Highly sensitive determination of low-level water content in organic solvents using novel solvatochromic dyes based on thioxanthone

Article abstract of DOI:10.1039/c3cc43608j

Two thioxanthone-based fluorescent probes exhibited prominent solvatofluorochromism, and they were further found to be useful as fluorescence indicators for the qualitative and quantitative detection of low-level water content in various solvent media.

Full text of DOI:10.1039/c3cc43608j

ChemComm
View Article Online
View Journal
COMMUNICATION
Highly sensitive determination of low-level water
content in organic solvents using novel solvatochromic
Cite this: DOI: 10.1039/c3cc43608j
dyes based on thioxanthone†
Li Ding,a Zhiyun Zhang,a Xin Lib and Jianhua Su*a
Received 14th May 2013,
Accepted 24th June 2013
DOI: 10.1039/c3cc43608j
www.rsc.org/chemcomm
Two thioxanthone-based fluorescent probes exhibited prominent traditional methods for the quantification of humidity in gas
solvatofluorochromism, and they were further found to be useful or liquid phases, for instance, psychrometry, Peltier dew point
as fluorescence indicators for the qualitative and quantitative methods, drying to constant weight, distillation, physical methods
detection of low-level water content in various solvent media.
(such as measuring density or electrical conductivity) and the most-
frequently-used Karl-Fischer method.14 Although these classical
Solvent polarity has profound effects on both absorption and procedures feature high sensitivities and might be applicable to a
emission spectral properties of fluorophores. Typically, since pola- large variety of samples, they also have some disadvantages: they
rities of the ground and excited states of a fluorophore are usually often require particular instruments and excessive sample pretreat-
different, a change in the solvent polarity will lead to differential ments, occasionally yield unsatisfactory results or consume too
stabilization of the ground and excited states, and thus, a change in much time for large-scale applications. Based on the current situa-
the energy gap between these electronic states. Consequently, tion, extensive research efforts have been aimed at developing easy-
variations in the position, intensity, and shape of the absorption to-use portable instruments or suitable online analysis methods.
and emission spectra can be direct measures of the specific inter- Recently, optical sensors for water sensing have drawn considerable
actions between the solute and solvent molecules.1–4
attention due to certain advantages especially over electrochemical
What is more, there also exist specific solvent effects: devices including immunity to electromagnetic interference, flex-
hydrogen bonding, preferential solvation, acid–base chemistry ibility in readout, possibility of remote and in situ monitoring as well
or charge-transfer interactions, to name a few. Hydrogen-bonding as ease and low-cost of fabrication.15 Particularly using fluores-
interactions in the excited state might lead to fluorescence quench- cence methods, parameters such as lifetime, intensity and wave-
ing due to rapid internal conversion or proton/electron transfer.5–8 length ratiometric sensing have been widely studied and applied
In addition to the above general and specific solvent– for detection of water content in organic solvents.16–21
fluorophore interactions, many fluorophores can form an inter-
Herein, two solvatochromic probes 2,7-bis(diphenylamino)-
nal charge transfer (ICT) state, or a twisted internal charge thioxanthen-9-one (BDPA-TXO) and 2,7-bis(4-triphenylaminyl)-
transfer (TICT) state. Following excitation there can be an thioxanthen-9-one (BTPA-TXO) were designed and synthesized
increase in charge separation within the fluorophore, which (Scheme 1 and Scheme S1, ESI†). Both the indicators underwent
can considerably affect the molecular orbital energy levels and pronounced intramolecular charge transfer (ICT) upon excitation,
accordingly influence the photophysical properties.9–12
and the fluorescence emission spectra exhibit prominent solvato-
The qualitative and quantitative detection of low-level water fluorochromism with the increase in solvent polarity. In addition,
content as impurity in organic solvents is of great significance in
several fields of chemistry and industrial processes such as pharma-
ceutical manufacturing, food processing, paper production, bio-
medical and environmental monitoring.13 There exist extensive
a Key Laboratory for Advanced Materials and Institute of Fine Chemicals,
East China University of Science & Technology, Shanghai 200237, China.
E-mail: bbsjh@ecust.edu.cn; Fax: +86 21-64252288
b Department of Theoretical Chemistry and Biology, School of Biotechnology,
KTH Royal Institute of Technology, Stockholm SE-10691, Sweden
† Electronic supplementary information (ESI) available. CCDC 921469 and
921470. For ESI and crystallographic data in CIF or other electronic format see
Scheme 1 X-ray crystal structure of (a) BDPA-TXO and (b) BTPA-TXO.
DOI: 10.1039/c3cc43608j
c
This journal is The Royal Society of Chemistry 2013
Chem. Commun.

