Tuning the pKa of Fluorescent Rhodamine pH Probes through Substituent Effects

Article abstract of DOI:10.1002/chem.201703176

Rhodamine spirolactams (RSLs) have recently emerged as popular fluorescent pH probes due to their fluorescence turn-on capability and ease of functionalization at the spirolactam nitrogen. Design of RSLs is often driven by biological targeting or compatibility concerns, rather than the pH sensitivity of the probe, and the relationship between RSL structure and pK_a is not well understood. To elucidate the relationship between pK_a values and the properties of substituents attached to the spirolactam nitrogen, a series of 19 aniline-derived RSLs is presented. RSLs derived from di-ortho-substituted anilines exhibit pK_a tunability across the moderately acidic region (ca. pH 4–6). Evaluation of pK_a data using the Fujita–Nishioka model for ortho substituent effects reveals that both steric and electronic substituent properties influence RSL pH responsiveness, with pK_a values increasing as substituent size and electron withdrawing character increase. These trends are attributed to changes in the RSL structure induced by large substituents, and to electronic influences on the protonated spirocyclic reaction intermediate. To demonstrate the practical applicability of these probes in completely aqueous environments, RSL-doped conjugated polymer nanoparticles that exhibit a ratiometric fluorescence response to changing pH levels are presented.

Full text of DOI:10.1002/chem.201703176

A Journal of
Accepted Article
Title: Tuning the pKa of Fluorescent Rhodamine pH Probes via
Substituent Effects
Authors: Sarah G. Stratton, Grace H. Taumoefolau, Grace E. Purnell,
Mona Rasooly, William L. Czaplyski, and Elizabeth J. Harbron
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: Chem. Eur. J. 10.1002/chem.201703176
Link to VoR: http://dx.doi.org/10.1002/chem.201703176
Supported by

10.1002/chem.201703176
Chemistry - A European Journal
Tuning the pKa of Fluorescent Rhodamine pH Probes via Substituent Effects
Sarah G. Stratton,[a] Grace H. Taumoefolau,[a] Grace E. Purnell,[a] Mona Rasooly,[a]
William L. Czaplyski,[a] and Elizabeth J. Harbron*[a]
Abstract: Rhodamine spirolactams (RSLs) have recently emerged as popular fluorescent pH
probes due to their fluorescence turn-on capability and ease of functionalization at the
spirolactam nitrogen. Design of RSLs is often driven by biological targeting or compatibility
concerns rather than the pH sensitivity of the probe, and the relationship between RSL structure
and pKa is not well-understood. Here we present a series of 19 aniline-derived RSLs designed to
elucidate the relationship between pKa values and the properties of substituents attached to the
spirolactam nitrogen. RSLs derived from di-ortho-substituted anilines exhibit pKa tunability
across the moderately acidic region (ca. pH 4 – 6). Evaluation of pKa data using the Fujita-
Nishioka model for ortho substituent effects reveals that both steric and electronic substituent
properties influence RSL pH responsiveness, with pKa values increasing as substituent size and
electron withdrawing character increase. These trends are attributed to changes in the RSL
structure induced by large substituents and to electronic influences on the protonated spirocyclic
reaction intermediate. To demonstrate the practical applicability of these probes in completely
aqueous environments, we present RSL-doped conjugated polymer nanoparticles, which exhibit
a ratiometric fluorescence response to changing pH levels.
[a] S.G. Stratton, G.H. Tamoefolau, G.E. Purnell, M. Rasooly, W.L. Czaplyski, Prof. E.J. Harbron
Department of Chemistry
The College of William and Mary
Williamsburg, VA 23187-8795 (USA)
E-mail: ejharb@wm.edu
This article is protected by copyright. All rights reserved.

10.1002/chem.201703176
Chemistry - A European Journal
Introduction
Accurate measurement of pH is crucial to understanding the many chemical and
biological processes that are influenced by proton concentration. Microelectrodes, magnetic
resonance techniques, and optical spectroscopy are among the methods routinely used to
measure pH.[1] From this group, fluorescence spectroscopy has emerged as the method of choice
for samples with spatial variation in pH.[2] Cells exemplify this type of sample: the pH required
for normal function ranges from alkaline (pH 8.0) in mitochondria to near-neutral in the cytosol
to acidic in lysosomes (pH 4.7).[3] An exquisite balance of biochemical processes is required to
maintain cellular pH levels,[3] and deviations from normal pH levels are associated with diseases
such as cancer.[4] Fluorescence can report on cellular pH values nondestructively and with high
spatiotemporal resolution and sensitivity.[2] Measurement of pH by fluorescence spectroscopy
utilizes probes with emission properties that are strongly pH-dependent.[2, 5] Based on their pKa
values, fluorescent pH probes for intracellular measurements are generally divided into those
useful in near-neutral (ca. pH 7) and acidic (pH 4.5-6) environments. Although probes that are
sensitive to acidic environments are critical for tracking processes within acidic organelles, fewer
have been developed compared to those for near-neutral environments.[6]
Rhodamine spirolactams (RSLs) show remarkable promise as fluorescent pH probes for
acidic environments. As shown in Scheme 1 for a generic rhodamine B derivative, RSLs exhibit
a pH-dependent equilibrium between two forms, the colorless and non-fluorescent spirolactam
and the colored and fluorescent rhodamine amide. Off-to-on colorimetric and fluorescence
switching is observed as the RSLs’ environment becomes more acidic, triggering opening of the
spirocyclic ring system to reveal the fully conjugated rhodamine dye. The fluorescent form of
the RSL possesses all of the desirable photophysical characteristics of its parent rhodamine dye,
including excitation and emission in the visible region of the spectrum, high extinction
coefficient, high fluorescence quantum yield, and good photostability.[7]
Due to these outstanding photophysical properties, RSL-based pH probes have
proliferated in the last few years and have been featured in reviews.[6, 8] Recent examples include
RSLs for the single-wavelength monitoring of pH fluctuations in lysosomes[9] and other
biological environments,[10] an RSL-fluorescein construct for wide range pH monitoring,[11] and
RSL-containing dyads for ratiometric pH measurements.[12] Acid-responsive RSLs have also
been incorporated into macromolecular systems for pH sensing,[13] tumor imaging,[14] and
2
This article is protected by copyright. All rights reserved.

