Polymer Testing 86 (2020) 106446

Contents lists available at ScienceDirect

Polymer Testing
journal homepage: http://www.elsevier.com/locate/polytest

Analysis Method

A spectrophotometric method for estimation of titanium alkoxide, a

catalytically active form of titanium in oligomer/ polymers
K Sattaiah Naidu a, **, Abhijeet S. Kate a, Vikas Kshirsagar b, R. Ganeshan a, Tukaram Gunale b,
Bing Zhou c, Samir Anapat b, Yusuf Sulub c, Arun Kumar a, Narayana Rao a
a
Analytical Department, SABIC, Bangalore, 562125, India
b
Petrochemical Department, SABIC, Bangalore, 562125, India
c
Analytical Department, SABIC, Mt Vernon, IN, 47620, USA

A R T I C L E I N F O A B S T R A C T

Keywords: A spectrophotometric method was developed for the selective estimation of titanium in alkoxide form in the
Spectrophotometric method presence of TiO2. Polymers like polybutylene terephthalate have been synthesized by using homogeneous
Titanium cresolate catalyst such as titanium tetra-isopropoxide (TTIP). These types of catalysts are susceptible to moisture and
Titanium alkoxide
convert to inactive TiO2. Hence, it is essential to have analytical method for selective estimation of alkoxide form
Titanium tetra-isopropoxide
Polybutylene terephthalate
of titanium. In this study, we have used O-cresol to convert TTIP to tetra O- cresyl titanate, a colored compound,
which was quantitatively analyzed by spectrophotometric technique. Polymer matrix was used in standard so-
lution to compensate the effect of matrix during UV-VIS measurements. Method validation was successfully
carried out for linearity, reproducibility, recovery, specificity and the observed acceptable results as per ICH
guidelines. 2-sample t-test was performed to check the variation observed between two analysts and the observed
p-value indicated the results were statistically similar to each other. Mechanism of color formation was estab-
lished by synthesis of colored complex called tetra O-cresyl titanate. The structure was supported by ICP, CHNS
and NMR analysis. The method is simple and easy to implement at manufacturing plant for initial monitoring of
active catalyst concentration in polymer/oligomer samples.

1. Introduction catalyst concentration does not remain stable and it becomes difficult to
maintain the catalytic activity. It is important for the process chemist to
Thermoplastic polymers such as terephthalates have been used know exact concentration of the catalyst (active form) prior to poly-
largely in plastic industries such as electrical, lighting, automotive, merization steps. If required, make up amount of TTIP can be added in
telecommunications, medical, food, appliances, business machines and the process to maintain the desired catalyst concentration. As discussed
textiles. Hence, the demand for polybutylene terephthalate (PBT) has earlier, in this process, titanium can be present in alkoxide form or TiO2
been growing tremendously [1–3]. To support the increasing demand, form. The investigation was focused on the quantitative estimation of
the producers are optimizing the PBT production process. Researchers the titanium alkoxide in the PBT polymer/oligomer matrix in the pres-
are screening different starting materials such as purified terephthalic ence of TiO2.
acid (PTA) or dimethyl terephthalate (DMT) [4]. PTA has several ad- Total Ti estimation can be done by ICP technique together with
vantages such as product quality and cost effectiveness, however, water ashing of sample followed by dissolution in H2SO4. However, this
is a byproduct of the reaction. Ti-based catalysts such as TTIP (titanium method cannot distinguish between TiO2 and alkoxide form of titanium.
tetraisopropoxide) are often used as a catalyst in the synthesis to The desired method should be able to distinguish between TiO2 and
accelerate the reaction by a factor of 100–1000 [5–7]. However, it has alkoxide form of titanium. These are homogenous catalysts and their
been observed that these catalysts degrade in the presence of water to concentration in polymerization reactions are in ppm level. The desired
form TiO2, which lead to reduced catalytic activity, increased quality method should have sufficient sensitivity to detect ppm level of titanium
issues and production cost for PBT [8,9]. alkoxide in oligomer/polymer samples. Here, we report a spectropho-
Due to the in-situ degradation in the presence of moisture, the tometric technique, which was developed to determine selectively,

* Corresponding author.
E-mail address: [email protected] (K. Sattaiah Naidu).