View Article Online
Communication
ChemComm
Table 1 Frontier molecular orbital contours and percentage of contributions
the emission bands were effectively quenched along with batho-
chromic shifts upon addition of increasing concentration of water
in four organic solvents. Consequently, these fluorescence probes
are useful for the determination of low-level water content with
theoretical detection limits of 19–60 ppm (by weight), which
approach modern coulometric Karl-Fischer analyzer’s detectability
with a detection limit of 1 ppm. Moreover, some rectangular
dyestuff test papers were also prepared to realize the possibility
of qualitative detection of low-level water content in several
common organic solvents.
HOMO
LUMO
Compound
BDPA-TXO
For both BDPA-TXO and BTPA-TXO, the absorption spectral
maximum is almost independent on the solvent polarity; the
magnitude of the solvatochromic shift from nonpolar solvent
(hexane) to polar solvent (ethanol, EtOH or dimethyl form-
amide, DMF) is 1–2 nm (Fig. S9, ESI†).
BTPA-TXO
Unlike the case of labs, the fluorescence emission spectra
exhibit prominent solvatofluorochromism. Some parameters of molecular orbitals (HOMOs) and the lowest unoccupied molecular
organic solvents with different polarities and the solution orbitals (LUMOs) (Table S3, ESI†). As shown in Table 1, the HOMOs
spectral properties of BDPA-TXO and BTPA-TXO are specifically for BDPA-TXO and BTPA-TXO are both delocalized over the whole
listed in Table S1 (ESI†). With the increase in solvent polarity, molecule, with substantial contribution from the diphenylamine
significant spectral bathochromic shifts (BDPA-TXO: 481 to and triphenylamine moieties (60% for BDPA-TXO and 86% for
596 nm, from hexane to EtOH; BTPA-TXO: 441 to 602 nm, from BTPA-TXO). It is also found that the LUMOs for both compounds
hexane to DMF, Fig. S10 and Table S1, ESI†) and evident are localized predominantly on the thioxanthone unit, where a
changes in fluorescence quantum yield (BDPA-TXO: 21% to significant contribution (36–37%) arises from the carbonyl group.
1%, from hexane to EtOH; BTPA-TXO: 9% to 1%, from hexane Thus the computed HOMO and LUMO indicate a charge transfer
to DMF, Table S1, ESI†) were observed. Both BDPA-TXO and process from the electron donating diphenylamine and triphenyl-
BTPA-TXO in various solvents exhibited dramatic fluorescent amine to the electron withdrawing thioxanthone. For both com-
color variations under illumination with 365 nm light: greenish pounds, the partial atomic charges localized on the carbonyl group
blue to orange red and blue to orange red, respectively (Fig. 1). changed from À0.16 to À0.35 upon excitation from the ground
Diphenylamine and triphenylamine are both well-known to be state to the first excited state, corresponding to a charge transfer of
very effective p-electron-donating groups with higher donor 0.19 e. These computed charges, obtained from the Mulliken
strength. In the donor–thioxanthone–donor structural framing of population analysis approach, provide insight into electron locali-
BDPA-TXO and BTPA-TXO, the diphenylamine and triphenylamine zation within the molecule and the direction of intramolecular
groups were the donor part while the carbonyl group belonging to charge transfer. Thus, the carbonyl group which accepted 0.19
thioxanthone served as the acceptor part. Consequently, the excita- electron upon photoexcitation is expected to play an important role
tion of the ICT state induced a charge shift from the donor to the in the photo-induced charge transfer process; perturbations on the
acceptor part, and the estimated dipole moments of the first oxygen atom (such as formation of hydrogen bonds with surround-
excited states of BDPA-TXO and BTPA-TXO are 17.1 and 29.6 Debye, ing protic solvent) are expected to result in considerable change in
respectively (Table S2, ESI†), which are sufficiently large to cause emission spectra.
significant solvatochromism according to previous studies.22 This
The experimental response curves for water content in four
dipole moment interacted with the polar solvent molecules to different organic solvents including dioxane, tetrahydrofuran,
reduce the energy of the excited state. As the solvent polarity dimethyl formamide, and acetonitrile were investigated. Owing
increased, this interaction became stronger, resulting in emission to the similar curve trend of the above four solvents, herein we
at lower energies or longer wavelengths. TDDFT calculations also take tetrahydrofuran for example to discuss the spectral changes
suggest that the vertical transitions at optimized S1 geometry, induced by the presence of water. In Fig. S11 (ESI†), the absorp-
which are of comparable energies to the observed emission wave- tion spectra of BDPA-TXO and BTPA-TXO in THF solution under-
lengths in experiments, take place between the highest occupied went slight changes from pure THF to THF containing 10% (v/v)
water, indicating that neither solvent polarity nor hydrogen
bonding had significant influence on the absorption spectra.
