A fast, high-accuracy universal polarimeter was developed using a charge-coupled device (CCD) spectrometer (CCD-HAUP), to carry out simultaneous optical anisotropic (linear birefringence, LB; linear dichroism, LD) and chiroptical (circular birefringence, CB; circular dichroism, CD) measurements on single crystals without any pretreatment, in the visible region between 400–680 nm. The principle of the HAUP method is to measure the intensities of emergent light passing through a polarizer, a crystal sample, and then an analyzer, as the azimuth angles of the polarizer and analyzer are independently altered. The CCD-HAUP has the unique feature that white transmitted light intensity can be measured using a CCD spectrometer, compared with the generalized HAUP (G-HAUP) system in which monochromatic transmitted light is measured using a photomultiplier. The CCD-HAUP measurements across the entire wavelength region are completed within the G-HAUP measurement time for a single wavelength. The CCD-HAUP drastically reduces the measurement time for a dataset to only 1.5 h, from the 24 h required for the G-HAUP system. LB, LD, CB, and CD measurements of single crystals of α-quartz and enantiomeric photomechanical salicylidenephenylethylamines before, during, and after ultraviolet light irradiation show results comparable to those obtained using the G-HAUP system. The newly developed system is very effective for samples susceptible to degradation induced by external stimuli, such as light and heat.

Solid-state chiral spectroscopy offers unique insight into complex chiral systems, which cannot be obtained by conventional solution spectroscopy. Every chiral crystal has an inherent optical activity, exhibiting chiroptical properties of circular birefringence (CB) and circular dichroism (CD).1 Moreover, many chiral and achiral crystals have optical anisotropy, such as linear birefringence (LB) and linear dichroism (LD). Measurement of these four optical properties is necessary for evaluating the optical response of chiral crystals. However, simultaneous measurement of CB, CD, LB, and LD is extremely difficult in practice; the CB and CD signals are overwhelmed by the LB and LD signals because the latter are 102–103 times larger than the former.2 Hence, conventional polarimeters and CD spectrophotometers cannot be applied to accurately measure CB and CD, except in one specific case, the measurement along the optic axis.

The high-accuracy universal polarimeter (HAUP) was developed in 19833 to measure the LB and CB of various crystals, such as glutamic acid4 and chiral cocrystals.5 The HAUP method measures the intensities of emergent light passing through a polarizer, a crystal sample, and then an analyzer as the azimuth angles of the polarizer and analyzer are independently altered. The extension of this method to the measurement of anisotropic colored materials has made possible the simultaneous measurement of LB, CB, LD and CD.6–10 Recently, the HAUP apparatus has been generalized using a conventional Xe lamp and a monochromator to obtain ultraviolet (UV) and visible region spectra (300–680 nm);11 monochromatic transmitted light is detected by a photomultiplier. Generalized HAUP (G-HAUP) has subsequently been applied to intercalated K4Nb6O17 crystals with an azobenzene derivative,11 and to laminated collagen membranes.12 

We recently reported the LB, LD, CB, and CD properties in photomechanical crystals of chiral salicylidenephenylethylamines before and during UV irradiation utilizing the G-HAUP.13 However, 100 transmitted light measurements were required for each individual single wavelength and a fine wavelength interval was required to measure small changes in LB, LD, CB, and CD spectra; thus, the measurements are time-consuming. Even under weak UV irradiation, the thickness of the chiral salicylidenephenylethylamine crystal decreased (by approximately –10%) during the three days required for the HAUP measurements, due to gradual sublimation from the top surface of the specimens via the photothermal effect under UV irradiation, despite temperature control with a thermostat. When LB, LD, CB, and CD spectra were obtained from raw experimental data via HAUP measurements, it was necessary to compensate for the decreasing thickness of the specimen. Hence, improvement of the G-HAUP apparatus is required to shorten the measurement time.

To achieve simultaneous LB, LD, CB, and CD measurements in a short time, we developed a fast, high-accuracy universal polarimeter using a charge-coupled device spectrometer (CCD-HAUP). This novel instrument is equipped with a CCD spectrometer that enables detection across all wavelengths of white transmitted light simultaneously. The CCD-HAUP measurements across the entire wavelength region are completed within the G-HAUP measurement time for a single wavelength. In this paper, we report the instrumentation and performance of the CCD-HAUP and applications to α-quartz and chiral photomechanical salicylidenephenylethylamine crystals.

The principle of the HAUP method is measurement of the intensities of emergent light passing through a polarizer, a crystal sample, and then an analyzer, as the azimuth angles of the polarizer and analyzer are independently altered. The basic principle of the HAUP method is explained here; details of the principle of the original3,14,15 and extended HAUP methods have been described in detail elsewhere.6,11,16 The proposed HAUP system is shown in Figure 1. The CCD-HAUP employs a simple optical system that consists of only two optical elements: a polarizer (P) and an analyzer (A). The vibrational direction of the transmitted light through P is orthogonal to that passing through A (i.e., the crossed Nicols position).