10.1002/chem.201703176
Chemistry - A European Journal
anticancer drug release.[15] The pKa values of RSLs vary with the spirolactam R group, with
many rhodamine B-derived probes intended for acidic pH monitoring falling in the pH 4-5 range.
Selection of the R group tends to be driven by concerns such as organelle targeting rather than
the goal of obtaining a specific pKa value. While this approach addresses biological
compatibility, it does not enable the targeting of specific processes for pH monitoring by RSLs.
Towards this end, we seek a systematic understanding of the structure-pKa relationship for RSLs.
Substituent effects on chemical reactivity can be driven by steric or electronic
interactions or a combination of both. Both steric and electronic effects have previously been
cited as influences on the pH-responsiveness of RSLs and related compounds. Yuan, et al. found
that RSLs with sterically demanding spirolactam R groups had higher pKa values than those with
smaller substituents.[16] In our own work, identical pKa values were observed across a series of
RSLs derived from para-substituted anilines with different substituents.[17] However, one
derivative prepared with a bulkier 2,4,6-substituted aniline exhibited a much higher pKa value,
supporting the finding of Yuan. These results provide anecdotal evidence that the pH-
responsiveness of RSLs can be tuned by controlling steric bulk near the spirolactam nitrogen.
This finding is in contrast to the behavior of rhodamine deoxy-lactams, RSL derivatives in which
the spirolactam carbonyl moiety has been removed. The pH-responsiveness of the deoxy-
lactams is correlated with the basicity of amine functionality, and both electronic[18] and steric[19]
effects on the basicity have been documented.
We seek to establish a structure-property relationship for the pKa values of RSL pH
probes for acidic environments. Here we present a series of aniline-derived RSLs in which the
substituents near the spirolactam nitrogen are varied in position and steric and electronic
properties. Upon establishing the tunability of pKa values, we employ simple and multiple linear
regression models to explore the origin of their variance. We also demonstrate the practical
utility of RSLs for ratiometric pH sensing in a completely aqueous environment by preparing
RSL-doped conjugated polymer nanoparticles.
Results and Discussion
RSL derivatives 1-19 (Table 1) were designed to probe the influence of the spirolactam R
group on the spirocyclic ring-opening reaction. To facilitate variation of substituent steric and
electronic properties, the R group was derived from substituted anilines and is denoted Ar in
3
This article is protected by copyright. All rights reserved.

10.1002/chem.201703176
Chemistry - A European Journal
Scheme 2. The rhodamine scaffold of RSL-based pH probes is commonly derived from
rhodamine B (RB) or rhodamine 6G (R6G), which differ in the substitution pattern of the
xanthene ring system (Scheme 2) and in their fluorescent properties. To address hypotheses
generated by our previous work on related derivatives,[17] we derived most of our compounds (1-
13) from RB, which fluoresces yellow-orange. With our long-term interest in fluorescence
resonance energy transfer (FRET), RB derivatives are also attractive as acceptor dyes due to
their low energy fluorescence.[20] Additionally, we prepared a smaller series of compounds
based on R6G (14-19), which emits green fluorescence with a much higher quantum yield than
RB.[21] The additional methyl substituents on R6G’s xanthene moiety facilitate exploration of
the combined effects of steric hindrance in the spirolactam substituent and the xanthene core.
Compounds 1-19 were prepared from the desired aniline and RB or the carboxylic acid form of
R6G via the rhodamine acid chloride. Structures of the RSLs were verified by high resolution
mass spectrometry and 1H and 13C NMR. All compound exhibited the characteristic resonance
of the spiro carbon in the 13C NMR spectrum (67.5 – 70.3 ppm for 1-19), verifying their
spirocyclic structures.
Consistent with other RSL derivatives, compounds 1-19 are all colorless and non-
fluorescent in their as-prepared spirocyclic forms, which absorb only in the ultraviolet with peaks
ca. 275 and 315 nm. Accordingly, they exhibit no absorbance or fluorescence in the visible
region of the spectrum. Addition of acid opens the spirolactam moiety to reveal the highly
colored and fully conjugated rhodamine dye structure (Scheme 1). The increases in visible
absorbance and fluorescence intensity upon HCl addition are shown in Figure 1 for
representative RB derivative 9, which has max,abs and max,fl values of 561 and 581 nm,
respectively. These values are typical for RB derivatives 1-13, which all have spectral maxima
within 561-563 nm for absorbance and 581-585 nm for fluorescence (Table S1). R6G
derivatives have similarly clustered values with ranges 532-533 nm and 551-555 nm for
absorbance and fluorescence, respectively. For both RB and R6G derivatives, there is no
consistent trend in max,abs and max,fl values with respect to the steric and electronic properties of
the aniline substituents.
We divided the RSLs into 4 overlapping series to facilitate comparison of pH-
responsiveness among the different derivatives. Series 1, comprised by RB derivatives 1-4 and
10, was designed to elucidate the aniline substitution pattern required to induce steric effects on
4
This article is protected by copyright. All rights reserved.

10.1002/chem.201703176
Chemistry - A European Journal
the pKa. The pKa value for each compound was determined via fluorescence titration with HCl.
From fluorescence spectra recorded after each addition of acid, the fluorescence intensity at the

max,fl was extracted and plotted as a function of pH to construct a titration curve. The pKa is
then determined by calculating the pH at which the fluorescence intensity is half of its maximum
value. Figure 2 shows representative titration curves for Series 1 compounds, all of which
contain 1 or 2 chlorine substituents in different positions relative to the aniline nitrogen, which is
part of the spirolactam ring in the closed form of the RSL. Substitution of Cl at the para
position (1), both meta positions (3), or a single ortho position (10) gave nearly
indistinguishable pKa values near 4.3, which is also the value obtained for an unsubstituted
control (5). In contrast, chlorine substitution of both ortho positions (4) or both ortho positions
with a para-nitro substituent (2) gave significantly shifted titration curves with pKa values that
are identical within experimental error at ca. 5.4. Series 1 compounds demonstrate that both
ortho positions of the aniline-derived ring must be substituted to effect a significant shift in the
pKa from “baseline” values for aniline-derived RSLs, shown here to be ca. 4.3.
To determine the extent to which di-ortho substitution can be used to tune the pKa of
aniline-derived RSLs, we prepared additional di-ortho-substituted derivatives based on RB
(Series 2, 4-9) and R6G (Series 3, 14-19). Substituents were selected from commercially
available di-ortho-substituted anilines to cover a range of steric and electronic substituent
properties. Titration curves for RB Series 2 compounds (Figure 3A) are distributed across the
acidic region, demonstrating a clear substituent dependence. The pKa values extracted from
these curves range from a low of 4.03 (7 OEt) to a high of 5.72 (9 iPr), a span of 1.7 pH units
(Table 1). R6G Series 3 titration curves (Figure 3B) span a slightly greater range of values than
their RB Series 2 counterparts, from a low of 3.81 (15 H) to a high of 5.87 (19 iPr), a difference
of 2.1 pH units. Unlike the fairly even distribution of pKa values observed for Series 2, Series 3
pKa values are more clustered, with the compounds having the highest pKa values (14, 18, 19)
narrowly distributed around an average of 5.60. We expected that R6G Series 3 would yield
higher pKa values than RB Series 2 due to steric hindrance from the additional methyl group on
the xanthene core of R6G. Indeed, identical substituents yielded higher pKa values in R6G than
RB for four of the six compounds, with the greatest difference between di-ortho-methyl-
substituted 8 and 18 (pKa = 0.42). This trend is consistent with the work of Yuan, et al., who
observed an even larger pKa difference between R6G and RB derivatives in which the
5
This article is protected by copyright. All rights reserved.