In contrast, upon addition of increasing concentration of water
in THF solution, the characteristic emission bands of BDPA-TXO
and BTPA-TXO were effectively quenched along with bathochromic
shifts from 509 nm to 542 nm and 519 nm to 564 nm, respectively
(Fig. S12, ESI†). These phenomena can be attributed to two main
aspects: general solvent effects, increasing addition of water changed
Fig. 1 Fluorescence photographs of (a) BDPA-TXO and (b) BTPA-TXO in organic
solvents of different polarity. Hex, hexane; THF, tetrahydrofuran; EA, ethyl
solvent polarity to some extent and caused spectral shifts; specific
solvent–fluorophore interactions, following excitation there was an
acetate; DCM, dichloromethane; DMF, dimethylformamide; MeCN, acetonitrile;
Bu, n-butanol; EtOH, ethanol.
c
Chem. Commun.
This journal is The Royal Society of Chemistry 2013

View Article Online
ChemComm
Communication
similar phenomena in the other three solvents (Fig. S19–S21, ESI†).
In a sense, the new sensing system provides a suitable platform to
qualitatively and roughly quantificationally detect low-level water
content in several common organic solvents.
In summary, we have developed two thioxanthone-based fluores-
cent probes whose fluorescence properties exhibited intriguing
dependence on solvent polarity, and their amazing multicoloured
performances in various solvents under illumination with 365 nm
light might prove to be of value as indicators for the empirical
determination of solvent polarity. Furthermore, the probes showed
fluorescence quenching upon addition of water in organic solvents
due to charge redistribution from an intramolecular charge transfer
excited state upon hydrogen bonding interaction with water. The
detection limits for the fluorometric method (low-ppm levels)
Fig. 2 (a) Fluorescence emission intensity of BDPA-TXO (lex = 346 nm) and
BTPA-TXO (lex = 357 nm) as a function of the water content in THF solution
(10.0 mM); dyestuff test papers absorbing a certain amount of (b) BDPA-TXO or
(c) BTPA-TXO and fluorescence photographs of THF solutions (5.0 mL) in the
presence of increasing amounts of water (0%, 0.5%, 2%, 5%, and 10%) after the
dyestuff test papers were soaked in.
increase in charge separation within the molecule inducing the performed using a laboratory-scale fluorescence spectrometer were
oxygen atom to become a stronger acceptor for hydrogen bonds, and found to be very close to the ones for the well-established Karl-
the resultant hydrogen bonding lowered the energy of the excited Fischer method. The fluorescence intensity of the solutions in
state, leading to significant changes in emission spectra. Further- which the dyestuff test paper was soaked also decreased upon
more, when the water content of THF solution was comparatively addition of water. The proposed new sensing system provides an
high, the fluorescence intensity quenched to an extremely low point, alternative and convenient approach for qualitative and quantitative
hence the determination of the emission spectra became uncertain. determination of water content in organic solvents.
Fig. 2a also detailedly exhibits how the measured fluorescence
intensity changed as a function of the water content of THF
solution. As can be seen from the graph, the fluorescence intensity
decreased dramatically below 1.00% (v/v): there existed reduction in
relative fluorescence intensity of nearly 45% for BDPA-TXO and 80%
Notes and references
1 (a) C. Reichardt, Chem. Rev., 1994, 94, 2319; (b) C. Reichardt, Org.
Process Res. Dev., 2007, 11, 105.
2 N. S. Bayliss, J. Chem. Phys., 1950, 18, 292.
for BTPA-TXO, whereas the emission band quenched moderately
when the water content was higher than 1.00% (v/v).
3 J. R. Lakowicz, Principles of fluorescence spectroscopy, Springer
Science+Business Media, New York, 3rd edn, 2006, ch. 6, pp. 205–215.
4 R. M. Hermant, N. A. C. Bakker, T. Scherer, B. Krijnen and J. W.
Verhoeven, J. Am. Chem. Soc., 1990, 112, 1214.
5 G. A. Jeffrey, An introduction to hydrogen bonding, Oxford University
Press, New York, 1997.
6 G. J. Huang, J. H. Ho, C. Prabhakar, Y. H. Liu, S. M. Peng and
J. S. Yang, Org. Lett., 2012, 14, 5034.
7 M. H. Abraham, Chem. Soc. Rev., 1993, 22, 73.
8 P. M. Pihko, Angew. Chem., Int. Ed., 2004, 43, 2062.
9 F. A. S. Chipem, A. Mishra and G. Krishnamoorthy, Phys. Chem.
Chem. Phys., 2012, 14, 8775.
