Structural variability of pendant groups within the interlayer region of zirconium arene-phosph(on)ates: Chemical and structural characterization of oxy- And methyl-linked 2-naphthyl phosphonates, and mixed oxy-linked derivatives

Article abstract of DOI:10.1039/c9dt03875b

Several new zirconium phosphonates incorporating the naphthalene ring and having the general formula Zr(O₃PR)₁(O₃PR′)₁ [R and R′ = -CH₂C₁₀H₇, -OC₁₀H₇, -CH₃, -OC₂H₅, -OH] have been synthesized. These materials were chemically characterized using thermal gravimetric analysis (percentage of organic content), infrared spectroscopy (presence of the desired organic functional groups), and solid-state ³¹P NMR (phosphorus environments), while the structural parameters were determined using X-ray powder diffraction (interlayer d spacings). The two new zirconium bis(phosphonates), Zr(O₃PC₁₀H₇)₂ and Zr(O₃PCH₂C₁₀H₇)₂, were found to have d spacings of 19.6 and 20.0 ?, respectively. Three of the four zirconium mixed phosphonates examined are found to be single-phase structures with random mixtures of the organic moieties within the interlayer, and possess d spacings (14.3, 15.3, and 16.1 ?) that are between those of the two parent zirconium bis(phosphonates). The fourth is found to be a staged or segregated structure and possesses a d spacing that is approximately a sum of the two parent zirconium bis(phosphonates), with a d spacing of 28.2 ?. Solid-state ³¹P NMR of Zr(O₃PCH₂C₁₀H₇)₂ revealed the presence of two isotropic resonances, which is interpreted in terms of two distinct, "locked-in" conformations of the -CH₂C₁₀H₇ group. The experimental d spacings of the zirconium bis(phosphonates) correlate well with a simple predictive model based on the effective length and predominant conformation of the organic functional group.

Full text of DOI:10.1039/c9dt03875b

Dalton
Transactions
View Article Online
View Journal
PAPER
Structural variability of pendant groups within the
interlayer region of zirconium arene-phosph(on)ates:
chemical and structural characterization of oxy-
and methyl-linked 2-naphthyl phosphonates,
and mixed oxy-linked derivatives†
Cite this: DOI: 10.1039/c9dt03875b
a
b
Jay C. Amicangelo
and Willem R. Leenstra
*
Several new zirconium phosphonates incorporating the naphthalene ring and having the general formula
Zr(O PR) (O PR’) [R and R’ = –CH , –OC10 , –CH , –OC , –OH] have been synthesized. These
3
1
3
1
2
10
C H
7
H
7
3
2 5
H
materials were chemically characterized using thermal gravimetric analysis (percentage of organic
content), infrared spectroscopy (presence of the desired organic functional groups), and solid-state
3
1
P NMR (phosphorus environments), while the structural parameters were determined using X-ray powder
diffraction (interlayer d spacings). The two new zirconium bis(phosphonates), Zr(O and
Zr(O PCH , were found to have d spacings of 19.6 and 20.0 Å, respectively. Three of the four
3 7 2
PC10H )
3
2 10 7 2
C H )
zirconium mixed phosphonates examined are found to be single-phase structures with random mixtures
of the organic moieties within the interlayer, and possess d spacings (14.3, 15.3, and 16.1 Å) that are
between those of the two parent zirconium bis(phosphonates). The fourth is found to be a staged or
segregated structure and possesses a d spacing that is approximately a sum of the two parent zirconium
3
1
bis(phosphonates), with a d spacing of 28.2 Å. Solid-state P NMR of Zr(O
presence of two isotropic resonances, which is interpreted in terms of two distinct, “locked-in” confor-
mations of the –CH group. The experimental d spacings of the zirconium bis(phosphonates) cor-
3 2 10 7 2
PCH C H ) revealed the
Received 30th September 2019,
Accepted 1st December 2019
2 10 7
C H
DOI: 10.1039/c9dt03875b
relate well with a simple predictive model based on the effective length and predominant conformation
rsc.li/dalton
of the organic functional group.
layers. Each zirconium is bonded to six oxygen atoms from
six different phosphate groups, giving it an octahedral bonding
1
. Introduction
α-Zirconium phosphate, Zr(O POH) ·H O, is a layered solid environment, and each phosphate group is bonded to three
3
2
2
material, whose crystal structure was first determined by different zirconium atoms. The fourth oxygen atom of each phos-
1
Clearfield and Smith. The structure is found to be com- phate group is pointing into the region between the layers,
prised of two-dimensional planes of zirconium(IV) atoms co- known as the interlayer region, and has a hydrogen atom
valently bonded to phosphate groups above and below the attached to it. The interactions between the layers of α-zirconium
planes, giving rise to the layers, and the three-dimensional phosphate is dominated by hydrogen bonding between the
2–4
structure is formed by stacking the two-dimensional layers layers and with an intercalated water molecule.
5
6
upon one another, with no covalent bonding between the
It was demonstrated by Alberti and co-workers, Maya, and
7
Dines and DiGiacomo that by using phosphonic acids
(
H O PR) rather than phosphoric acid (H PO ), where R is an
2 3 3 4
a
organic functional group, a series of new compounds known
as zirconium phosphonates, Zr(O PR) , or more precisely as
School of Science, Penn State Erie, The Behrend College, Erie, PA 16563, USA
b
Department of Chemistry, University of Vermont, Burlington, VT 05405, USA.
3
2
E-mail: Willem.Leenstra@uvm.edu
zirconium bis(phosphonates), could be synthesized.
Zirconium bis(phosphonates) possess a layered structure
similar to that of zirconium phosphate, but with the organic
functional groups now projecting into the interlayer region.
For these materials, the principal interactions between the
layers are the van der Waals forces between the interlayer func-
†
Electronic supplementary information (ESI) available: Synthetic methodologies
for Zr(ONap) , Zr(ONap) (Me) , 2-napthylmethylphosphonic acid, and its precur-
sor diethyl 2-napthylmethylphosphonate. Sets of TGA traces, IR spectra, solid-
2
1
1
3
1
state P NMR spectra, and XRD patterns for zirconium phosph(on)ate com-
pounds examined but not explicitly displayed in the paper are given as Fig. 1S
through 4S, respectively. See DOI: 10.1039/c9dt03875b
This journal is © The Royal Society of Chemistry 2019
Dalton Trans.

View Article Online
Paper
Dalton Transactions
tional groups. One of the key structural parameters used to
In some cases, particularly if the two organic groups are of
characterize these materials is the interlayer spacing or d significantly different size or possess different types of inter-
spacing, which is the distance between the adjacent zirconium molecular interactions (hydrophilic versus hydrophobic for
planes, and it was found that the interlayer spacing varied in example), structures in which the different organic groups seg-
accordance with the size of the organic functional group. Since regate into adjacent layers, known as “staging”, have also been
9
,10,13,15–18
those initial reports, a large number of zirconium bis(phos- observed.
For a system with a staged structure, the
phonates) have been synthesized, with a variety of organic observed interlayer spacing is found to be larger than the inter-
functional groups, such as alkyl chains, functionalized alkyl layer spacings of either of the two individual zirconium bis
groups, and various rings of both the aromatic and non-aro- (phosphonate) compounds and often is found to be an
8
–10
matic type.
By co-precipitating the zirconium(IV) with two different ual zirconium bis(phosphonates), as is depicted in Fig. 1(b).
phosphonic acids, H PR and H PR′, mixed zirconium The structure displayed in Fig. 1(b) is an idealized situation in
phosphonates having the general formula Zr(O PR) (O PR′)2−x which the long functional groups are all segregated into the
Typically, the observed inter- lower layer and the short functional groups all segregated into
layer spacing for the mixed zirconium phosphonate, the upper layer. Within the possible structural arrangments for
Zr(O PR) (O PR′)2−x, is found to be intermediate between the staged zirconium phosph(on)ates, there are some other more
two zirconium bis(phosphonates), Zr(O PR) and Zr(O PR′) , subtle functional group arrangements that include some
approximate sum of the interlayer spacings of the two individ-
2
O
3
2 3
O
3
x
3
1
1–14
have been synthesized.
3
x
3
3
2
3
2
which is generally taken to indicate that the mixed compounds mixing of functional groups within one of the layers (for
contain a random mixture of the organic groups, i.e., a “solid example, see Fig. 1 of ref. 13 or Fig. 4(c) of ref. 18) and alter-
solution” with all layers at identical stoichiometry. This is pic- nate arrangments of adjacent functional groups (for example,
torially shown in Fig. 1(a), with the different organic functional see Scheme 3 of ref. 16). These tend to be more rare than the
groups represented by the long and short columns. Note that basic structure presented in Fig. 1(b) and were not observed
the structure shown in Fig. 1(a) is an idealized case of a for the materials presented in the current work.
1
: 1 mixture in which each long functional group is opposite a
Zirconium phosph(on)ates are a class of materials that have
short functional group. In that case, the arrangement gives several potential practical applications in a variety of areas
1
9
20–22
23
rise to an interlayer spacing that is intermediate between the such as catalysis, ion exchange,
photophysics, nuclear-
2
4,25
26
27,28
interlayer spacings of the two zirconium bis(phosphonates), waste remediation,
for which the two interlayer structures are represented in for example. Several reviews of zirconium phosph(on)ate chem-
Fig. 1(c). It is reasonable to propose such an arrangement as istry have appeared in the literature
biotechnology, and drug delivery,
8
–10
and demonstrate the
the extended network seeks a global energy minumum by max- growth of this field of solid-state inorganic chemistry.
imizing van der Waals internactions.
The present work describes a chemical and structural inves-
tigation of zirconium phosph(on)ates containing methyl-
naphthalene or oxynaphthalene rings as the primary organic
pendant groups. The materials can be generally organized into
two categories: (a) zirconium bis(phosph(on)ates) containing
the naphthalene ring systems mentioned above and (b) zirco-
nium mixed phosph(on)ates containing the naphthalene ring
systems mixed with alkyl or alkoxy chains. The motivation for
synthesizing and thoroughly examining the chemical and
structural aspects of the present materials stems from our
general interest in chromophore-containing layered materials
–
zirconium phosph(on)ates containing methylnaphthalene
or oxynaphthalene ring systems have, to date, not been
investigated.
We have found that the zirconium phosphate crystal lattice
structure provides a unique environment for the examination
of the photophysics of these aromatic pendant groups. In fact,
in our earlier report concerning the emission spectra of two
zirconium bis(arenephosphonates) with naphthalene and
anthracene rings, we observed that both of the zirconium bis
(
arenephosphonates) displayed prominent excimer emission,
Fig. 1 Schematic representations of the three general structural possi-
but that the ratio of excimer to monomer emission intensities
in these systems was drastically different. Specifically, the
anthracene system displays almost no monomer emission,
whereas the naphthalene system displays comparable excimer
and monomer fluorescence intensities. We attributed this
bilities for a zirconium mixed phosph(on)ates, Zr(O
3
PR)
x
(O
3
PR’)2−x, with
29
two different organic functional groups, represented by the large and
small columns, where (a) the functional groups are randomly dispersed,
(
b) the functional groups are segregated or staged, and (c) the functional
groups precipitate into distinct co-crystallites.
Dalton Trans.
This journal is © The Royal Society of Chemistry 2019