FIG. 1.

Schematic drawing (a) and photograph of the CCD-HAUP apparatus (b). Sample stage has a temperature control unit. Here, θ represents the azimuth angle from an arbitrary origin and Y represents the deflecting angle of A from the crossed Nicols position.

FIG. 1.

Schematic drawing (a) and photograph of the CCD-HAUP apparatus (b). Sample stage has a temperature control unit. Here, θ represents the azimuth angle from an arbitrary origin and Y represents the deflecting angle of A from the crossed Nicols position.

Close modal

Due to this simple optical configuration, systematic errors, except those relating to P and A, are excluded.15 In the HAUP method, systematic errors originating from parasitic ellipticities of P and A (p and q, respectively) and a small error angle (δY, attributed to the displacement of the crossed Nicols configuration) are evaluated and eliminated. Here, we define θ as an azimuth angle of P from an arbitrary origin, Y as the azimuth angle of A from the crossed Nicols position of the arbitrary origin of P, Y’ as the azimuth angle of A from δY, θ0 as an extinction position angle of P from the arbitrary origin obtained by

((I/I0)θ)Y=0=0,

and θ’ as the azimuth angle of P from θ0. That is, θ=θ0+θ’ and Y=δY+Y’. The values of θ’ and Y’ can be measured accurately in the practical HAUP experiment. The clockwise direction from the viewpoint of an observer is defined as the positive direction for all angles.

From the expressed P, S, and A in the Jones matrix as written in the supplementary material, the ratio, Γ, of the intensity of transmitted light and intensity of incident light, I and I0, respectively, are represented as follows:

Γ(θ,Υ)=I/I0=A(θ)+B(θ)Υ+CΥ2,
(1)
A(θ)=H11+Hθ12+Hθ213,
(2)
B(θ)=H21+Hθ22,
(3)
C=H31,
(4)

where

H11 independent of θ and Υ,

H12=0,
H13=eE+eE2cosΔ,
(5)
H21=b1p+b2q+a1δΥ+2c1k2c2(sinΔ)k,
(6)
H22=2(eEcosΔ),
(7)
H23=eE,
(8)

where

a1=2sin2ΔeE+eE2cosΔ,
b1=2(eEcosΔ)sinΔeE+eE2cosΔ,
b2=2(eEcosΔ)sinΔeE+eE2cosΔ,
c1=KK2+1=E/Δ(E/Δ)2+1,
c2=1K2+1=1(E/Δ)2+1.

The extinction position angle, θ0, can be expressed under the same approximation as follows:

θ0=a2(p+q)b2δΥ+c1k+c2k+N,
(9)

where

a2=sinΔeE+eE2cosΔ,
b2=eEcosΔeE+eE2cosΔ.

Here, N is the difference of θ0 from its absolute value, Δ and E represent the phase difference and the total LD of the sample, respectively, and k and k’ represent the ellipticity derived from the CB and CD of the sample, respectively. Using these quantities, LB, LD, CB, and CD are expressed as follows:

LB=nsnfΔλ2πd,
(10)
LD=msmfEλ2πd,
(11)
CB=nLnRΔkλπd=2kLB,
(12)
CD=mLmRΔkλπd=2kLB,
(13)

where subscripts s and f are axes of the slow and fast light rays, subscripts R and L are the right and left circular polarization, n is the refractive index, m is the absorption coefficient, λ is the wavelength of incident light, and d is the thickness of the sample, respectively.

In the HAUP method, the values of LB, LD, CB, and CD are determined by the following procedure:

  • The intensities of light, I, are measured as quadratic functions of θ and Υ. The values of I0 · H″ij (i, j = 1, 2, 3) at each θ position and θ0 are determined by least-squares fittings using Eqs. (1)–(4). The values of Δ, E and I0 are calculated from the θ dependences of H″13, H″22 and H″31, i.e., Eqs. (5), (7) and (8). It should be noted that this procedure provides only the recorded phase difference, Δr(0Δrπ). The real value of Δ is obtained using the relationship Δ=2nπ±Δr, where n is usually measured using an Ehringhaus compensator. From Eqs. (10) and (11), LB and LD were obtained by normalizing with the sample thickness.