10.1002/chem.201703176
Chemistry - A European Journal
spirolactam nitrogen was functionalized with an adamantyl group.[16] The Series 2 and 3
compounds demonstrate that we can use substituent effects to manipulate the pKa across the
acidic region.
Our ultimate goal is to define a structure-property relationship for the pH-responsiveness
of RSLs. Linear free energy relationships (LFERs) can supply a quantitative understanding of
how substituent properties influence a reaction or molecular property. To test our initial
hypothesis that increasing steric hindrance at the spirolactam nitrogen would increase the pKa of
RSLs, we examined univariate correlations of our Series 2 and 3 pKa data with standard steric
parameters. The steric parameters[22] included in this analysis were A-values (Winstein-Holness
values), Charton’s , Meyer’s Va, interference values, Sterimol B1 values, and Taft ES values[23]
as modified by Kutter and Hansch.[24] The extent of correlation between the pKa values and
substituent steric properties varied considerably depending on the steric parameter used. Three
parameters – Sterimol B1, interference values, and ES – yielded correlations for which the r2
value was greater than 0.6 for both Series 2 and 3 data. The strongest correlation was observed
for Es (Figure 4), which is correlated with substituents’ van der Waals radii.[24] The Es-pKa
correlation was more significant and had a steeper slope for R6G Series 3 than for RB Series 2,
suggesting that steric effects might be more dominant for R6G compounds. However, the
significant scatter in Figure 4 indicates that steric properties alone cannot account for the
observed range of pKa values in Series 2 and 3.
The scatter in the univariate correlations suggests that both electronic and steric
substituent properties may influence the pKa values. Several LFERs have been developed for the
treatment of data for ortho-substituted compounds, which may have steric, inductive, and
resonance effects that are distinct from those with more distant meta and para substituents.
Building on the work of Taft[23] and Charton,[25] Fujita and Nishioka developed a 3-parameter
LFER for ortho effects.[26] Their original model divides the effect of ortho substituents into
contributions from ordinary polar, proximity polar, and steric effects to yield Equation 1, where
o, F, and Es are substituent parameters, and , f, and  are the respective susceptibility
constants.[26] The ordinary polar effect encompasses both inductive and resonance effects and is
log 푘표 = ꢀꢁ표 + 푓퐹 + 훿퐸푠 + 푐
(1)
6
This article is protected by copyright. All rights reserved.

10.1002/chem.201703176
Chemistry - A European Journal
given by o ≡ p, the standard Hammett constant. The Swain-Lupton field parameter F [27] is
used to correct for the greater inductive and field effects of ortho over para substituents, and the
Taft-Kutter-Hansch Es values are used as the steric parameter. Alternatively, the Fujita-Nishioka
model can be reformulated[26] to express electronic effects as a linear combination of field and
resonance effects as shown in Equation 2, where R is the Swain-Lupton-Hansch resonance
log 푘표 = 푓퐹 + 푟푅 + 훿퐸푠 + 푐
(2)
parameter[27-28] and f, r, and  are the susceptibility constants for the field, resonance, and steric
parameters. Eqs. 1 and 2 have successfully accounted for substituent effects on kinetics,
thermodynamics, and chemical shifts.[26]
Multiple linear regressions of Series 2 and 3 data were performed according to Eqs. 1 and
2 using literature values for , Es, F, and R (Table 2) with the value of each substituent constant
doubled to account for the di-ortho substitution of these compounds.[29] The Fujita-Nishioka
models given by Eqs. 1 and 2 yielded significant correlations, explaining 95% and 98% of the
variance in pKa values for RB and R6G derivatives, respectively. As shown in Eqs. 3A (RB) and
3B (R6G), regression by Eq. 1 curiously yields electronic coefficients with opposite signs. The
푝퐾푎 = 1.25ꢁ푝 − 0.44퐹 − 0.48퐸푠 + 4.28 (R2 = 0.953)
푝퐾푎 = 1.05ꢁ푝 − 0.21퐹 − 0.72퐸푠 + 3.82 (R2 = 0.983)
(3A)
(3B)
positive value for  indicates that pKa values increase as the electron withdrawing nature of the
substituent increases, but the negative value for f indicates the opposite trend. This result
suggests that regression by Eq. 1 may overestimate the influence of inductive and field effects
and underestimate the contribution of resonance effects to the pKa values. Indeed, regression by
Eq. 2 with separate field and resonance parameters yields coefficients with the same sign that
paint a more consistent picture, as shown in Eqs. 4A (RB) and 4B (R6G). Regression according
to Charton’s model, which also uses separate parameters for inductive and resonance effects,
gives extremely similar results with high significance (Figure S3), validating the use of Eq. 2 for
our system.
푝퐾푎 = 0.72퐹 + 1.26푅 − 0.47퐸푠 + 4.28 (R2 = 0.953)
푝퐾푎 = 0.78퐹 + 1.06푅 − 0.72퐸푠 + 3.82 (R2 = 0.983)
(4A)
(4B)
Figure 5 depicts the excellent agreement between experimental and predicted values for
regression of Series 2 and 3 by Eq. 2. To evaluate the relative contributions of field, resonance,
and steric effects to the pKa values, we determined the standardized regression coefficients and
7
This article is protected by copyright. All rights reserved.