View Article Online
Dalton Transactions
Paper
effect to the differing amounts of π-overlap within these ring ates, as well as for the one phosphonic acid and its precursor,
systems, with the former having a larger degree of overlap of are provided as ESI.†
neighboring chromophores, thus leading to increased excimer
formation.
Purchased chemicals were used as received. Hydrofluoric
acid was purchased from J.T. Baker, Inc.; CDCl was purchased
3
These interesting results led us to the idea of tuning the from Cambridge Isotope Labs, Inc.; zirconyl chloride, ethanol,
amount of excimer formation in these systems statistically (a) acetone, and diethyl ether were purchased from Aldrich
by introducing alkyl spacer groups, in effect diluting the Chemical Co. 2-Naphthyl phosphate monosodium salt was
chromophores and decreasing the amount of excimer for- purchased from Lancaster Synthesis; 2-napthylmethyl-
mation, and (b) by varying the size of the spacer groups (i.e., phosphonic acid was synthesized as described in the ESI.†
the volume available to a chromophore for exploring its confor- Ethyl phosphate was purchased from TCI America as a mixture
mational degrees of freedom is varied) in turn modulating the of the monoethyl and diethyl forms, and monoethyl phosphate
amount of excimer formation further. Our initial chemical and was isolated from the mixture by an extraction procedure
31
structural investigation of mixed arene/alkoxy zirconium similar to that described by Stewart and Crandall. Solvents
phoshp(on)ates was concerned with materials of the type used for UV-vis measurements were HPLC or spectroscopy
3
0
Zr(O
present paper, we follow up on our initial chemical and struc-
tural investigations and present data on number of nate precursor compound were characterized by melting point
additional materials within the Zr(O PR) (O
3
PC10
H
7
)
1
(O
3
PR)
1
, where R = OC
2
H
5
and OC
4
H
9
.
In the grades.
2-Napthylmethylphosphonic acid and its diethyl phospho-
a
1
13
31
3
x
3
PR′)2−x motif in or boiling point, mass spectrometry, NMR ( H, C, and P),
which an additional linking moiety is inserted between the IR and UV-vis spectroscopies. All zirconium phosph(on)ate
naphthalene ring and the phosphorus atom: (i) R = R′ = materials were fully characterized by thermal gravimetric ana-
3
1
CH
2
C
10
H
7
; (ii) R = R′ = OC10
H
7
; (iii) R = OC10
7 3
H , R′ = CH ; (iv) lysis (TGA) and IR, UV-vis, and solid-state P NMR spectrosco-
R = OC10
H
7
, R′ = OC , and (v) R = OC10H
2
H
5
7
, R′ = OH (random pies. Structural parameters of the zirconium phosph(on)ates
and staged). The chemical characterization of these new (d-spacings) were determined using X-ray powder diffraction
materials using thermal gravimetric analysis (percentage of (XRD).
organic content), infrared spectroscopy (presence of the
31
Zirconium bis(2-naphthylmethylphosphonate),
Zr(O PCH C H ) , [Zr(CH Nap) ]
desired organic functional groups), and solid-state P NMR
phosphorus environments) is discussed in section 3.1, while
3
2
10 7 2
2
2
(
the characterization of the primary structural parameter (inter- To a solution of zirconyl chloride (0.206 g, 0.639 mmol) in
layer d spacing) using X-ray powder diffraction is discussed in 24.0 mL of 75% ethanol/H O was added 1.25 mL of 2.04 M HF
2
section 3.2. Two of the key/novel findings of our study are the (2.57 mmol). This solution was added dropwise to a solution
formation of a “staged” structure for a mixed zirconium phos- of 2-napthylmethylphosphonic acid (0.284 g, 1.277 mmol) in
phate involving the OC H and OH functional groups, and 25.0 mL of 75% ethanol/H O with stirring. Afterwards, 5.0 mL
1
0
7
2
the presence of two distinct conformations of the functional of 75% ethanol/H
group for the zirconium bis(phosphonate) with the CH
functional group. A photophysics-oriented investigation of the six days. The solution was cooled to room temperature, the
subject materials will be attempted in the future. solid collected by centrifugation and washed successively with
5% ethanol/H O, acetone, diethyl ether and air dried to yield
Zr(CH Nap) as a white solid (0.302 g, 89%):
2
O was used to rinse the zirconyl chloride
2
C
10
H
7
flask and the entire mixture was refluxed under nitrogen for
7
2
3
1
P NMR
500 MHz, solid state) δ 5.2, 2.1; XRD d001 (Å) 20.0; IR (KBr,
2
2
(
2
. Experimental section
−
1
cm ) 3442, 3055, 3023, 2921, 2911, 2892, 2853, 1632, 1600,
540, 1510, 1440, 1403, 1384, 1364, 1264, 1248, 1209, 1168,
2
.1 Materials and procedures
1
Several novel zirconium phosph(on)ates containing naphtha- 1145, 1089, 1035, 956, 906, 865, 834, 816, 802, 766, 746, 725,
lene rings have been synthesized, and in one case (R = 669, 642, 620, 600, 521, 506, 478; TGA (150 °C to 1000 °C, Zr
CH C H ), the phosphonic acid and its diethyl phosphonate (CH Nap) → ZrP O ) % Wt loss: Exptl (calcd) 49.1 (50.1); UV
2
10
7
2
2
2 7
precursor were also synthesized. The materials studied (glycerol mull, nm) λmax 225, 279, 291, 298, 322.
include zirconium phosph(on)ates with one type of functional
Zirconium 2-naphthyl phosphate ethyl phosphate, Zr
group [Zr(O PR) , zirconium bis(phosph(on)ates)] as well
3
2
(
O POC H ) (O POC H ) [Zr(ONap) (OEt) ]
3 10 7 1 3 2 5 1 1 1
as materials containing two different functional groups
Zr(O PR) (O PR′)2−x, zirconium mixed phosph(on)ates].
[
3
x
3
A
solution comprised of zirconyl chloride (0.173 g,
Because the syntheses for the zirconium phosph(on)ates are 0.537 mmol) in 19.0 mL of water with 1.08 mL of 1.98 M HF
very similar (mixing of a zirconyl chloride solution in HF with (2.14 mmol) was added dropwise to a solution of 2-naphthyl
a solution of the stoichiometric amounts of the appropriate phosphate monosodium salt (0.132 g, 0.537 mmol) and mono-
phosphonic acid(s) or phosphates followed by refluxing for ethyl phosphate (0.070 g, 0.537 mmol) in 20.0 mL of water
several days, then filtering or centrifuging and drying), only with stirring, 5.0 mL of water was used to rinse the zirconyl
the representative synthetic particulars are detailed here. The chloride flask and the entire mixture was refluxed under nitro-
synthetic methodologies for the other zirconium phosph(on) gen for 18 hours. After cooling to room temperature, the solid
This journal is © The Royal Society of Chemistry 2019
Dalton Trans.