  • For extracting CB and CD from experimental data, reasonable procedures for removing the systematic errors are essential for the HAUP method. Because the value of p is independent of the sample setting, it has already been determined by a preceding experiment with an achiral crystal LiNbO3 or MgF2; using this procedure, HAUP measurements of various crystals have been successfully conducted.6,11,12,17,18 For this study, the p value was determined as after mentioned by the wavelength dependence G-HAUP measurements of an achiral MgF2 crystal. By assuming that k is dependent on λ and k′ = 0 in the wavelength regions, where the sample has no absorption, the optimum values of q, δY, and N are obtained by least-squares fittings using Eqs. (6) and (9) and the p value. The values of k and k′ are calculated by introducing the resulting systematic errors, Δ and E to Eqs. (6) and (9). Finally CB and CD are determined from Eqs. (12) and (13).

The CCD-HAUP system (Figure 1) is equipped with a 150-W xenon discharge lamp (BSO-X150, Bunkoukeiki) as a light source. White light passes through a polarizer P, sample S, and analyzer A. Glan–Thompson prisms made of CaCO3 crystal purchased from Karl Lambrecht were used as P and A. The azimuth angles of the polarizer and analyzer were independently and precisely altered using precision rotation stages (RA10A-W, Kohzu Precision). The sample, S, was fixed on a thin Cu plate with a small opening with a diameter of about 0.5 mm to pass the light beam. The white light intensity transmitted through P, S, and A is detected at once by a CCD spectrometer (Glacier X, B&W Tek). This is the most significant change in the optical system from the G-HAUP in which the monochromatic light intensity transmitted through P, S, and A is detected by a photomultiplier and a lock-in amplifier. Using a CCD spectrometer, transmitted light in the range 350–680 nm can be measured simultaneously. This optical system allows us to quickly measure the LB, LD, CB, and CD spectra in the wavelength region from 350 to 680 nm, because the entire measurement of the CCD-HAUP is completed within the measurement time of the G-HAUP for a single wavelength. The sample temperature was controlled with a Peltier element and a proportional-integral-derivative (PID) control unit (Bunkoukeiki) in the range from 15 to 80°C.

To evaluate the performance of the CCD-HAUP system, HAUP measurements in the range 425–600 nm were performed on an achiral MgF2 crystal as a standard sample, which was cut in the plane perpendicular to the a axis with a thickness of 52.6 μm, and placed on a thin Cu plate with a pinhole 0.7 mm in diameter. The raw experimental results (Δ, E, H″21, and θ0) from the analysis of the MgF2 crystal using the CCD-HAUP and G-HAUP7 methods were comparable (Figure 2). These four parameters are very important for obtaining LB, LD, CB, and CD, as described in the previous chapter; hence, LB, LD, CB, and CD should be measured using the CCD-HAUP system.

FIG. 2 .

Raw experimental results from the analysis of an achiral MgF2 crystal on the (100) face measured using the G-HAUP (black open circle) and CCD-HAUP (black solid circle) at 20° C: (a) the phase difference (real value), (b) total linear dichroism, (c) H21 and (d) extinction position angle.

FIG. 2 .

Raw experimental results from the analysis of an achiral MgF2 crystal on the (100) face measured using the G-HAUP (black open circle) and CCD-HAUP (black solid circle) at 20° C: (a) the phase difference (real value), (b) total linear dichroism, (c) H21 and (d) extinction position angle.

Close modal

The accuracy of the Δ, E, H″21, and θ0 values were estimated to be 0.7 × 10−1, 0.9 × 10−2, 0.1 × 10−3, and 0.5 × 10−3, by averaging the deviations of the experimental values obtained using the CCD-HAUP from the experimental values obtained using the G-HAUP. The precision of Δ, E, H″21, and θ0 were also estimated to be 0.3 × 10−2, 0.3 × 10−2, 0.1 × 10−4, and 0.7 × 10−5, as the standard errors of three MgF2 experiments.

Figure 3 shows the LB and LD spectra of the MgF2 crystal on the (100) face. The LB values, obtained using the CCD-HAUP, were consistent with those obtained using the G-HAUP, as well as with published values (1.18 × 10−2 at 600 nm).19,20 The measured LD values were zero, because MgF2 crystals are transparent in the visible region. The measurement time of the CCD-HAUP was 1.5 h, while that of the G-HAUP was 24 h. The results show that the LB spectrum of an anisotropic crystal can be measured rapidly using our CCD-HAUP system, with almost the same accuracy as the G-HAUP.

FIG. 3.

(a) LB and (b) LD spectra of an achiral MgF2 crystal on the (100) face measured using the G-HAUP (black open circle) and CCD-HAUP (black solid circle).

FIG. 3.

(a) LB and (b) LD spectra of an achiral MgF2 crystal on the (100) face measured using the G-HAUP (black open circle) and CCD-HAUP (black solid circle).