https://doi.org/10.1016/j.polymertesting.2020.106446
Received 1 October 2019; Received in revised form 27 January 2020; Accepted 15 February 2020
Available online 25 February 2020
0142-9418/© 2020 Elsevier Ltd. All rights reserved.
K. Sattaiah Naidu et al. Polymer Testing 86 (2020) 106446

Fig. 1. Reaction of TTIP with O-cresol in presence of HCl (DCM was used as solvent).

catalytically active Ti content that is titanium alkoxide in oligomer/ flask then added 5 mL of o-cresol: DCM (3:1) mixture. Heated the so-
polymer samples. Early observations showed that O-cresol when reacted lution at 70 ! C till the PBT dissolves completely. Cooled the solution at
with TTIP, Titanium (IV) butoxide and Titanium (IV) ethoxide results to room temperature and added 100 μL of 3% HCl solution and mixed
give yellow to red colored complex. In this study, TTIP was selected for properly. Diluted the solution up to 25 mL with DCM.
the detailed analytical method development and validation study. In
short, the method involves the reaction of alkoxide form of Ti with o- 2.4. Preparation of blank solution (DMT matrix)
cresol & dichloromethane (DCM) mixture in HCl environment, resulting
into a stable yellow colored complex i.e. tetra O- cresyl titanate, which Weighed accurately about 0.12 g of DMT in 25 mL volumetric flask
was quantified by spectrophotometer at wavelength of 405 nm. DCM then added 5 mL of o-cresol: DCM (3:1) mixture. Heated the solution at
was used as solvent in this method. 70 ! C till DMT dissolves completely. Cooled the solution to room tem-
perature, added 100 μL of 3% HCl solution, and mixed thoroughly.
2. Experimental section Diluted the solution up to 25 mL with DCM.

2.1. Materials and methods 2.5. Sample measurement protocol

All the chemicals used were AR grade. O-cresol, Dichloromethane Instrument was calibrated before use for the measurements. Cleaned
(DCM), Isopropyl alcohol (IPA), Hydrochloric acid (HCl), TTIP, Zirco- the cuvette and rinsed with the desired solution at least three times
nium (IV) propoxide, Germanium (IV) isopropoxide, Tin (IV) isoprop- before measuring. Filled the cuvette and kept in an appropriate place in
oxide, Titanium ethoxide and Titanium butoxide were procured from the instrument. Initially, performed auto-zero and baseline correction
Sigma-Aldrich. UV-2550 UV–Vis Spectrophotometer - Shimadzu or with blank solution and kept the blank solution in reference cuvette.
equivalent was used for the spectrophotometric measurement at wave- Measured the absorbance of different linearity levels and plot a cali-
length of 405 nm, slit width of 2 nm, threshold of 0.001, scan mode: bration curve. Measured the absorbance of sample solution against the
single and scan speed: medium. 1H and 13C spectra were obtained using blank at 405 nm utilizing a 1 cm path length cell. Calculated the con-
Avance Bruker 300 MHz instrument. centration of sample solution and then applied dilution factor to get final
concentration of active Ti in the sample.
Note that the UV–Vis spectrophotometer measurements of the solu-
2.2. Reagents
tions need to perform within 4 h of sample preparation.
O-cresol: DCM (3:1): To prepare 1000 mL of the solvent mixture,
2.6. Synthesis of tetra O- cresyl titanate
melt o-cresol in a water bath at 80 ! C. Mix 750 mL of o-cresol with 250
mL of DCM in a bottle and mix and cap it well.
Titanium tetracresolate synthesis was carried out in a 50 mL capacity
3% (V/V) HCl solution in Isopropyl alcohol: 1 mL of 30% suprapure
three neck-round bottom flask (RBF) dipped in oil bath. One neck of RBF
HCl was diluted to 10 mL with isopropyl alcohol.
was fitted with water condenser and oil bubbler. Thermo-well was fitted
in one neck to measure actual temperature of reaction mixture. Middle
2.3. Sample preparation neck of RBF was connected to distillation condenser to collect distillate
(isopropyl alcohol). Heidolph heating and stirring hotplate was pro-
2.3.1. Standard solution preparation vided for heating the oil bath. The equipment was provided with auto-
matic temperature control and variable stirrer speeds. o-cresol (8.634g,
2.3.1.1. Stock solution preparation. Added 631 μL of TTIP in 50 mL of 79.84 mmol) and titanium isopropoxide (5.153g, 18.13 mmol) were
IPA. Mixed thoroughly and diluted to 100 mL with IPA (Stock-1: 1000 weighed and added in 50 mL round bottom flask at room temperature
ppm of Ti). Diluted 5 mL–10 mL with IPA (Stock "2: 500 ppm). All the [10]. Reaction mixture was heated to 180 ! C for 2 h. Isopropyl alcohol
above manipulations were carried out in glove box under nitrogen (4.15g, 95.23% yield) was collected in collection vessel. Titanium tet-
atmosphere. racresolate (8.93g, 94.70% yield) catalyst along with cresol was ob-
tained as residue.
2.3.1.2. Calibration standard solutions preparation. Weighed 0.12 g of
DMT in five volumetric flask of 25 mL capacity. Added 5 mL of o-cresol: 3. Results and discussion
DCM (3:1) mixture and 5, 10, 25, 50 and 100 μL of 500 ppm stock so-
lution for preparing 0.1, 0.2, 0.5, 1 and 2 ppm respectively (5 levels). This exercise refers to the development of a spectrophotometric
Heated the solution at 70 ! C until it dissolves completely. Cooled the method for selective determination of alkoxide form of titanium. The
solution to room temperature and added 100 μL of 3% HCl solution in method is based on the in-house observation of interaction of TTIP with
IPA. Mixed it thoroughly and diluted the solution up to 25 mL with DCM. phenol to give red color [10] The similar phenomenon was observed
when TTIP reacted with o-cresol: dichloromethane mixture. The plau-
2.3.2. PBT sample preparation sible mechanism of color development would be the formation of
Weighed accurately about 0.1 g of PBT sample in 25 mL volumetric tetra-O-cresyl titanate after reaction of TTIP with O-cresol as shown in