10.1002/chem.201703176
Chemistry - A European Journal
their ratios. For RB Series 2, the standardized coefficients (0.30F, 0.36R, -0.54Es) demonstrate
that steric and electronic effects have roughly equal effects on pKa values, with Es accounting for
45% of the variance in pKa and electronic effects split roughly evenly among field (25% F) and
resonance (30% R). As expected due to the bulkier structure of the R6G xanthene core, the
standardized coefficients for R6G Series 3 (0.33F, 0.30R, -0.81Es) reveal a greater influence of
steric effects, with 56% of the variance in pKa due to Es, 23% to F, and 21% to R. For both
Series 2 and 3 compounds, regression by Eq.2 demonstrates that a high pKa is promoted by
substituents that are large and electron-withdrawing by induction and resonance.
Many LFER analyses,[30] including recent applications of the Fujita-Nishioka model,[31]
treat polysubstituted aromatics with the assumption that substituent effects are additive. To
determine whether this assumption is valid for our compounds and gain further insight into the
structure-pKa correlation, we prepared a limited set of mono-ortho-substituted RB derivatives
(Series 4, 10-13). The pKa values of Series 4 compounds (Table 1) are tightly clustered and
identical within experimental error, with the exception of one slightly higher value (10, Cl). A
multiple linear regression performed on the combined set of Series 2 and 4 pKa values using Eq.
2 (Figure S1) failed to predict the nearly identical Series 4 values and was not as successful at
predicting Series 2 values as when these were treated separately. This result suggests that
substituent effects in the Series 2 di-ortho compounds are much stronger than would be expected
based on their mono-ortho counterparts and the assumption of additive effects. The extreme
steric crowding in di-ortho-substituted aromatics has previously been shown to cause
anomalously high reactivity compared to mono-ortho compounds in a kinetic study of Z-
arylhydrazones of oxadiazoles by Spisani and coworkers.[31b] Here, the di-ortho-substituted
compounds exhibit substituent effects on their pKa values that likewise exceed those expected by
simple additivity.
To understand the structural basis for the variance in pKa values, we attempted to obtain
crystal structures of the di-ortho-substituted compounds. Of the Series 2 and 3 compounds, we
were able to grow high quality crystals only for 9, the RB di-o-iPr derivative. The obtained
crystal structure for 9 shows a remarkable deformation of the xanthene ring system, which is
generally expected to be planar. As shown in Figure 6, the xanthene system in 9 is deformed
into a shape reminiscent of butterfly wings, with the outside rings of the xanthene bent away
from the bulky iPr substituents of the aniline-derived substituent. We quantified the xanthene’s
8
This article is protected by copyright. All rights reserved.

10.1002/chem.201703176
Chemistry - A European Journal
deviation from planarity by averaging the C13-O1-C1-C2 and C1-O1-C13-C12 dihedral angles
to obtain a value of 154°. The crystal structure for 5 (RB H) was previously reported in the
literature[32] and exhibits an average xanthene dihedral of 174°, which is much closer to the
expected planar structure. This result suggests that deformation of the xanthene ring structure
may be required for high pKa values such as that observed for 9.
To test the deformation-pKa hypothesis, we used Gaussian[33] to calculate optimized
geometries for Series 2 compounds at the HF/6-311(d,p) level, which was selected because it
yielded structures with dihedral angles that matched those obtained from crystal structures for 5
and 9. The calculated xanthene dihedral angles (Table 2) decrease (i.e., become more deformed)
as substituent size increases. These values show good correlation with the Es steric parameter (r2
= 0.872, Figure S2), supporting the idea the RSL structure becomes more deformed and more
highly strained as the di-ortho substituents increase in size. We found 6 additional literature
crystal structures for RB-based spirolactams derived from mono-ortho-substituted anilines.[34]
The xanthene dihedral angle was >170° for all of these compounds except for one with an o-SMe
substituent[34c] that had a slightly more deformed dihedral angle (166°). Thus, it appears that di-
ortho substitution is required to obtain the most deformed structures (dihedral < 160°), which are
those with the highest pKa values.
The effect of the xanthene deformation and substituent electronics on the pKa can be
rationalized in conjunction with the mechanism shown in Scheme 1. Protonation of the closed
form is expected to be rapid and at equilibrium prior to the ring-opening reaction.[35] Thus,
differences in closed-open equilibrium are likely driven by the relative thermodynamic stabilities
of the protonated intermediate and the fluorescent forms of the dye, which include two tautomers
that are indistinguishable by fluorescence. With its delocalized positive charge, the protonated
spirocyclic intermediate is destabilized by substituents in the spirolactam R group that are
electron withdrawing by induction or resonance. Stephenson and coworkers previously observed
the same electronic effect on the thermodynamic stabilities of a series of pyridine-substituted
RSLs, noting that the equilibrium shifted to the open form of the dye as the electron-withdrawing
nature of the R group increased.[36] The protonated intermediate for our di-o-substituted
compounds would be additionally destabilized by deformation of its xanthene ring structure as
described above. Both the electronic and structural forms of destabilization are expected to be
relieved by the spirocyclic ring-opening reaction, which yields a less sterically congested product
9
This article is protected by copyright. All rights reserved.

10.1002/chem.201703176
Chemistry - A European Journal
in which the positive charge is more extensively delocalized. To probe the effect of substituents
that are both electronically and sterically destabilizing, we attempted to synthesize the RB di-o-
NO2 derivative, which has a predicted pKa value of 8.04 according to Eq. 4A, the regression
equation for RB di-ortho-substituted compounds. However, our efforts were unsuccessful,
perhaps because the spirocyclic form is simply too destabilized for the desired product to form.
Having demonstrated that we can vary the pKa of RSLs in systematic fashion via
substituent effects, we next set out to assess the general suitability of these RSLs for practical
applications. The utility of RSL dyes is often limited by two key features, solubility and single-
color emission. Most RSLs have negligible water solubility in their closed forms and frequently
require an organic cosolvent. Indeed, 1-19 were titrated in 1:1 v/v EtOH:H2O for this reason.
RSLs like 1-19 can report on the pH of their surrounding medium as demonstrated above, but
this measurement is of fluorescence intensity at a single wavelength. Fluorescence intensity
during such measurements can be affected by factors unrelated to the pH, including differences
in probe concentration and instrumental or environmental fluctuations. Ratiometric
measurements address this problem by using a ratio of fluorescence intensities from two
different fluorophores in place of the single-wavelength measurement. Although ratiometric
measurements are more desirable for this reason, they exacerbate solubility problems because
covalently attaching an additional fluorophore to an RSL further reduces its water solubility in
most cases.
An ideal RSL pH probe would be capable of ratiometric sensing in an aqueous
environment without organic cosolvent. We developed highly fluorescent conjugated polymer
nanoparticles (CPNs) doped with 9 that meet these criteria. CPNs are stably suspended in water,
and those that are non-covalently doped with fluorogenic dyes can be employed as ratiometric
sensors.[37] Undoped CPNs of the conjugated polymer PFBT emit green-yellow fluorescence
(max,fl = 535 nm) that overlaps the absorbance of RSLs like 9 (max,abs = 561 nm), making the
CPNs and 9 a suitable fluorescence resonance energy transfer (FRET) donor-acceptor pair. At
high pH, all 9 dyes will be in their non-fluorescent spirolactam form, and only green-yellow
fluorescence from the CPNs will be observed (Figure 7A). As the pH is lowered into 9’s
responsive range, 9 dyes will begin to convert to their rhodamine forms, activating FRET from
CPN donor chromophores to 9. As more 9 dyes open, CPN fluorescence will decrease, and
orange fluorescence from 9 will increase. In the context of ratiometric pH sensing with RSLs,
10
This article is protected by copyright. All rights reserved.