10 (a) Z. Q. Guo, W. H. Zhu and H. Tian, Chem. Commun., 2012, 48, 6073;
(b) L. Ding, Q. Zou, Y. Qu and J. H. Su, RSC Adv., 2012, 2, 4754.
11 (a) D. F. Perepichka, I. F. Perepichka, M. R. Bryce, N. I. Sokolov and
A. J. Moore, J. Mater. Chem., 2001, 11, 1772; (b) Y. M. Yang, Q. Zhao,
W. Feng and F. Y. Li, Chem. Rev., 2013, 113, 192.
What is more, the fluorescence intensity as a function of water
concentration showed a near-linear relationship below 1.00% (v/v),
as a consequence, at this low-level water concentration the quench-
ing of the fluorescence intensity due to the presence of water in THF
solution was described by a linear equation, so the detection limit
was estimated from the slope of the linear equation on the basis of
3 times the standard deviation as measured for the solution of the
indicator in the pure solvent (Fig. S13, ESI†).
The response curves including both absorption and fluorescence
spectra for water content in the other three different solvents
(dioxane, dimethyl formamide, and acetonitrile) showed a similar
trend to that in tetrahydrofuran (Fig. S14–S18, ESI†), and the
detection limits are summarized in Table S4 (ESI†). From represen-
12 Y. Y. Wu, Y. Chen, G. Z. Gou, W. H. Mu, X. J. Lv, M. L. Du and
W. F. Fu, Org. Lett., 2012, 14, 5226.
13 M. M. F. Choi and O. L. Tse, Anal. Chim. Acta, 1999, 378, 127.
tative data given in Table S5 (ESI†), it can be seen that the content of 14 K. Fischer, Angew. Chem., 1935, 48, 394.
15 (a) A. Fong and G. M. Hieftje, Anal. Chem., 1995, 67, 1139;
water in THF samples as determined by BDPA-TXO was in good
agreement with standard concentrations with a maximum relative
error less than 5%.
(b) M. S. Baptista, C. D. Tran and G. H. Gao, Anal. Chem., 1996,
68, 971; (c) A. Hakonen, Anal. Chem., 2009, 81, 4555.
16 (a) Q. Chang, Z. Murtaza, J. R. Lakowicz and G. Rao, Anal. Chim.
¨
Acta, 1997, 350, 97; (b) A. Hakonen and N. Stromberg, Analyst, 2012,
To develop an effective method for the qualitative detection of
water in various solvents, some rectangular dyestuff test papers
which absorbed a small amount of fluorescence probe were pre-
pared, and they were then put into the organic solutions which
contained various contents of water. Here, we also take tetrahydro-
furan as a representative. As shown in Fig. 2b and c, immediately
after the dyestuff test papers were soaked in the THF solutions with
different contents of water, the colour of the solutions changed
dramatically under illumination with 365 nm light: with the
increase in the water content, the fluorescence of the samples
exhibited obvious red shifts along with a significant decrease in
the fluorescence intensity. Analogous to THF solutions, it showed
137, 315.
17 H. Mishra, V. Misra, M. S. Mehata, T. C. Pant and H. B. Tripathi,
J. Phys. Chem. A, 2004, 108, 2346.
18 X. Yang, C. G. Niu, Z. J. Shang, G. L. Shen and R. Q. Yu, Sens.
Actuators, B, 2001, 75, 43.
19 C. M. Rushworth, Y. Yogarajah, Y. Zhao, H. Morgan and C. Vallance,
Anal. Methods, 2013, 5, 239.
20 Y. Ooyama, A. Matsugasako, Y. Hagiwara, J. Ohshita and Y. Harima,
RSC Adv., 2012, 2, 7666.
ˇ
21 (a) S. Ishihara, J. Labuta, T. Sikorsk´y, J. V. Burda, N. Okamoto,
H. Abe, K. Ariga and J. P. Hill, Chem. Commun., 2012, 48, 3933;
(b) A. Hakonen and N. Stromberg, Chem. Commun., 2011, 47, 3433.
22 (a) P. Pasman, F. Rob and J. W. Verhoeven, J. Am. Chem. Soc., 1982,
104, 5127; (b) L. Yang, X. Li, J. B. Yang, Y. Qu and J. L. Hua, ACS Appl.
Mater. Interfaces, 2013, 5, 1317.
¨
c
This journal is The Royal Society of Chemistry 2013
Chem. Commun.