2
K. Sattaiah Naidu et al. Polymer Testing 86 (2020) 106446

Fig. 2. Overlaid spectrum of different concentration of Ti in DMT matrix.

Fig. 1 [11]. The hypothesis was confirmed by reaction of TTIP with

O-cresol (1:4 M ratio) at 180 ! C and under N2 atmosphere. Isopropyl
alcohol was collected separately from distillation column. The structure
of tetra-O-cresyl titanate was confirmed by NMR analysis. The purity of
product was checked with CHNS and ICP analysis [see Table S1 of the
Supporting Information]. It was found to be highly moisture sensitive.
Dissolution in dichloromethane-d2 followed by NMR spectral charac-
terization of the product evidenced it to be an mixture of complex [Ti
(O–C6H4–CH3)4] and free ligand (O-cresol). The 1H NMR spectrum
features two sharp singlets at δ 2.01 and 2.12 ppm, which are attribut-
able to methyl group of requisite complex and along with the free ligand
(O-cresol). On the other hand, the 13C{1H} NMR spectrum consists of
two sets of signals which are associated with titanium complex and free
O-cresol, details are given [see Figure S1&2 of the Supporting
Information].
In early trials, solutions with various concentrations of TTIP in iso-
propanol were analyzed by spectrophotometer. TTIP is not stable in Fig. 3. Linearity graph of different concentration of Ti in DMT matrix.
open environment and hence stock solution was prepared in inert at-
mosphere by using glove box. Linear correlation was observed between dimethyl terephthalate (DMT) were used for early trials, however due to
absorbance and concentration of titanium. However, when TTIP was the limited solubility of PTA in the selected solvents, DMT was chosen as
spiked with PBT matrix, interference was seen from matrix as shown in blank matrix. Linearity was plotted and the observed regression coeffi-
[see Figure S3 of the Supporting Information]. Other concern was the cient was greater than 0.995 (Figs. 2 and 3). The method was further
instability of developed color and hence it was difficult to implement the fine-tuned by optimizing various parameters such as weight of sample,
method as such. Experiments were conducted to improve the color and solvent volume, heating procedure to dissolve sample, moisture effect
solution stability such as effect of heat and pH. In presence of HCl, the and instrument parameters.
color intensity increases drastically and color of the solution remained The method was validated by using a polymer sample name Sample-
stable even after 12 h as shown in [see Figure S4a&b of the Supporting X, which had only TTIP and no TiO2. Sample-X was synthesized by DMT
Information]. This was the breakthrough of the method development route in which TTIP remained intact, as methanol was the byproduct of
process, as with the improved method, the desired sensitivity could be DMT process and hence no degradation expected due to the absence of
achieved and can be implemented easily. moisture. In this case, the total titanium is expected to be in the alkoxide
Further trials were carried out to resolve issue of matrix in- form and so estimation by spectrophotometric technique and ICP tech-
terferences, in which calibration standards containing PBT matrix, and nique can be comparable. It was observed that in spectrophotometric
the blank solution containing the same matrix were used in reference method, background correction with DMT matrix (Blank solution) is
cell. These experiments were planned to address the issue of background very crucial and needs to have a matrix blank in reference cell while
interferences. Selection of PBT matrix that can be used as blank for carrying out spectrophotometric measurements.
interference, turned out to be crucial factor. It was difficult to find The expected concentration of Ti in the sample-X was in low ppm
oligomer or polymer without catalyst and hence monomers were used level and hence 0.1–2 ppm Ti was selected as a range for the calibration
for matrix matching study. Purified terephthalic acid (PTA) and standard measurements. Excellent linearity was observed with

