomol Soedin A 11:16701680
13. Polic AL, Duever TA, Penlidis A (1998) Case studies and literature review on the estimation of
copolymerization reactivity ratios. J Polym Sci Pol Chem 36:813822
14. Tidwell PW, Mortimer GA (1965) An improved method of calculating copolymerization reactivity ratios.
J Polym Sci A 3:369387
15. Habibi A, Vasheghani-Farahani E, Semsarzadeh MA, Sadaghiani K (2003) A generalized least square
model for the determination of monomer reactivity ratios in free radical copolymerization systems.
Macromol Theor Simul 12:184195
16. Habibi A, Vasheghani-Farahani E, Semsarzadeh MA, Sadaghiani K (2003) Monomer reactivity ratios for
lauryl methacrylateisobutyl methacrylate in bulk free radical copolymerization. Polym Int 52:14341443
17. Ashenagar S, Ziaee F (2013) Estimation of reactivity ratios of styrene/butyl acrylate copolymer by
ordinary and generalized least square methods. Iran Polym J 22:511518
18. Ayaz N, Bezgin F, Demirelli K (2012) Polymers based on methacrylate bearing coumarin side group:
synthesis via free radical polymerization, monomer reactivity ratios, dielectric behavior, and thermal
stabilities. ISRN Polym Sci 2012:113
19. Fu Z, Fan Y, Fan Z (2011) Temperature-structure dependence of poly (1-octene-co-t-butyl acrylate)
prepared by conventional free radical polymerization. Iran Polym J 20:223235
20. Kumar KR, Feroz SK, Rao PR (2008) Reactivity ratios, thermal and di-electrical properties of 1, 11 azobis
(cyclohexanecarbonitrile) initiated nitrile copolymer. Asian J Res Chem 1:5863

You might also like