View Article Online
Paper
Dalton Transactions
was collected by filtration, washed successively with water, a System 7/4 console and a model 3600 Data Station. All TGA
acetone, diethyl ether and air dried to afford Zr(ONap) (OEt) scans were obtained on 10 to 15 mg of solid, with flowing com-
1
1
3
1
−1
as a white solid (0.205 g, 87.3%): P NMR (500 MHz, solid pressed air, and a 10 °C min heating rate. The experimental
state) δ −22.8, −29.1; XRD d001 (Å) 16.1; IR (KBr, cm− ) 3058, reproducibility of % weight loss values for the zirconium pho-
1
3
1
9
027, 2981, 2931, 2909, 2869, 1632, 1601, 1580, 1548, 1512, phonates synthesized was determined to be ±0.2%. All NMR
467, 1442, 1384, 1366, 1357, 1252, 1220, 1169, 1120, 1068, spectra were acquired on a Bruker ARX 500 MHz spectrometer
88, 943, 878, 851, 807, 789, 744, 702, 646, 606, 539, 513, 475; interfaced with an Aspect Station computer system and using
TGA (150 °C to 1000 °C, Zr(ONap) (OEt) → ZrP O ) % Wt Bruker UXNMR software. Spectrometer operating frequencies
1
1
2 7
1
13
Loss: Exptl (Calcd) 36.1 (39.4); UV (glycerol mull, nm) λmax 226, were 500.13 MHz, 125.76 MHz, and 202.40 MHz for H, C,
3
1
1
13
2
29, 282, 292, 309, 316, 325.
and P nuclei, respectively. For H and C NMR spectra,
either tetramethylsilane or the resonance of the deuterated
Zirconium 2-naphthyl phosphate phosphate,
Zr(O POC H ) (O POH) [Zr(ONap) (OH) ]
31
solvent was used as an internal reference, while for P NMR
3
10 7 1
3
1
1
1
3 4
spectra, 85% H PO was used as an external reference, unless
3
1
13
A sample of Zr(ONap) (OEt) (34.7 mg) was heated in a tube otherwise noted. Spectra of P and C nuclei were obtained
furnace for 24 hours at 270 °C with flowing oxygen. The temp- with H decoupling, using standard Bruker pulse sequences.
erature was then raised to 350 °C and the solid heated six Solid-state P NMR was performed using 7 mm zirconia
1
1
1
3
1
hours with flowing oxygen, after which Zr(ONap) (OH)1 was rotors, Delrin caps, and magic-angle spinning (MAS) at 4 kHz.
1
obtained as a brown solid: XRD d001 (Å) 14.3; IR (KBr, cm−
1
)
The average error in the solid-state P NMR chemical shifts
056, 1850, 1777, 1719, 1625, 1605, 1584, 1510, 1482, 1466, was estimated to be ±0.2 ppm. Absorbance spectra were
446, 1426, 1384, 1366, 1351, 1249, 1216, 1059, 748, 609, 522, obtained on a PerkinElmer Lambda 6 UV-vis spectrometer.
31
3
1
4
82; TGA (150 °C to 1000 °C, Zr(ONap)
1
(OH)
1
→ ZrP
2
O
7
) % Wt Precursor compounds were dissolved in an appropriate HPLC
or spectral grade solvent and the absorbance spectrum
measured in a 1 cm pathlength quartz cuvette. Zirconium
phosphonates were ground to a fine powder and then mixed
Loss: Exptl (Calcd) 28.3 (35.2).
Staged zirconium 2-naphthyl phosphate phosphate,
Zr(O POC H ) (O POH) [stg Zr(ONap) (OH) ]
3
10 7 1
3
1
1
1
with glycerol to form a mull or suspension, with “concen-
To a solution of 2-naphthyl phosphate monosodium salt trations” generally ranging from 5 × 10− mg zirconium
2
0.190 g, 0.772 mmol) in 10.0 mL of water was added 1.48 mL phosph(on)ate per mg glycerol to 1 × 10− mg zirconium
3
(
of 0.522 M phosphoric acid (0.772 mmol). Zirconyl chloride phosph(on)ate per mg glycerol. The absorbance spectrum of
0.249 g, 0.772 mmol) was then added to this mixture with stir- the glycerol mull was obtained by placing the mull between
(
ring followed by 0.40 mL of 2.04 M HF (0.816 mmol). The two 0.18 cm thick quartz plates, separated by a 0.12 cm thick
entire mixture was refluxed under nitrogen for six days. The plastic spacer, with a rectangular opening in the center
suspension was cooled to room temperature, 20 mL of absol- (1.1 cm × 1.9 cm).
ute ethanol added and the mixture placed in a freezer for two
hours. The solid was collected by filtration and washed succe-
sively with water, acetone, diethyl ether and air dried to afford 3. Results and discussion
3
1
1 1
stg Zr(ONap) (OH) as a white solid (0.271 g, 85.7%): P NMR
3
.1 Chemical characterization
(
(
500 MHz, solid state) δ −21.6, −30.0; XRD d001 (Å) 28.2; IR
KBr, cm− ) 3051, 1635, 1600, 1578, 1559, 1540, 1510, 1467,
1
3.1.1 Thermal gravimetric analysis. Thermal gravimetric
analysis (TGA) can be used as an analytical tool for the confir-
mation of the composition of the organic components of zirco-
nium phosphonates. It was shown by Costantino and by
Alberti that heating zirconium phosphonates in the presence
of oxygen or air to temperatures between 1000 °C and 1200 °C
converted the compounds to zirconium pyrophosphate,
1
9
5
441, 1419, 1383, 1356, 1253, 1222, 1172, 1081, 1060, 1001,
59, 878, 853, 804, 770, 760, 739, 699, 648, 636, 618, 608, 597,
63, 531, 483, 435; TGA (150 °C to 1000 °C, Zr(ONap) (OH) →
32
1
1
1
3
2 7
ZrP O ) % Wt Loss: Exptl (Calcd) 31.9 (35.2).
2
.2 Instrumental methods
ZrP
2 7
O , in either the α-layered (d001 = 6.1 Å) or the cubic (d001 =
X-ray powder diffraction patterns were obtained with either a
Scintag X1 or a Philips PW1877 automated powder diffract-
ometer, using Cu Kα radiation and an internal silicon powder
standard with all samples. Patterns were generally measured
between 1.00° and 35.00° (2θ), with a step size of 0.02° and
X-ray tube settings of 45 kV and 40 mA. The average error in
the d-space values for all synthesized zirconium phosph(on)
ates is estimated to be ±0.3 Å. Infrared spectra were recorded
on a PerkinElmer System 2000 FT-IR instrument, as KBr
pellets for solids and on NaCl plates for neat liquids, using
4.8 Å) structure. Therefore knowing the final composition of
the pyrolysis product, the experimental % weight loss deter-
mined can be compared to the theoretical % weight loss (cal-
culated based on the eqn (1) shown below).
FW ZrðO3PRÞ ðO PR′Þ
ꢀ FW ZrP2O7
x
3
2ꢀx
%
weight loss ¼
ꢁ 100
ð1Þ
FW ZrðO3PRÞ ðO3PR′Þ
x
2ꢀx
In general, the TGA traces of zirconium phosph(on)ates
examined show loss of surface-bound or intercalated water
PerkinElmer TGS-2 Thermogravimetric Analyzer interfaced to below 150 °C, followed by volatilization of the organic frag-
−
1
4
cm
resolution. TGA scans were performed on a
Dalton Trans.
This journal is © The Royal Society of Chemistry 2019