3
K. Sattaiah Naidu et al. Polymer Testing 86 (2020) 106446

Table 1 The concept was tested on other metal isopropoxides such as Zirco-
Reproducibility study with Sample-X sample. nium (IV) propoxide, Tin (IV) isopropoxide and Germanium (IV) iso-
Replicate # Titanium content (ppm) propoxide and observed that only TTIP gave colored solution after
reaction with O-cresol [see Figure S5 of the Supporting Information].
Rep1 94.06
Rep2 102.13 Further, two other titanium alkoxides were tested such as titanium
Rep3 91.80 butoxide and titanium ethoxide and both titanium alkoxides showed
Rep4 94.49 reddish yellow color after reaction with O-cresol [see Figure S6 of the
Rep5 91.40 Supporting Information]. These observations suggested that the method
Rep6 88.91
Average 93.80
could be extrapolated to any titanium alkoxide catalysts.
STD DEV 4.55
%RSD 4.85 4. Conclusions

A simple spectrophotometric technique was developed for specific

Table 2 determination of alkoxide form of titanium in PBT oligomer/polymer
Recovery study by using Sample X (Ti content – 93.8 ppm) as matrix. samples. This method involves the reaction of titanium alkoxide with o-
Spiked Ti (μg) Recovered Ti (μg) % Recovery cresol in acidic environment, resulting into a stable yellow color com-
plex, which has been quantified by spectrophotometer at 405 nm.
10 11.05 110.5
20 22.05 110.3
Mechanism of color formation was established by synthesis of colored
30 27.47 91.6 complex called tetra O-cresyl titanate. The structure was supported by
NMR, ICP and CHNS analysis. The method was successfully validated
and implemented for analysis of in-house oligomer/polymer samples.
Table 3
Study on blending of two PBT samples. Authors contrubution
A Poly 69 407.15 ppm
B R350 93.8 ppm All author were involved in the work and in writing the manuscript.
AþB Expected(ppm) Observed(ppm)
Declaration of competing interest
Mix (50:50) 250.475 231.8
Mix (20:80) 156.47 141.56
Mix (80:20) 344.48 321.82
The authors declare that they have no known competing financial
interests or personal relationships that could have appeared to influence
the work reported in this paper.

Table 4
Study on Moisture interference in Sample –X. CRediT authorship contribution statement