10.1002/chem.201703176
Chemistry - A European Journal
the CPNs have two key advantages over most small molecule FRET donors: 1) they are stably
suspended in water, obviating the need for an organic cosolvent and 2) they act as light
harvesters,[38] amplifying the fluorescence intensity of the rhodamine acceptor dyes beyond what
would be observed upon direct excitation of the dye. These features have been observed
previously in similar CPNs doped with a pH-sensitive fluorescein derivative.[39]
Doped CPNs with an average diameter of 14.7 +/- 1.5 nm (Figure S4) were titrated to
determine the response of their fluorescence to pH changes. As the pH was lowered,
fluorescence intensity from the CPNs decreased while that of 9 increased, and an isoemissive
point was observed at 559 nm (Figure 7B). These as-expected changes indicate FRET from the
CPNs to 9 as the dyes are converted to their rhodamine form by acid. The fluorescence changes
are quantified by the ratio of fluorescence intensity of 9 at 577 nm to that of the CPNs at 535 nm
as a function of pH. Figure 7C shows ratiometric data from two representative samples, each of
which was titrated from high to low pH with acid and then from low to high pH with base. At pH
<4.7, the CPNs exhibit changes in their absorbance spectra consistent with irreversible structural
changes, so we focused on the pH 4.7 – 10 range. The relationship between the I577/I535 ratio and
pH shows no dependence on the direction of titration, is consistent across the different datasets,
and has the appearance of a partial titration curve. Since the ratio continues to increase at the
low pH end of the titrations, we could not establish a pKa for 9 in the aqueous environment of the
nanoparticles. However, the ratio serves as a tool for pH readout within the 4.7 – 10 range,
similar to the 5.0 – 8.0[39b] and 4.8 – 13[40] ranges given for previous CPN-based pH sensors.
An additional feature of the doped CPNs is their ability to amplify the fluorescence
intensity of 9. CPNs have extremely large extinction coefficients and exhibit exciton diffusion
that funnels FRET donor energy to acceptor dyes with unusually high efficiency.[38] The
resulting amplification of acceptor intensity can be quantified by recording the fluorescence of
each sample twice, once upon excitation of the dyes via FRET from the CPNs (exc = 450 nm)
and again upon direct excitation of the dyes (exc = 535 nm). Figure 7D shows that the
fluorescence intensity of 9 in a representative sample of doped CPNs is 25 times greater when
excited via FRET than when excited directly. This amplification translates to stronger signal
and/or the ability to reduce excitation intensity when using CPNs as pH sensors.
11
This article is protected by copyright. All rights reserved.

10.1002/chem.201703176
Chemistry - A European Journal
Conclusions
We have demonstrated that the pH responsiveness of aniline-derived RSLs can be
systematically tuned via substituent effects. Di-ortho-substituted RB- and R6G-derived pH
probes exhibit pKa values that span the biologically relevant pH 4 – 6 range. LFER analysis
using the Fujita-Nishioka model indicates that 45% (RB) to 56% (R6G) of the variance in pKa is
due to substituent steric effects, with the remainder attributed to inductive and resonance effects.
Increasing the size and electron-withdrawing character of the substituents shifts the pKa to higher
values. Crystal structures and calculated geometries suggest that di-ortho-substitution induces
deformation of the xanthene moiety to an extent that depends on substituent size. We explored
the applicability of RSLs to ratiometric pH sensing in a purely aqueous environment by
preparing RSL-doped conjugated polymer nanoparticles. The doped CPNs yielded a consistent
ratiometric fluorescence response across the pH 4.7 – 10 range and exhibited a substantial
enhancement of rhodamine fluorescence intensity due to the light harvesting effect of the
nanoparticles. These results will guide future efforts to develop RSL-based pH probes with pKa
values targeted to specific processes.
Experimental Section
Absorbance and fluorescence studies. Absorbance and fluorescence measurements were made
on a Varian Cary50 and Varian Eclipse, respectively. Stock solutions of 1-19 were prepared in
1:1 (v/v) ethanol:water at a concentration of 20 M or 2 M (4, 8, 14-19) and were stirred at
room temperature for 1 h prior to absorbance and fluorescence studies. Absorption and
fluorescence titrations were conducted by adding small aliquots of 0.1 M HCl or 0.1 M NaOH to
a stirring stock solution (100 mL) of 1-19. Aliquots were removed for spectroscopic
measurement (exc,fl = 535 nm) after the pH stabilized and were then returned to the stock
solution. Reported pKa values are the average of a minimum of 4 separate titration experiments.
Materials and methods. Reagents were purchased from Acros and used as received except 2,6-
diethoxyaniline, which was synthesized by a literature procedure.[41] Poly[9,9-dioctylfluorenyl-
2,7-diyl)-co-1,4-benzo-{2,1′-3}-thiadiazole)] (PFBT) with an average molecular weight of
138,000 and polydispersity 3.1 was obtained from American Dye Source (Quebec, Canada) and
used as received. Granular poly(vinyl butyral-co-vinyl alcohol-co-vinyl acetate) (PVB-VA-VA)
12
This article is protected by copyright. All rights reserved.