View Article Online
Dalton Transactions
Paper
ments between 400 °C and 800 °C. As examples, the TGA weight loss of Zr(ONap)
1
(OEt)
1
is between the values of the two
traces of zirconium bis(2-naphthyl phosphate) [Zr(ONap)2], bis(phosphate) compounds. This is in accordance with the
zirconium 2-naphthyl phosphate ethyl phosphate difference in the masses of the organic components in these
(OEt) ], and zirconium bis(ethyl phosphate) compounds. It is evident from Fig. 2 that the organic frag-
Zr(OEt) ] are displayed in Fig. 2. The TGA traces of zirconium ments of Zr(OEt)2 begin to volatilize at a lower temperature
[
[
Zr(ONap)
1
1
2
bis(2-naphthylmethylphosphonate)[Zr(CH
-naphthyl phosphate methylphosphonate [Zr(ONap)
zirconium 2-naphthyl phosphate phosphate [Zr(ONap) (OH) ], lists the experimentally determined and theoretically calcu-
2
Nap)
2
], zirconium (∼300 °C) as compared to those of Zr(ONap)
2
(∼450 °C), which
2
1
(Me) ], reflects the greater thermal stability of the arene ring. Table 1
1
1
1
and staged zirconium 2-naphthyl phosphate phosphate [stg lated % weight loss values for the zirconium phosph(on)ates
Zr(ONap) (OH) ] are given as Fig. 1S in the ESI.† Note that the synthesized. Also listed in Table 1 are the % weight loss values
1
1
3
0
compound referred to as stg Zr(ONap) (OH) was determined for zirconium bis(ethyl phosphate) [Zr(OEt) ], zirconium bis
to possess a segregated or staged interlayer structure from the (methylphosphonate) [Zr(Me)
X-ray diffraction analysis (vide infra).
1
1
2
3
3,34
2
],
and α-zirconium phos-
for comparison with the zirconium mixed
3
5
phate [Zr(OH)
2
]
As can be observed in Fig. 2, the trace of Zr(ONap) displays phosph(on)ates. Although the range of temperatures over
2
a larger % weight loss than the trace of Zr(OEt)
2
and the % which the pendant-group loss varies (see Fig. 2 and 1S†), an
approximate midpoint of the temperature loss range is indi-
cated in Table 1. The agreement between the experimentally
determined and the theoretically calculated weight-loss values
2 2
are generally very good, except for Zr(ONap) , Zr(OEt) , and
Zr(ONap) (OH) , with the experimental values for these latter
1
1
three being lower than their theoretically calculated values, on
the order of 14–20%. This systematic discrepancy toward lower
values has been discussed in our previous chemical and struc-
tural report on mixed arene/alkoxy zirconium phoshp(on)ates,
as well as other studies, and is attributed to the incomplete
combustion of the organo-phosphate groups and the sub-
sequent trapping or surface deposition of carbon from the
organic fragments, as evidenced by a slight grayish color of the
1
3,30,34,36
final solid after TGA analysis.
This results in an artifi-
cial increase in the amount of solid that remains, thereby arti-
ficially lowering the weight-loss values.
Fig. 2 also shows that the organic functional groups of
1 1
Zr(ONap) (OEt) do volatilize at noticeably different temperatures
from one another and therefore the TGA of this compound
could theoretically yield the relative stoichiometry directly,
however, the volatilization regions are not sufficiently well sep-
arated to allow for an accurate determination. For the other
mixed compounds studied, the two functional groups did not
volatilize separately and therefore the TGA curves of these com-
pounds will not directly yield the relative stoichiometry.
However, even with the lack of distinct volatilization, the stoi-
Fig. 2 Thermal gravimetric analysis curves of Zr(OEt)
and Zr(ONap)
2 1 1
, Zr(ONap) (OEt) ,
2
.
3 2 3 1 3 1
Table 1 TGA data for the zirconium bis(phosph(on)ates), Zr(O PR) , and zirconium mixed phosph(on)ates, Zr(O PR) (O PR’)
T
(°C)
loss
Exptl percent
weight loss
Theor percent
weight loss
Percent
difference
a
Compound
Zr(CH Nap)
x
2
2
525
600
450
650
49.1
43.5
17.4
4.9
12.0
33.6
36.1
28.3
31.9
50.1
50.5
21.8
5.0
12.0
34.9
39.4
35.2
35.2
2.0
Zr(ONap)
2
13.9
20.2
2.0
0.0
3.7
8.4
19.6
9.4
b
Zr(OEt)
2
c
Zr(Me)
Zr(OH)
2
d
2
Zr(ONap)
Zr(ONap)
Zr(ONap)
1
1
1
(Me)
(OEt)
(OH)
1
575
575
550
550
0.94
0.78
0.75
0.90
1
1
e
stg Zr(ONap)
1
(OH)
1
a
b
c
d
Calculated for the mixed compounds using eqn (2) as described in the text. Taken from ref. 30. Taken from ref. 33 and 34. Taken from ref.
5. Stg refers to the structural classification of a staged compound.
e
3
This journal is © The Royal Society of Chemistry 2019
Dalton Trans.

View Article Online
Paper
Dalton Transactions
chiometric parameter x, which is defined as the number of
moles of H O PR incorporated into the zirconium mixed
2
3
1
3
phosph(on)ate, can be estimated using eqn (2),
1
W150 ꢀ W1000
W150
xR þ ð2 ꢀ xÞR′ ꢀ O2
2
ΔW ¼
¼
ð2Þ
ZrP2O6 þ xR þ ð2 ꢀ xÞR′
where W150 and W1000 are the % weight loss values at 150 °C
chosen to represent a state without water content) and at
000 °C, respectively; R and R′ are the formula weights of the
different organic fragments; and formulas ZrP O and O also
(
1
2
6
2
represent their formula weights. Using the appropriate weights
for the mixed compounds, and inserting these into eqn (2),
yields values of x listed in Table 1. The values of x determined
1 1 1 1
this way for Zr(ONap) (Me) and stg Zr(ONap) (OH) are close
to unity, demonstrating that these compounds are indeed
comprised of an approximately equal mole ratio of the two
functional groups. The deviations of the calculated values of x
1 1 1 1
for Zr(ONap) (OEt) and Zr(ONap) (OH) from unity are
slightly larger, however, we believe that these compounds are
also comprised of approximately equimolar compositions of
the two functional groups, and that the discrepancies are pri-
marily caused by incomplete combustion or surface deposition
of carbon from the organic fragments, as mentioned above.
3
.1.2 Infrared spectroscopy. We chose to use IR spec-
troscopy to confirm the presence of the desired functional
groups in all of the materials and for a qualitative confir-
mation of the thermal analysis measurements. The observed
IR vibrational frequencies for each of the zirconium phosph
Fig. 3 Infrared spectra of Zr(ONap)
KBr pellets in the 3100 to 2800 cm and 1700 to 400 cm spectral
regions.
2
−
, Zr(ONap)
1
(OEt)
1
2
, and Zr(OEt) as
(
on)ates synthesized are listed at the end of each synthesis
description and, as examples, the IR spectra of Zr(ONap)
Zr(ONap) (OEt) and Zr(OEt) are shown in Fig. 3. The IR spectra
1
−1
2
,
1
1
2
of Zr(CH
Zr(ONap)
nium phosph(on)ates, strong absorptions between 1100 cm
2
Nap)
2
, Zr(ONap)
1 1 1 1
(Me) , Zr(ONap) (OH) , and stg
1
(OH)
1
are given as Fig. 2S in the ESI.† For all the zirco-
−
1
compounds. Similar results are observed for Zr(CH
group Fig. 3 also shows that the mixed compound, Zr(ONap)
1
Nap)
(OEt)
.
,
2
2
and 900 cm− due to the stretching modes of the –PO
1
3
1
−
1
and 400 cm−1 does indeed possess both functional group types, as evidenced
and weaker absorptions between 600 cm
corresponding to the bending modes of the –PO group are by the clear presence of distinct vibrations due to the naphthyl
3
3
7,38
observed.
For the compounds containing arene rings, the ring and the ethyl group. A similar analysis involving its dis-
1
(Me) ,
IR spectra display absorptions for the aromatic C–H stretching tinct functional group vibrations also held for Zr(ONap)
1
modes (3100 cm− to 3000 cm ), the aromatic –CvC– ring Zr(ONap)
1
−1
(OH)
3.1.3 Solid-state
bending modes (1300 cm to 1000 cm ), and the out-of- NMR-active nucleus, solid-state P NMR can be used to deter-
and stg Zr(ONap)
(OH)
.
1
1
1
1
−
1
−1
31
stretching modes (1600 cm to 1400 cm ), the in-plane C–H
P NMR. Since phosphorus-31 is an
−
1
−1
31
−1
−1 39
plane C–H ring bending modes (900 cm to 675 cm ).
mine the magnetic and chemical environment of the phos-
Materials containing the alkyl groups possess absorptions for phorus atoms in zirconium bis(phosph(on)ate) materials and
the asymmetric and symmetric C–H stretching modes between confirm the presence of chemically distinct phosphorus atoms
1
−1
31
3
000 cm− and 2800 cm ; bands for the asymmetric and sym- in zirconium mixed phosph(on)ates. The isotropic P chemi-
−
1
metric methyl bending modes, near 1450 cm
and cal shifts for the zirconium phosph(on)ates studied are given
1
375 cm− , respectively; and an absorption for the methylene in Table 2. Also listed are the known isotropic P chemical
1
31
−
1 39
30
scissoring mode close to a value of 1465 cm
asymmetric methyl bending peak can sometimes be obscured nium bis(methylphosphonate) [Zr(Me) ],
2
by the methylene scissoring band.
.
Note that the shifts for zirconium bis(ethyl phosphate) [Zr(OEt)
2
], zirco-
and α-zirconium
for comparison with the zirconium
3
3,34
4
0
2
phosphate [Zr(OH) ]
3
1
As can be seen in Fig. 3, the IR spectra of Zr(ONap) and mixed phosph(on)ates. As examples, the solid-state P NMR
2
Zr(OEt)
2
2 2 2 1 1
display absorptions that are attributable to the respec- spectra of Zr(CH Nap) , Zr(ONap) , and Zr(ONap) (OEt)
3
1
tive functional group vibrations described above, confirming are shown in Fig. 4. The solid-state P NMR spectra of
that the desired organic groups are intact and present in these Zr(ONap)
(Me) and stg Zr(ONap) (OH) are shown in Fig. 3S.†
1 1 1
1
Dalton Trans.
This journal is © The Royal Society of Chemistry 2019