Active Ti Content (addition Active Ti Content (addition Active Ti Content

K Sattaiah Naidu: Conceptualization, Methodology, Investigation,
of water before complex of water after complex (no addition of
formation) formation) water) Writing - original draft, Validation. Abhijeet S. Kate: Data curation,
Methodology, Visualization, Writing - review & editing. Vikas Kshir-
95.2 ppm 92.46 ppm 93.8 ppm
sagar: Formal analysis, Investigation. R. Ganeshan: Data curation,
Formal analysis. Tukaram Gunale: Project administration. Bing Zhou:
regression coefficient greater that 0.995 (Fig. 2). Dichloromethane was Investigation. Samir Anapat: Writing - review & editing. Yusuf Sulub:
used as diluent since PBT matrix gets precipitate out in isopropanol Writing - review & editing. Arun Kumar: Supervision, Writing - review
(IPA). Sample was dissolved in 5 mL of o-cresol: dichloromethane (3:1) & editing. Narayana Rao: Supervision, Writing - review & editing.
mixture and then added 100 μL of 3% HCl (in IPA) and diluted to 25 mL
with dichloromethane. Spectrophotometric measurements were carried Acknowledgment
out at 405 nm. Observed concentration of active Ti in sample X was 93.8
ppm, which was similar to ICP-OES results 95.4 ppm. The data indicated We are grateful to the SABIC for financial support.
that method could be implemented if TiO2 does not interfere in the
process of color formation. Appendix A. Supplementary data
An experiment was performed with the addition of high TiO2 con-
centration (1% solution) to the matrix to check for interferences. Color Supplementary data to this article can be found online at https://doi.
was not formed and observed no absorbance at 405 nm, which suggested org/10.1016/j.polymertesting.2020.106446.
TiO2 does not interfere in this method and it is specific for alkoxide form
of Ti. Repeated analysis (6 replicates) of sample-X showed very good References
reproducibility as % RSD was less than 5% (Table 1). Excellent recovery
was observed, when spiking study was performed (Table 2). The new [1] H.A. Alidedeoglu, P. Kumar, WO 2016202615 A1, 2016.
spectrophotometric method was further tested with check standards, [2] Mark, Bikales, Overberger, Menges, Encyclopedia of Polymer Science and
Engineering, second ed., vol. 12, Wiley-Interscience, New York, 1988 https://doi.
which were prepared by blending two sets of polymer samples. The org/10.1002/pol.1988.140260115.
observed results were matching very well with the theoretical value [3] James M. Margolis (Ed.), Engineering Thermoplastics: Properties and Applications,
(Table 3). Experiments were conducted to prove that the presence of Marcel Dekker, New York, 1985, https://doi.org/10.1002/pol.1986.140240616|.
[4] H.L. Paranjpe, B.B. Mathur, Polyester raw material DMT or TPA, Chem. Age India
moisture in solvent does not affect the results measured by the spec- 38 (4) (1987) 177.
trophotometric technique (Table 4). Method robustness was checked by [5] J. Zhu, J. Cai, W. Xie, P. Chen, M. Gazzano, M. Scandola, R.A. Gross, Poly(butylene
analyzing sample X by two Scientists and results compared with a sta- 2,5-furan dicarboxylate), a biobased alternative to PBT: synthesis, physical
properties, and crystal structure, Macromolecules 46 (2013) 796–804, https://doi.
tistical tool, 2-sample t-test. The observed p value (0.195 Value) indi- org/10.1021/ma3023298.
cated that the results were statistically similar [see Table S2 of the [6] E. Walch, R.J. Gaymans, Synthesis and properties of poly(butylene terephthalate)-
Supporting Information]. b-polyisobutylene segmented block copolymers, Polymer 35 (3) (1994) 636–641,
https://doi.org/10.1016/0032-3861(94)90523-1.

4
K. Sattaiah Naidu et al. Polymer Testing 86 (2020) 106446

[7] H. A Alidedeoglu, T. Ding, G Kannan, US Patent US 2014/0031457 A1, January 30, [10] F. Mijangos, F. Varona, N. Villota, Changes in solution color during phenol
2014,. oxidation by fenton reagent environ, Sci. Technol. 40 (2006) 5538–5543, https://
[8] D.A.H. Hanaor, I. Chironi, I. Karatchevtseva, G. Triani, C.C. Sorrell, Single and doi.org/10.1021/es060866q.
mixed phase TiO2 powders prepared by excess hydrolysis of titanium alkoxide, [11] S.C. Chaudhry, Anup Kumar, S.S. Bhatt, S. Neeraj, Synthesis and reactions of
Adv. Appl. Ceram. 111 (2012) 149–158, https://doi.org/10.1179/ tetrakis(4-methoxyphenoxo)titanium(IV), Synth. React. Inorg. Met. Org. Chem. 22
1743676111Y.0000000059. (1992) 1–9, https://doi.org/10.1080/15533179208020635.
[9] Chung Y. Kim, Hyun N. Cho, Young C. Kim, Process for the Preparation of
Polybutylene Terephthalate Based Polymer, US5237042 A, 1993.