10.1002/chem.201703176
Chemistry - A European Journal
1
was obtained from Sigma-Aldrich. H and proton-decoupled 13C NMR spectra were recorded on
a Varian Mercury 400 (1H 400 MHz, 13C 100 MHz) and referenced to TMS (1H) or CDCl3 at
77.0 ppm (13C). Chemical shifts are reported in ppm and coupling constants in Hz. Mass spectra
were measured through positive electrospray ionization (w/ NaCl) on a Bruker 12 Tesla APEX–
Qe FTICR–MS with and Apollo II ion source.
Conjugated polymer nanoparticle preparation and characterization. Stock solutions of
conjugated polymer PFBT (1 mg/mL) and stabilizing polymer PVB-VA-VA (1 mg/mL) in
anhydrous THF were stirred overnight under argon. A precursor solution was prepared by
combining portions of the PFBT and PVB-VA-VA solutions with additional THF to a final
concentration of 0.040 mg/mL PFBT and 0.016 mg/mL PVB-VA-VA. This precursor solution
was filtered through a 0.7 m filter to remove any aggregates and then sonicated for 30 s to
ensure homogeneity. A portion of the precursor solution (1 mL) was injected into sonicating
ultrapure water (8 mL), and the resulting aqueous suspension of CPNs was sonicated for an
additional 2 min. Argon was bubbled through the CPN suspension on a 65 °C hot plate for 30
min. to remove THF. To dope the CPNs with RSL 9, a solution of 9 in THF (3.6 L, 10 mg/mL)
was injected into the CPN suspension. The aqueous suspension of 9-doped CPNs was then
filtered through a 0.22 m filter to remove any aggregated dyes or CPNs. CPN size distributions
were measured in aqueous suspension by dynamic light scattering using a Nicomp N3000
Submicron Particle Sizer (Particle Sizing Systems), and the size distribution is shown in Figure
S4. Titrations of 9-doped CPNs were conducted by adding small aliquots of 2.5 mM HCl or 5
mM NaOH to a 2.5 mL portion of the aqueous dispersion of CPNs in a quartz cuvette. After the
pH stabilized, absorption and fluorescence spectra were recorded. Ratios of peak fluorescence
intensities were computed using IgorPro 6.05.
General synthetic procedures. Compounds 1, 2, and 5 were reported previously.[17] Detailed
characterization data for compounds 3-4 and 6-19 are given in the Supporting Information. To
prepare compounds 3-4 and 6-13, phosphorus oxychloride (3 mmol) was added dropwise to a
stirred solution of rhodamine B (0.6 mmol) in 1,2-dichloroethane (8 mL). The desired aniline (1
mmol) was added dropwise, and the reaction mixture was refluxed for 24 hours. The solution
was washed with 2 M HCl (2x), 2 M NaOH (3x), and brine (2x). The organic layer was dried
13
This article is protected by copyright. All rights reserved.

10.1002/chem.201703176
Chemistry - A European Journal
with MgSO4 and concentrated in vacuo. The crude product was purified by manual flash column
chromatography.
To prepare compounds 14-19, the carboxylic acid form of rhodamine 6G (0.5 mmol) was
dissolved in anhydrous 1,2-dichlorethane (60 mL) with sonication and added to a flame-dried
round bottom flask equipped with molecular sieves. After stirring overnight under argon, the
solution was placed in an ice bath, and the desired aniline (3 mmol) was added and stirred for 30
min. Phosphorous oxychloride (0.5 mmol) was added dropwise, and the mixture was stirred for
48 hours. The solution was filtered and then washed with chilled 2M NaOH (3x) and brine (1x).
After drying with Na2SO4, the solution was concentrated in vacuo. The crude product was
purified by manual flash column chromatography.
Acknowledgements
We gratefully acknowledge support of this work by the Camille and Henry Dreyfus Foundation
through a Henry Dreyfus Teacher-Scholar Award; the National Science Foundation (CHE-
1464699); and the Charles Center of the College of William and Mary via student fellowships
(G.H.T., S.G.S, M.R.). We thank Prof. Robert D. Pike of the College of William and Mary for
the x-ray crystal structure.
References
[1]
[2]
[3]
[4]
F. B. Loiselle, J. R. Casey, Methods Mol. Biol. 2010, 637, 311-331.
J. Han, K. Burgess, Chem. Rev. 2010, 110, 2709-2728.
J. R. Casey, S. Grinstein, J. Orlowski, Nat Rev Mol Cell Biol 2010, 11, 50-61.
B. A. Webb, M. Chimenti, M. P. Jacobson, D. L. Barber, Nat Rev Cancer 2011, 11, 671-
677.
[5]
a) J. Yin, Y. Hu, J. Yoon, Chem Soc Rev 2015, 44, 4619-4644; b) Y. Ni, J. Wu, Org
Biomol Chem 2014, 12, 3774-3791; c) N. Boens, V. Leen, W. Dehaen, Chem Soc Rev
2012, 41, 1130-1172; d) R. Wang, C. Yu, F. Yu, L. Chen, C. Yu, TrAC, Trends Anal.
Chem. 2010, 29, 1004-1013.
[6]
[7]
[8]
X. Chen, T. Pradhan, F. Wang, J. S. Kim, J. Yoon, Chem. Rev. 2012, 112, 1910-1956.
M. Beija, C. A. M. Afonso, J. M. G. Martinho, Chem. Soc. Rev. 2009, 38, 2410-2433.
Z. Yang, J. Cao, Y. He, J. H. Yang, T. Kim, X. Peng, J. S. Kim, Chem. Soc. Rev. 2014,
43, 4563-4601.
[9]
a) H.-S. Lv, S.-Y. Huang, Y. Xu, X. Dai, J.-Y. Miao, B.-X. Zhao, Bioorg. Med. Chem.
Lett. 2014, 24, 535-538; b) H. Li, C. Wang, M. She, Y. Zhu, J. Zhang, Z. Yang, P. Liu,
Y. Wang, J. Li, Anal. Chim. Acta 2015, 900, 97-102; c) X.-L. Shi, G.-J. Mao, X.-B.
Zhang, H.-W. Liu, Y.-J. Gong, Y.-X. Wu, L.-Y. Zhou, J. Zhang, W. Tan, Talanta 2014,
14
This article is protected by copyright. All rights reserved.