View Article Online
Dalton Transactions
Paper
3
1
Table 2 Solid-state P NMR isotropic chemical shifts for the zirconium phorus atoms in the two compounds, with the phosphorus
bis(phosph(on)ates), Zr(O PR) , and zirconium mixed phosph(on)ates,
Zr(O PR) (O PR’)
3
2
atoms in Zr(ONap) being much more shielded than those of
2
3
1
3
1
2 2 2
Zr(CH Nap) . This is understandable given that in Zr(ONap) ,
the phosphorus atoms are connected to an oxygen atom which
would increase the shielding due to the lone pairs on the
oxygen, as compared to the phosphorus atoms being con-
Compound
Zr(CH Nap)
δ
arene (ppm)
δalkyl (ppm)
2
2
5.2, 2.1
−29.5
Zr(ONap)
2
a
2 2 2
nected to a CH group in Zr(CH Nap) . More striking are
Zr(OEt)
2
−19.8, −20.8, −21.8
6.9
−18.7
7.0
−22.8
−21.6
b
Zr(Me)
Zr(OH)
2
their peak patterns. For Zr(ONap) , what is seen in the
2
c
2
31
P spectrum is typical of that of zirconium bis(phosph(on)ate)
Zr(ONap)
Zr(ONap)
1
(Me)
(OEt)
1
−28.1
−29.1
−30.0
compounds, in that a single, high-intensity isotropic reso-
1
1
d
1
stg Zr(ONap)
1
(OH)
nance (−29.5 ppm) along with multiple, lower-intensity spin-
3
1
a
b
c
ning sidebands is observed. In contrast, the P spectrum of
Zr(CH Nap) is unusual in that two isotropic resonances
with unequal intensities are observed at 5.2 ppm and at 2.1 ppm
Δ = 635 Hz). We do not believe that these two resonances are
due to spin–spin splitting. The hydrogens on the CH group
Taken from ref. 30. Taken from ref. 33 and 34. Taken from ref. 40.
Stg refers to the structural classification of a staged compound.
d
2
2
(
2
are the nearest NMR-active nuclei, and due to the two hydro-
gens on the methylene group, a spin triplet would be the
expected splitting pattern if these hydrogen atoms were coup-
3
1
1
ling to the phosphorus. Additionally, typical two-bond P– H
coupling constants are on the order of 10–20 Hz (ref. 39) and
the splitting between the two observed peaks in the spectrum
of Zr(CH Nap) is 635 Hz. Rather, the pair of resonances are
2
2
believed to be due to the distinct, “locked-in” conformations
of the naphthyl rings as displayed in Fig. 5.
The structural models for the multiple conformations of Zr
(
2 2
CH Nap) shown in Fig. 5 were obtained by consideration of
the results of low-level molecular mechanics calculations on
the corresponding phosphonic acid, 2-naphthylmethyl-
phosphonic acid (2NMPA). The molecular mechanics calcu-
lations were performed with the Discover energy minimization
and dynamics module within the InsightII v95 software package
(
Molecular Simulations, Inc., San Diego, CA) using the Central
41
Valences Force Field (CVFF). For 2NMPA, molecular mechanics
calculations suggest that there are three minimum-energy struc-
1 2 m
tures along the C –C –C –P dihedral angle coordinate, which is
one of the primary variable coordinates for this molecule. The
lowest-energy minimum is a structure in which the C –C –C –P
1
2
m
m 2
dihedral angle is 89.2° and the P–C –C angle is 113.0° (some-
2
what distorted from a pure sp conformation). The other two
minima are equal in energy to each other, and only 3.5 kcal
−1
mol higher than the lowest-energy-minimum structure. These
two structures both have a value of 120.2° for the P–C –C angle
and they differ in that one has a C –C –C –P dihedral angle of
m
2
1
2
m
0
.0° while the other has a dihedral angle of 180.0°.
In considering which of these structures might be reason-
3
1
Fig. 4 Solid-state P NMR spectrum of Zr(CH
2 2 2
Nap) , Zr(ONap) , and Zr
(
ONap) (OEt) . The peaks marked with an asterisk are spinning side- able models for the conformation of the methylnaphthyl
1
1
bands. The isotropic resonances associated with the methyl-
naphthylphosphonate, the naphthyl phosphate, and the ethyl phosphate
groups, are marked with δMNap, δONap, and δOEt, respectively.
groups in the Zr(CH
2 2
Nap) solid, the structure in which the
C
1
–C –C –P dihedral angle is 90.0° was ruled out because of
2
m
the lateral space restrictions of the neighboring arene mole-
cules which are estimated to be 5.3–5.4 Å apart based on the
refined crystal structure of zirconium bis(phenylphospho-
3
1
42
As can be seen from Table 2 and Fig. 4, the isotropic
P
nate). The cross-sectional size of a naphthalene ring (dis-
chemical shifts for the two zirconium bis(phosph(on)ate) com- tance from H to H or H to H in Fig. 5), is estimated to be
1
4
5
8
pounds examined in this study, Zr(CH
2
Nap)
2
and Zr(ONap)
2
,
1 2 m
approximately 5 Å. Therefore, the structures with C –C –C –P
indicate very different chemical environments for the phos- dihedral angles of 0.0° and 180.0° were considered the most
This journal is © The Royal Society of Chemistry 2019
Dalton Trans.

View Article Online
Paper
Dalton Transactions
mation shown in Fig. 5(a). Since the phosphorus environment
is chemically equivalent for the two conformations, the relax-
ation time for the phosphorus atom in each type of confor-
mation would expected to be the same. Therefore, using the
3
1
relative intensities from the P solid state NMR, we estimate
that the ratio of the two conformations present in the
Zr(CH
0% that of Fig. 5(a).
This type of structural observation (distinct conformations
2 2
Nap) solid is approximately 60% that of Fig. 5(b) and
4
for one type of organic functional group) has been reported
previously for zirconium bis(monochlorophosphonate),
4
3
Zr(O
method, Clearfield and co-workers were able to solve the
structure of zirconium bis(monochlorophosphonate),
Zr(O PCH Cl) , and found that there were two crystallographi-
3
PCH
2
Cl)
2
.
Using the Rietveld structure refinement
3
2
2
cally distinct chlorine atoms in the structure at the same x and
z coordinates but differing in their y coordinates. In other
words, they were related by a change in the dihedral angle
about the P–C bond. In our previous report on the chemical
and structural characterization of mixed arene/alkoxy zirco-
nium phosph(on)ates, we made a similar observation (multiple
3
1
isotropic resonances) in the P solid-state NMR of zirconium
bis(ethyl phosphate) and zirconium bis(butyl phosphate) that
were also attributed to “locked-in” conformations of the alkoxy
3
0
chains. The three isotropic resonances of zirconium bis(ethyl
phosphate) are listed in Table 2.
3
1
The P NMR spectrum of Zr(ONap)
spectrum of Zr(CH Nap) , only displays a single isotropic reso-
nance at −29.5 ppm, even though one might expect that
Zr(ONap) would also possess the two conformations described
Nap) . Molecular mechanics performed on
–P dihedral angle of the precursor, 2-naphthyl phosphate (2NP), which has an
2
, in contrast to the
2
2
2 2
Fig. 5 Idealized structural conformations of Zr(CH Nap) using (a) a P–
2
C
(
1
m
–C
b) a P–C
80.0°.
2
angle of 120.0° and a C
1
–C
2
–C
m
–P dihedral angle of 0.0° and above for Zr(CH
–C –C
2
2
m
–C angle of 120.0° and a C
2
1
2
m
oxygen atom between the phosphorus and 2-position carbon
of the ring, predicts that there is only one minimum-energy
structure along the C –C –O–P dihedral angle coordinate; a
1
2
reasonable for the distinct, “locked-in” conformations of the structure in which the dihedral angle has a value of 91.6° and
methylnaphthyl groups in Zr(CH Nap)
the P–O–C angle is at a value of 116.0°. The modeling calcu-
2
2
.
2
In order to understand how the presence of these two dis- lations clearly point to an absence of structures in which the
tinct conformations of the methylnaphthyl groups in Fig. 5 dihedral angle is at a value of 0.0° or 180.0° by not falling to
3
1
would lead to distinct P resonances, one has to consider the these local energy minima for the isolated 2NP precursor
effect of the aromatic ring in the two conformations on the molecule. This result is most likely due to shorter P–O and
neighboring phosphorus atoms rather than the effect on the O–C bond lengths in the minimum-energy structure of 2NP,
2
phosphorus atom that the methylnaphthyl group is attached which are 1.62 Å and 1.38 Å, respectively, as compared to the
to. Due to the close proximity of the aromatic ring, a phos- P–Cm and C –C bond lengths of 1.75 Å, and 1.54 Å, respect-
m
2
phorus atom to the left of the methylnaphthyl group pictured ively, in the minimum-energy structures of 2NMPA. The
in Fig. 5(a) would experience increased shielding due to the shorter bond lengths in 2NP apparently cause more steric
aromatic ring current anisotropy and the resonance of this repulsions at C –C –O–P dihedral angles of 0.0° and 180.0°
1
2
phosphorus atom would be expected to be shifted upfield than in 2NMPA and therefore the structures with these di-
when compared to the analogously neighboring phosphorus hedral angles are no longer energy minima in the isolated
atom in Fig. 5(b), in which the aromatic ring is farther away. 2NP.
This would be a case of magnetically inequivalent but chemi-
Even though the minimum energy structure for the isolated
cally equivalent phosphorus atoms. Based on this argument, 2NP possesses a C –C –O–P dihedral angle of approximately
1
2
3
1
the P resonance at 5.2 ppm can be assigned to phosphorus 90.0°, when it is bound to the zirconium layer in Zr(ONap)2,
atoms next to aromatic rings with the conformation shown in this dihedral angle will no longer be possible due to the lateral
Fig. 5(b) and the P resonance at 2.1 ppm can be assigned to space restrictions of 5.3–5.4 Å for the aromatic rings discussed
phosphorus atoms next to aromatic rings with the confor- above. Therefore the C –C –O–P dihedral angle would be
3
1
1
2
Dalton Trans.
This journal is © The Royal Society of Chemistry 2019