10.1002/chem.201703176
Chemistry - A European Journal
130, 356-362; d) K.-K. Yu, K. Li, J.-T. Hou, H.-H. Qin, Y.-M. Xie, C.-H. Qian, X.-Q.
Yu, RSC Adv. 2014, 4, 33975-33980; e) H. Li, H. Guan, X. Duan, J. Hu, G. Wang, Q.
Wang, Org. Biomol. Chem. 2013, 11, 1805-1809.
[10] a) J.-L. Tan, F. Zhang, T.-T. Yang, Y. Liu, M.-X. Zhang, Z.-B. Li, H. Zuo, Spectrochim
Acta A Mol Biomol Spectrosc 2015, 140, 489-494; b) Z.-Q. Hu, M. Li, M.-D. Liu, W.-M.
Zhuang, G.-K. Li, Dyes Pigm. 2013, 96, 71-75; c) M. Devi, A. Dhir, C. P. Pradeep, Eur.
J. Org. Chem. 2015, 2015, 4650-4657; d) A. Liu, M. Hong, W. Yang, S. Lu, D. Xu,
Tetrahedron 2014, 70, 6974-6979; e) M.-C. Xia, L. Cai, S. Zhang, X. Zhang, Anal.
Chem. 2017, 89, 1238-1243; f) H. Yu, G. Li, B. Zhang, X. Zhang, Y. Xiao, J. Wang, Y.
Song, Dyes Pigm. 2016, 133, 93-99; g) Y. Xue, W. Liang, Y. Li, Y. Wu, X. Peng, X. Qiu,
J. Liu, R. Sun, J. Agric. Food. Chem. 2016, 64, 9592-9600.
[11] M. H. Lee, J. H. Han, J. H. Lee, N. Park, R. Kumar, C. Kang, J. S. Kim, Angew. Chem.,
Int. Ed. 2013, 52, 6206-6209.
[12] a) X.-F. Zhang, T. Zhang, S.-L. Shen, J.-Y. Miao, B.-X. Zhao, RSC Advances 2015, 5,
49115-49121; b) U. Reddy G, A. H. A, F. Ali, N. Taye, S. Chattopadhyay, A. Das, Org.
Lett. 2015, 17, 5532-5535; c) N. I. Georgiev, A. M. Asiri, A. H. Qusti, K. A. Alamry, V.
B. Bojinov, Dyes Pigm. 2014, 102, 35-45; d) J. Fan, C. Lin, H. Li, P. Zhan, J. Wang, S.
Cui, M. Hu, G. Cheng, X. Peng, Dyes Pigm. 2013, 99, 620-626; e) S.-L. Shen, X.-F.
Zhang, S.-Y. Bai, J.-Y. Miao, B.-X. Zhao, RSC Adv. 2015, 5, 13341-13346.
[13] J. Hu, L. Dai, S. Liu, Macromolecules 2011, 44, 4699-4710.
[14] C. Rivas, G. J. Stasiuk, J. Gallo, F. Minuzzi, G. A. Rutter, N. J. Long, Inorg. Chem. 2013,
52, 14284-14293.
[15] X. Zhao, X. Jia, L. Liu, J. Zeng, K. Tian, T. Zhou, P. Liu, Biomacromolecules 2016, 17,
1496-1505.
[16] L. Yuan, W. Lin, Y. Feng, Org. Biomol. Chem. 2011, 9, 1723-1726.
[17] W. L. Czaplyski, G. E. Purnell, C. A. Roberts, R. M. Allred, E. J. Harbron, Org. Biomol.
Chem. 2014, 12, 526-533.
[18] a) Q. A. Best, C. Liu, P. D. van Hoveln, M. E. McCarroll, C. N. Scott, J. Org. Chem.
2013, 78, 10134-10143; b) A. Bender, Z. R. Woydziak, L. Fu, M. Branden, Z. Zhou, B.
D. Ackley, B. R. Peterson, ACS Chem. Biol. 2013, 8, 636-642.
[19] C. Liu, Q. A. Best, B. Suarez, J. Pertile, M. E. McCarroll, C. N. Scott, J. Fluoresc. 2015,
25, 231-237.
[20] E. S. Childress, C. A. Roberts, D. Y. Sherwood, C. L. M. LeGuyader, E. J. Harbron,
Anal. Chem. 2012, 84, 1235-1239.
[21] R. F. Kubin, A. N. Fletcher, J. Lumin. 1982, 27, 455-462.
[22] S. Winstein, N. J. Holness, J. Am. Chem. Soc. 1955, 77, 5562-5578.
[23] R. W. Taft, Jr., in Steric effects in organic chemistry (Ed.: M. S. Newman), John Wiley &
Sons, New York, 1956, pp. 556-675.
[24] E. Kutter, C. Hansch, J. Med. Chem. 1969, 12, 647-652.
[25] M. Charton, Prog. Phys. Org. Chem 1971, 8, 235-318.
[26] T. Fujita, T. Nishioka, Prog. Phys. Org. Chem 1976, 12, 49-89.
[27] C. G. Swain, E. C. Lupton, J. Am. Chem. Soc. 1968, 90, 4328-4337.
[28] C. Hansch, A. Leo, S. H. Unger, K. H. Kim, D. Nikaitani, E. J. Lien, J. Med. Chem.
1973, 16, 1207-1216.
15
This article is protected by copyright. All rights reserved.

10.1002/chem.201703176
Chemistry - A European Journal
[29] Additional regressions were run with recalculated F and R values from C. Hansch, A.
Leo and R.W. Taft, Chem. Rev., 1991, 91, 165, but these values did not appreciably alter
the regression equation or significance.
[30] H. H. Jaffé, Chem. Rev. 1953, 53, 191-261.
[31] a) R. Grainger, J. Cornella, D. C. Blakemore, I. Larrosa, J. M. Campanera, Chemistry – A
European Journal 2014, 20, 16680-16687; b) F. D'Anna, F. Ferroni, V. Frenna, S.
Guernelli, C. Z. Lanza, G. Macaluso, V. Pace, G. Petrillo, D. Spinelli, R. Spisani,
Tetrahedron 2005, 61, 167-178; c) F. D’Anna, V. Frenna, C. Z. Lanza, G. Macaluso, S.
Marullo, D. Spinelli, R. Spisani, G. Petrillo, Tetrahedron 2010, 66, 5442-5450.
[32] W.-J. Deng, D. Sun, B.-Y. Su, S.-P. Wang, H. Zheng, Acta Crystallogr. Sect. Sect. E:
Struct. Rep. Online 2009, 65, o1464-o1464.
[33] Gaussian 09, Revision D.01, M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria,
M. A. Robb, J. R. Cheeseman, G. Scalmani, V. Barone, G. A. P. B. Mennucci, H.
Nakatsuji, M. Caricato, X. Li, H. P. Hratchian, A. F. Izmaylov, J. Bloino, G. Zheng, J. L.
Sonnenberg, M. Hada, M. Ehara, K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T.
Nakajima, Y. Honda, O. Kitao, H. Nakai, T. Vreven, J. A. Montgomery, Jr., J. E. Peralta,
F. Ogliaro, M. Bearpark, J. J. Heyd, E. Brothers, K. N. Kudin, V. N. Staroverov, R.
Kobayashi, J. Normand, K. Raghavachari, A. Rendell, J. C. Burant, S. S. Iyengar, J.
Tomasi, M. Cossi, N. Rega, J. M. Millam, M. Klene, J. E. Knox, J. B. Cross, V. Bakken,
C. Adamo, J. Jaramillo, R. Gomperts, R. E. Stratmann, O. Yazyev, A. J. Austin, R.
Cammi, C. Pomelli, J. W. Ochterski, R. L. Martin, K. Morokuma, V. G. Zakrzewski, G.
A. Voth, P. Salvador, J. J. Dannenberg, S. Dapprich, A. D. Daniels, O. Farkas, J. B.
Foresman, J. V. Ortiz, J. Cioslowski, and D. J. Fox, Gaussian, Inc., Wallingford CT,
2009.
[34] a) P. Du, S. J. Lippard, Inorg. Chem. 2010, 49, 10753-10755; b) H. Zheng, G.-Q. Shang,
S.-Y. Yang, X. Gao, J.-G. Xu, Org. Lett. 2008, 10, 2357-2360; c) M. Yang, W. Meng, Q.
Ding, N. Su, X. Liu, M. Zhang, B. Yang, New J. Chem. 2015, 39, 4790-4795; d) N.
Chatterjee, S. B. Maity, A. Samadder, P. Mukherjee, A. R. Khuda-Bukhsh, P. K.
Bharadwaj, RSC Advances 2016, 6, 17995-18001; e) Y. Long, in CSD Communication,
2016.
[35] H. Montenegro, M. Di Paolo, D. Capdevila, P. F. Aramendia, M. L. Bossi, Photochem.
Photobiol. Sci. 2012, 11, 1081-1086.
[36] M. J. Maher, K. Yehl, F. Haque, A. Faint, K. D. Shimizu, C. J. Stephenson, Sens.
Actuators, B 2014, 200, 1-8.
[37] Z. Tian, J. Yu, C. Wu, C. Szymanski, J. McNeill, Nanoscale 2010, 2, 1999-2011.
[38] Y. Jiang, J. McNeill, Chem. Rev. 2017, 117, 838-859.
[39] a) W. Xu, S. Lu, M. Xu, Y. Jiang, Y. Wang, X. Chen, J. Mater. Chem. B 2016, 4, 292-
298; b) Y.-H. Chan, C. Wu, F. Ye, Y. Jin, P. B. Smith, D. T. Chiu, Anal. Chem. 2011, 83,
1448-1455.
[40] H. Cui, Y. Chen, L. Li, Y. Wu, Z. Tang, H. Fu, Z. Tian, Microchimica Acta 2014, 181,
1529-1539.
[41] W. R. Carroll, C. Zhao, M. D. Smith, P. J. Pellechia, K. D. Shimizu, Org. Lett. 2011, 13,
4320-4323.
16
This article is protected by copyright. All rights reserved.

10.1002/chem.201703176
Chemistry - A European Journal
Scheme 1. RSL acid-base equilibrium.
Table 1. RSL derivatives and pKa values.
Aniline
RB
cmpd
#
RB
pKa
R6G
cmpd
#
R6G
pKa
4-Cl
1
4.21 ± 0.01
5.46 ± 0.03
4.34 ± 0.05
5.38 ± 0.08
4.31 ± 0.02
4.52 ± 0.08
4.03 ± 0.07
4.90 ± 0.11
5.72 ± 0.13
4.43 ± 0.08
4.33 ± 0.01
4.31 ± 0.02
4.28 ± 0.02
2,6-Cl2-4-NO2
3,5-Cl2
2,6-Cl2
H
2
3
4
14
15
16
17
18
19
5.62 ± 0.04
3.81 ± 0.02
4.30 ± 0.04
4.15 ± 0.05
5.32 ± 0.07
5.87 ± 0.06
5
2,6-F2
2,6-OEt2
2,6-Me2
2,6-iPr2
2-Cl
6
7
8
9
10
11
12
13
2-F
2-Me
2-iPr
17
This article is protected by copyright. All rights reserved.

10.1002/chem.201703176
Chemistry - A European Journal
Scheme 2. RSL structures.
Figure 1. Absorption (A) and fluorescence (B) of compound 9 in 1:1 (v/v) ethanol-water during
titration between pH 9.15 and 4.05.
Figure 2. Normalized fluorescence intensity at the max,fl as a function of pH for Series 1
compounds in 1:1 v/v ethanol-water.
18
This article is protected by copyright. All rights reserved.

10.1002/chem.201703176
Chemistry - A European Journal
Figure 3. Normalized fluorescence intensity at the max,fl as a function of pH for RB Series 2 (A)
and R6G Series 3 (B) compounds in 1:1 v/v ethanol-water.
Figure 4. Experimental pKa values for RB Series 2 (red) and R6G Series 3 (blue) compounds vs.
the Taft-Kutter-Hansch Es steric parameter.
19
This article is protected by copyright. All rights reserved.

10.1002/chem.201703176
Chemistry - A European Journal
Table 2. Substituent parameters for LFER analysis.
[a]
Substituent
Es[a]
p
F[a]
R[a]
angle[b]
Cl
-0.97
0
0.23
0
0.41
0
-0.15
0
155 (4)
173 (5)
169 (6)
162 (7)
155 (8)
152 (9)
H
F
-0.46
-0.55
-1.24
-1.71
0.06
-0.24
-0.17
-0.15
0.43
0.22
-0.04
-0.05
-0.34
-0.44
-0.13
-0.10
OEt
Me
iPr
[a] Parameters derived from ref. 26. Parameters were doubled in
analyses of di-ortho-substituted compounds.
[b] Calculated xanthene angles for the designated derivatives as
described in the text.
Figure 5. Predicted vs. experimental pKa values for RB Series 2 (A) and R6G Series 3 (B)
compounds. Predicted values were obtained by multiple linear regression according to the
modified Fujita-Nishioka model in Eq. 2.
Figure 6. Crystal structure of 9.
20
This article is protected by copyright. All rights reserved.

10.1002/chem.201703176
Chemistry - A European Journal
Figure 7. (A) Graphic representation of 9-doped CPNs with pH-dependent ratiometric
fluorescence properties. (B) Fluorescence spectra of 9-doped CPNs in aqueous suspension as a
function of pH, where the spectra with the highest and lowest pH values are shown in black and
labeled with their pH values. (C) Ratio of 9 to CPN fluorescence intensities (I577/I535) as a function
of pH for 2 separate samples during titrations from high to low pH with acid (red and green
symbols) and from low to high pH with base (orange and blue symbols). The sigmoidal curve is
presented as a guide to the eye. (D) Relative fluorescence intensity of 9 at 577 nm at pH 5.11
when excited directly at 535 nm (red) or via FRET from the CPNs at 450 nm (blue). The
fluorescence intensities were corrected for CPN emission at 577 nm.
21
This article is protected by copyright. All rights reserved.