View Article Online
Dalton Transactions
Paper
forced to be close to the value of 0.0° or 180.0°. Due to the Table 3 Experimental and predicted interlayer spacing of the zirconium
bis(phosph(on)ates), Zr(O
Zr(O PR) (O PR’)
3
PR)
2
, and zirconium mixed phosph(on)ates,
shorter P–O and O–C2 bond lengths mentioned above, it is
expected that there would be considerable steric repulsions of
the ring protons with the zirconium layer oxygen atoms for
3
1
3
1
a
Compound
Zr(CH Nap)
d001 (Å) Expt
d001 (Å) Pred
both the 0.0° or 180.0° conformations of Zr(ONap) . However,
2
2
2
20.0
19.6
19.7
19.1
11.9
8.4
because there would only be one proton interacting with the
zirconium layer oxygen atoms for the 180.0°, it is thought that
Zr(ONap)
Zr(OEt)
2
b
c
2
11.6 , 11.7
d
e
this would be the energetically favored conformation in the Zr(Me)
2
8.8 , 8.8
f
Zr(OH)
2
7.6
7.8
2
interlayer region of Zr(ONap) and therefore we believe that all
Zr(ONap)
Zr(ONap)
1
1
1
(Me)
(OEt)
(OH)
1
15.3
13.8
15.5
13.4
27.2
(or nearly all) of the naphthoxy groups exist in a conformation
1
16.1
14.3
28.2
with a dihedral angle close to 180.0°, as in Fig. 5(b) but with Zr(ONap)
1
g
h
1 1
stg Zr(ONap) (OH)
an oxygen linkage rather than via a methylene group.
a
This assignment also is most consistent with the experi-
Predicted value based on the simple geometrical model described in
b
c
d
mental d spacing of Zr(ONap) . The experimental finding that the text. Taken from ref. 30. Taken from ref. 5. Taken from ref. 33.
2
e
f
g
Taken from ref. 44. Taken from ref. 1. Stg refers to the structural
serves as indirect evidence for the idea that there is only one
conformation in Zr(ONap) is an observed disparity in the
FWHM of the 001 XRD peaks (vide infra): 0.17° for Zr(ONap)2 [7.6 Å].
and 0.33° for Zr(CH Nap) . Since the crystallinity of zirconium
h
classification of a staged compound. Predicted value based on the
sum of the experimental d spacings of Zr(ONap) [19.6 Å] and Zr(OH)
2
2
2
2
2
phosph(on)ates, which is reflected by the FWHM of the 001
XRD peak, does change with HF concentration and reflux
time, the two syntheses for the FWHM comparison were
carried out using identical amounts of HF and reflux times.
Therefore the decreased FWHM for Zr(ONap) can be taken as
2
an indication of more efficient packing of the organic groups
in Zr(ONap)
the one-conformation vs. two-conformation hypothesis.
For the mixed systems, Zr(ONap) (Me) , Zr(ONap)
, the P resonances are sufficiently
2 2 2
as compared to such in Zr(CH Nap) , in line with
1
1
1 1
(OEt) ,
3
1
1 1
and stg Zr(ONap) (OH)
resolved that identification and confirmation of the two
chemically distinct phosphorus atoms was possible. Due to
the method of synthesizing the mixed compound Zr
(
ONap) (OH) , a sufficient amount of this material was not
1 1
3
1
produced to permit a P solid-state NMR to be obtained. It is
worth noting that the resonances for the naphthyl phosphate
phosphorus atoms in the mixed compounds are slightly
2
shifted with respect to the resonances of Zr(ONap) . We
believe this is due to slight perturbations in the magnetic
environment of the phosphorus atoms in the mixed com-
pounds, which now have different neighboring groups as com-
2
pared to that of Zr(ONap) . Also, the three ethyl phosphate
resonances observed for Zr(OEt)2 have now collapsed into a
single isotropic resonance and shifted slightly downfield. This
is presumably due to restricted conformational freedom of the
ethyl group as a result of the presence of the large, bulky
naphthalene ring within the interlayer.
Fig. 6 X-ray diffraction patterns displaying the 00n peaks of Zr(ONap)
Zr(ONap) (OEt) , Zr(ONap) (OH) , and stg Zr(ONap) (OH)
2
,
3
.2 Structural investigations
1
1
1
1
1
1
.
For crystalline zirconium phosph(on)ates, the interlayer dis-
tance or d spacing of the zirconium planes can be determined
from the 00n peaks in the powder X-ray diffraction (XRD) of the ESI.† The d-space values for zirconium bis(ethyl
5
,30
pattern via the Bragg equation, nλ = 2d sin θ. The d-space phosphate) [Zr(OEt)
2
],
and α-zirconium phosphate [Zr(OH)
in this study are given in Table 3. The XRD patterns been included in Table 3 in order to allow a comparison of the
of Zr(ONap) Zr(ONap) (OEt) Zr(ONap) (OH)
and stg d-space values for the zirconium mixed phosph(on)ates to the
Zr(ONap) (OH)
are shown in Fig. 6, while the XRD patterns zirconium bis(phosphonates). As can be seen from Fig. 6 and
of Zr(CH Nap) and Zr(ONap) (Me) are given in Fig. 4S Fig. 4S,† most of the XRD patterns display a strong 001 peak
zirconium bis(methylphosphonate)
3
3,44
1
values for the zirconium phosph(on)ate materials synthesized [Zr(Me)
2
],
2
] have also
2
,
1
1
,
1
1
,
1
1
2
2
1
1
This journal is © The Royal Society of Chemistry 2019
Dalton Trans.

View Article Online
Paper
Dalton Transactions
(lowest-angle peak in the pattern) followed by higher order 00n domly distributed structure will show a single 001 peak at a
peaks at larger angles and lower intensities. The large-angle d-space value equal to or between those of the two parent zir-
region of the XRD patterns, which is not shown in the figures, conium bis(phosph(on)ates). The ordered, segregated struc-
consist of lower intensity, unresolved peaks.
.2.1 Bis(phosph(on)ates). The d spacings observed for the spacing is that of a sum of the two parent zirconium bis
two zirconium bis(phosphonates) synthesized in this study, (phosph(on)ates). The third possibility, distinct co-crystallites,
Zr(ONap) and Zr(CH Nap) , are 19.6 and 20.0 Å, respectively. will give rise to an XRD pattern that contains two 001 peaks,
ture, on the other hand, will display an 001 peak whose d
3
2
2
2
These are similar to the d spacing of 19.1 Å we observed for zir- one peak at the d spacing for one of the parent zirconium bis
conium bis(naphthylphosphonate) in our previous report on (phosph(on)ates) and the other peak at the value for the other
the chemical and structural characterization of mixed arene/ parent zirconium bis(phosph(on)ate), as if for a simple physi-
alkoxy zirconium phoshp(on)ates, which also contained a cal mixture of the two compounds.
3
0
naphthalene ring system in the interlayer. A simple, ideal-
As can be seen from an inspection of Table 3, the magni-
can be used to obtain tudes of the d spacings of Zr(ONap) (Me) , Zr(ONap) (OEt)
predicted d spacings based on twice the zirconium phospho- and Zr(ONap) (OH) are between the d-space values of their
3
0,45
ized geometrical modeling method
1
1
1
1
,
1
1
1
,5
nate layer thickness (3.2 Å), twice the normal distance (pro- parent zirconium bis(phosph(on)ates), indicative of single-
jection along the P–C or P–O bond axis) from the phosphorus phase, randomly mixed structures for each of these materials.
atom to the farthest hydrogen atom in the organic functional In the case of Zr(ONap) (OH) , it is worth noting that this ran-
1
1
group, plus twice the magnitude of the hydrogen atomic domly mixed material was not formed by the direct co-precipi-
radius (0.7 Å). Predicted d-space values can be obtained for tation method but rather was obtained in a circuitous route via
both of the structural conformations depicted in Fig. 5 for the pyrolysis of the ethyl groups in the randomly mixed
both Zr(ONap) and Zr(CH Nap) in order to determine which Zr(ONap) (OEt) material. This was done because, as will be
2 2 2 1 1
conformational model is most consistent with the observed d discussed below, direct co-precipitation of the two precursor
spacings. For Zr(CH Nap) , the conformation displayed in phosphates with zirconyl chloride leads to a segregated or
2
2
Fig. 5(a) yields a predicted d spacing of 15.5 Å while the confor- staged structure for the resulting solid.
mation displayed in Fig. 5(b) yields a predicted d spacing of For the purpose of calculating the predicted interlayer spa-
9.7 Å. A similar calculation can be performed for Zr(ONap)2 cings for these random-structure zirconium mixed (phosph
after replacing the methylene group with an oxygen, adjusting (on)ates), we employed the same type of primitive predictive
1
4
5
the P–O and C
using a P–O–C2 angle of 120.0° to be consistent with the zirconium bis(phosph(on)ates). Namely, it is presumed that
Zr(CH Nap) calculations. With this analysis, one obtains pre- the naphthalene rings are opposite a methyl, ethoxy, and
dicted values of 14.9 Å for the analogous conformation to that hydroxyl group (depending on the functional group that is
depicted in Fig. 5(a) and 19.1 Å for the analogous confor- being mixed), which would allow “collapse” of the
2
–O bond lengths to their standard values, and model that was used to estimate the interlayer spacings of the
2
2
a
mation to that depicted in Fig. 5(b). Based on the comparison “scaffolded” structure to a lower d-space value than for the zir-
of the experimental d spacings with both of the predicted d conium bis(phosph(on)ates). Under these assumptions, one
spacings for Zr(ONap) and Zr(CH Nap) , the structural model obtains estimated d-space values of 13.8, 15.5, and 13.4 Å for
2
2
2
with the conformation displayed in Fig. 5(b) is the one Zr(ONap)
1 1 1 1 1 1
(Me) , Zr(ONap) (OEt) , and Zr(ONap) (OH) respect-
most consistent with the experimental d-space values for each ively. As can be seen by comparing these predicted d-space
3
1
compound. As discussed above in the P solid-state NMR values to the experimental values, the predicted values are
section, for Zr(CH Nap) methylnaphthyl groups in the solid lower than the experimental values by 0.6–1.5 Å. A likely expla-
that possess the conformation shown in Fig. 5(a) are present at nation for the deviations is that with a stoichiometrically equal
40% abundance. However, the experimental d-space number of naphthoxy groups and methyl, ethoxy, and hydroxyl
2
2
a
measurements suggest that the 60% presence of conformer groups, i.e., a compound with x = 1, there are a sufficient
Fig. 5(b) is sufficient to exhibit the higher d spacing as the number of naphthalene rings that are near another naphtha-
overall value.
.2.2 Mixed phosph(on)ates. For
lene ring from the opposite layer, essentially “propping up”
zirconium mixed the interlayer distance up to a slightly larger value than that
3
a
phosph(on)ate, there are generally three different structural based on the simple prediction. Although zirconium mixed
possibilities for the arrangement of the organic moieties: (i) a phosph(on)ates with x values less than 1 (i.e., a preponderance
random distribution within the interlayer (all layers therefore of the smaller pendant groups) were not synthesized for the
equivalent), (ii) an ordered segregation or staging of the func- present series of compounds, we envision that d-space values
tional groups in an alternating fashion (adjacent layers not for such stoichiometries would be closer to the primitive
equivalent), and (iii) precipitation into co-crystallites or predictions.
4
+
domains of the bis(phosph(on)ates), in other words two indi-
As mentioned above, direct co-precipitation of Zr with
vidual phases. These three basic structural possibilities are naphthyl phosphate (O POC H ) and phosphate groups
3
10
7
represented schematically in Fig. 1.
3
(O POH) does not lead to a randomly mixed structure in which
Each of these situations will give rise to distinct behavior the 001 diffraction peak is at a d-space value equal to or
which will be evident in the XRD powder patterns. The ran- between those of the two parent zirconium bis(phosph(on)
Dalton Trans.
This journal is © The Royal Society of Chemistry 2019

View Article Online
Dalton Transactions
Paper
ates), Zr(CH
2
Nap)
2
and Zr(ONap)
2
,
and four zirconium
mixed phosph(on)ates, Zr(ONap) (Me) , Zr(ONap) (OEt) ,
1
1
1
1
Zr(ONap)
1
(OH)
1
, and stg Zr(ONap)
1
1
(OH) .
Chemically, the compositions of all materials synthesized
and studied were confirmed by TGA within an acceptable
experimental error of 2–20%. The presence of the desired func-
tional groups in all of the materials was determined by IR
3
1
spectroscopy. Solid-state P NMR spectroscopy revealed the
presence of magnetically inequivalent phosphorus environ-
2 2
ments (multiple resonances) for Zr(CH Nap) , which is attribu-
ted to the presence of two conformationally distinct, “locked-
in” geometries of the methylnaphthyl groups involving the di-
2
hedral angle about the CH group (0.0 and 180°), whereas
Zr(ONap) only displayed one dominant phosphorus environ-
2
ment (single resonance), as expected for zirconium bis
(
phosph(on)ates). The lack of multiple conformations for
Zr(ONap) is attributed to increased steric interactions between
2
Fig. 7 Idealized structural representation of stg Zr(ONap)
showing the predicted sum of d-space values for a staged compound.
1
(OH)
1
the naphthalene ring and the zirconium phosphate layer
oxygen atoms for the 0.0° conformation, due to the shorter
P–O and C–O bonds, thereby giving rise to only one favored
conformation (180°) in the interlayer region. Solid-state
P NMR spectroscopy was also used to confirm the presence
of two chemically distinct phosphorus environments for
3
1
ates), but rather leads to a structure that produces the XRD
pattern labeled stg Zr(ONap) (OH) in Fig. 6. The d-space value
1
1
Zr(ONap)
Zr(ONap)
1
1
(Me)
(OH)
1
, Zr(ONap)
1 1 1 1
(OEt) , Zr(ONap) (OH) , and stg
for 001 peak of this compound is 28.2 Å, which is clearly larger
1
. These results thus indicate a new structural
than the d-space values of either Zr(ONap) or Zr(OH) (19.6 Å
2
2
type of layered materials: a mixed zirconium phosph(on)ate in
which the homogeneous distribution within the interlayer
region refers not to chemical composition, but instead
denotes a mix of conformers.
and 7.6 Å, respectively). This indicates that this material pos-
sesses a staged or segregated structure in which the repeating
unit consists of layers of alternating composition; an idealized
model of this structure is displayed in Fig. 7. A similar type of
staging or segregation to this has been observed previously for
a zirconium mixed phosph(on)ate material involving phenyl
Structurally, the interlayer spacings of Zr(CH
2 2
Nap) and
Zr(ONap) , 19.6 and 20.0 Å, respectively, were found to be con-
2
1
6,17
sistent with the general size of the naphthalene pendant group
when compared to the previously reported interlayer spacing
for a similar compound containing naphthalene rings, zirco-
and hydroxyl moieties.
Based on the experimental d-space
, the expected d-space value
values for Zr(ONap) and Zr(OH)
2
2
for the idealized staged structure displayed in Fig. 7 is 27.2 Å,
which is close to the experimentally observed value of 28.2 Å.
The discrepancy between the two values is thought to be a
result of possible incorporation of a layer of coordination
water molecules within the hydroxy layer, leading to a slightly
larger distance than 7.6 Å for that portion of the structure or
slight differences in hydration between this mixed compound
and the parent compounds. The reason that the two functional
groups segregate into alternate layers when directly precipi-
tated is believed to be due to the radical difference in the
nature of the two groups, one being extremely hydrophobic
3
0
nium bis(naphthylphosphonate). Using a simple, idealized
geometrical modeling method, predicted d space values were
calculated based on the two possible conformations (0.0
and 180°) of the ring systems and we find that the models
with the 180° conformations are the most consistent with
2 2
the experimental d space values for both Zr(CH Nap) and
Zr(ONap) . For the mixed compounds Zr(ONap) (Me) and
2
1
1
Zr(ONap)
1
1
(OEt) , both were determined to be single-phase
compounds that possess structures with random distributions
of the two organic pendant groups within the interlayer, with
d-space values of 15.3 and 16.1 Å, respectively. Two different
mixed compounds were synthesized that both incorporated
the ONap and OH functional groups in approximately equal
(ONap) and the other (OH) being extremely hydrophilic, there-
fore giving rise to an “immiscibility” of the two groups within
the solid. This property is brought out during the slow-growth,
six-day self-assembly process detailed in the Experimental
section.
1 1
amounts: Zr(ONap) (OH) in which the two pendant groups
are randomly distributed (synthesized by controlled pyrolysis
of Zr(ONap) (OEt) ), d-space value of 14.3 Å, and stg
1
1
Zr(ONap)
1
(OH)
1
in which the pendant groups are segregated
into separate layers (synthesized by direct co-precipitation),
d-space value of 28.2 Å. The segregated structure in the latter
4
. Conclusions
Six novel zirconium phosph(on)ates incorporating naphtha- material is attributed to the “immiscibility” of the two groups
lene rings have been synthesized, chemically characterized, within the solid due to the very different hydrophobicities of
and structurally examined: two zirconium bis(phosph(on) the two functional groups.
This journal is © The Royal Society of Chemistry 2019
Dalton Trans.

View Article Online
Paper
Dalton Transactions
1
6 J. D. Wang, A. Clearfield and G. Z. Peng, Mater. Chem.
Phys., 1993, 35, 208–216.
Conflicts of interest
There are no conflicts to declare.
17 A. Clearfield, J. D. Wang, Y. Tian, E. Stein and C. Bhardwaj,
J. Solid State Chem., 1995, 117, 275–289.
1
8 J. C. Amicangelo and W. R. Leenstra, J. Am. Chem. Soc.,
998, 120, 6181–6182.
9 A. Clearfield, J. Mol. Catal., 1984, 27, 251–262.
1
Acknowledgements
1
Financial support of this work was provided by the University 20 G. Alberti, Acc. Chem. Res., 1978, 11, 163–170.
of Vermont through UCRS Grant No. PSCI94-2. We thank 21 Inorganic Ion Exchange Materials, ed. A. Clearfield, CRC
Dr Christopher Landry for the use of the Scintag X1 automated
Press, Boca Raton, FL, 1982, ch. 1–3.
X-ray powder diffractometer.
22 G. Alberti and U. Costantino, in Intercalation Chemistry, ed.
M. S. Whittingham and A. J. Jacobsen, Academic,
New York, 1985, ch. 5.
2
3 L. A. Vermeulen and M. E. Thompson, Chem. Mater., 1994,
References
6, 77–81.
1
2
3
4
5
A. Clearfield and G. D. Smith, Inorg. Chem., 1969, 8, 431– 24 J. D. Burns, A. Clearfield, M. Borkowski and D. T. Reed,
36. Radiochim. Acta, 2012, 100, 381–387.
J. M. Troup and A. Clearfield, Inorg. Chem., 1977, 16, 3311– 25 J. D. Burns, T. C. Shehee, A. Clearfield and D. T. Hobbs,
314. Anal. Chem., 2012, 84, 6930–6932.
J. Albertsson, A. Oskarsson, R. Tellgren and J. O. Thomas, 26 B. Bujoli, P. Janvier and M. Petit, in Metal Phosphonate
4
3
J. Phys. Chem., 1977, 81, 1574–1578.
E. R. Losilla, M. A. G. Aranda and S. Bruque, J. Solid State
Chem., 1996, 125, 261–269.
Chemistry: From Synthesis to Applications, ed. A. Clearfield
and K. Demadis, The Royal Society of Chemistry, London,
2011, ch. 13.
G. Alberti, U. Costantino, S. Allulli and N. Tomassini, 27 A. Díaz, M. L. González, R. J. Pérez, A. David, A. Mukherjee,
J. Inorg. Nucl. Chem., 1978, 40, 1113–1117.
L. Maya, Inorg. Nucl. Chem. Lett., 1979, 15, 207–212.
A. Báez, A. Clearfield and J. L. Colón, Nanoscale, 2013, 5,
11456–11463.
6
7
M. B. Dines and P. M. DiGiacomo, Inorg. Chem., 1981, 20, 28 J. González-Villegas, Y. Kan, V. I. Bakhmutov, A. García-
9
2–97.
Vargas, M. Martínez, A. Clearfield and J. L. Colón, Inorg.
Chim. Acta, 2017, 468, 270–279.
8
A. Clearfield, in Progress in Inorganic Chemistry, ed.
K. D. Karlin, John Wiley and Sons, New York, 1998, vol. 47, 29 J. C. Amicangelo and W. R. Leenstra, J. Am. Chem. Soc.,
pp. 371–510. 2003, 125, 14698–14699.
A. Clearfield, in Metal Phosphonate Chemistry: From 30 J. C. Amicangelo and W. R. Leenstra, Inorg. Chem., 2005,
Synthesis to Applications, ed. A. Clearfield and 44, 2067–2073.
K. Demadis, The Royal Society of Chemistry, London, 2011, 31 D. C. Stewart and H. W. Crandall, J. Am. Chem. Soc., 1951,
ch. 1. 73, 1377–1378.
0 R. Vivani, F. Costantino and M. Taddei, in Metal 32 U. Costantino and A. L. Ginestra, Thermochim. Acta, 1982,
Phosphonate Chemistry: From Synthesis to Applications, 58, 179–189.
ed. A. Clearfield and K. Demadis, The Royal Society of 33 W. R. Leenstra and J. C. Amicangelo, Inorg. Chem., 1998,
Chemistry, London, 2011, ch. 2. 37, 5317–5323.
1 M. B. Dines, P. M. DiGiacomo, K. P. Callahan, P. C. Griffith, 34 J. C. Amicangelo, Ph.D. Dissertation, University of
R. H. Lane and R. E. Cooksey, in Chemically Modified Vermont, 1999.
Surfaces in Catalysis and Electrocatalysis, ed. J. S. Miller, ACS 35 A. Clearfield and J. A. Stynes, J. Inorg. Nucl. Chem., 1964,
9
1
1
Symposium Series 192, American Chemical Society,
Washington, DC, 1982, ch. 13.
2 G. Alberti, U. Costantino and R. Guiletti, Gazz. Chim. Ital.,
26, 117–129.
36 C. Y. Yang, Ph.D. Dissertation, Texas A&M University,
1986.
1
1
1
1
1
983, 113, 547–552.
37 S. E. Horsely, D. V. Nowell and D. T. Stewart, Spectrochim.
Acta, Part A, 1974, 30, 535–541.
38 X. Mathew and V. U. Nayar, Infrared Phys., 1988, 28, 189–
194.
39 R. M. Silverstein, F. X. Webster and D. J. Kiemle,
Spectrometric Identification of Organic Compounds, John
Wiley and Sons, New York, 2005.
3 G. Alberti, U. Costantino, J. Kornyei and M. L. Giovagnotti,
React. Polym., 1985, 4, 1–10.
4 G. Alberti, U. Costantino and G. Perego, J. Solid State
Chem., 1986, 63, 455–460.
5 G. Alberti, in Solid-State Supramolecular Chemistry: Two-
and Three-Dimensional Inorganic Networks, ed. G. Alberti
and T. Bein, Comprehensive Supramolecular Chemistry, 40 N. J. Clayden, J. Chem. Soc., Dalton Trans., 1987, 1877–
vol. 7, ed. J. L. Atwood, J. E. D. Davies, D. D. MacNicol, 1881.
F. Vögtle, and J.-M. Lehn, Elsevier Science Ltd., Tarrytown, 41 (a) S. Lifson, A. T. Hagler and P. Dauber, J. Am. Chem. Soc.,
NY, 1996, ch. 5.
1979, 101, 5111–5121; (b) A. T. Hagler, S. Lifson and
Dalton Trans.
This journal is © The Royal Society of Chemistry 2019

View Article Online
Dalton Transactions
Paper
P. Dauber, J. Am. Chem. Soc., 1979, 101, 5122–5130; 44 M. B. Dines and P. C. Griffith, J. Phys. Chem., 1982, 86, 571–576.
c) A. T. Hagler, P. Dauber and S. Lifson, J. Am. Chem. Soc., 45 Except for the specific angles mentioned in the text,
(
1
979, 101, 5131–5141.
all other bond distances and angles used in the
idealized structures were the standard values given in the
CRC Handbook of Chemistry and Physics, ed. R. C. Weast,
CRC Press, Boca Raton, FL, 70th edn, 1989, pp. F-188–
F-192.
4
2 M. D. Poojary, H. L. Hu, F. L. Campbell and A. Clearfield,
Acta Crystallogr., Sect. B: Struct. Sci., 1993, 49, 996–1001.
3 D. M. Poojary, C. Bhardwaj and A. Clearfield, J. Mater.
Chem., 1995, 5, 171–174.
4
This journal is © The Royal Society of Chemistry 2019
Dalton Trans.