Close modal

CCD-HAUP measurements were conducted to obtain both the chiroptical and anisotropic properties of chiral crystals. To confirm that the CB can be measured accurately, α-quartz plate crystal (P3221) was measured initially. We purchased a (100) α-quartz plate crystal (P3221) from Crystal Base. The surfaces of the (100) plate crystal were polished with diamond slurry (abrasive grain diameter: 3.0 μm) and colloidal silica (32.5 nm) using polishing machines (Precision Lapping Machine, Nano Factor Tokyo, Japan and Metaserv 2000, Illinois, USA, respectively) to decrease the thickness and obtain flat and smooth surfaces. After that, the specimen was mounted on a Cu plate with a pinhole approximately 0.7 mm in diameter; the thickness was 35.7 μm.

See supplementary material of Figure S1 shows the raw experimental results (Δ, E, H″21, θ0) from the analysis of the α-quartz plate (P3221) measured using the CCD-HAUP and G-HAUP systems. We made successive approximations for determining the value of q and δY, by assuming that k is dependent on λ as follows:

k=sλ+t,
(14)

where s and t are constants, and k′ = 0 in the measured wavelength region. The optimum values of q, δY and N, and the constants s and t were obtained by least-squares fittings using Eqs. (6) and (9) and the p value. For this study, the p value was determined to be 1.0 × 10−4 by the wavelength dependence G-HAUP measurements at 20 °C of an achiral MgF2 crystal that was cut in the plane perpendicular to the a axis.

The systematic error parameters q and δY of the CCD-HAUP measurement determined by least-squares fittings are given below:

q=3.51×104
δΥ=2.92×103

The values of q and δY obtained by least-squares fittings for any other experimental data in this study varied between 10−4–10−3 and 10–5–10–3, respectively. In comparison with previous studies,5,12,13,21 these values are acceptable.

The values of k and k′ were calculated by introducing the obtained systematic errors, s, t, Δ, and E to Eqs. (6) and (9), as shown in See supplementary material of Figure S2. Finally CB and CD were determined from Eqs. (12) and (13).

Figure 4 shows the LB, LD, CB and CD spectra of α-quartz plate (P3221) on the (100) face at 20°C measured using the G-HAUP and CCD-HAUP systems. The LB value (0.009 at 570 nm) obtained using the CCD-HAUP is also consistent with a previous study (0.009 at 633 nm).15 The LB and CB values obtained using the CCD-HAUP are consistent with those obtained using the G-HAUP. The measurement time of the CCD-HAUP was 1.5 h, while that of the G-HAUP was 24 h. The results show that we can rapidly and simultaneously measure the LB and CB spectra of α-quartz plate (P3221) on the (001) face using our CCD-HAUP system. The measured LD and CD values (Figure 4b and d) were zero, because α-quartz is transparent in the visible region.

FIG. 4.

Optical anisotropic and chiroptical spectra of α-quartz plate (P3221) on the (100) face: (a) LB, (b) LD, (c) CB, and (d) CD. These properties were measured using the G-HAUP (black open circle) and CCD-HAUP (black solid circle).

FIG. 4.

Optical anisotropic and chiroptical spectra of α-quartz plate (P3221) on the (100) face: (a) LB, (b) LD, (c) CB, and (d) CD. These properties were measured using the G-HAUP (black open circle) and CCD-HAUP (black solid circle).

Close modal

To confirm that the LB, LD, CB and CD can be measured simultaneously using the CCD-HAUP, chiral crystals of both enantiomeric (S)- and (R)-salicylidenephenylethylamines in enol form [enol-(S)-1 and enol-(R)-1] were measured. Enol-(S)-1 and enol-(R)-1 have a photochromic nature in the crystalline state caused by photoinduced proton transfer (Figure 5).22 The plate-like enol-(S)-1 and enol-(R)-1 crystals bend reversibly upon UV irradiation.23 The compounds enol-(S)-1 and enol-(R)-1 were synthesized according to the published protocol.24 Thin, plate-like single crystals of enol-(S)-1 and enol-(R)-1 were prepared by slow sublimation at 70–80 °C. The top surface was assigned to be the (001) face in the longitudinal direction along the a axis, based on comparison with the plate-like bulk single crystals obtained by recrystallization from solution. Using silicone grease, the samples were fixed on a thin Cu plate with a pinhole of diameter 0.5 mm.

FIG. 5.

Photoinduced hydrogen transfer reaction of enantiomeric salicylidenephenylethylamines enol-(S)-1 and enol-(R)-1.

FIG. 5.

Photoinduced hydrogen transfer reaction of enantiomeric salicylidenephenylethylamines enol-(S)-1 and enol-(R)-1.

Close modal

See Figure S3 of the supplementary material shows the raw experimental results (Δ, E, H″21, θ0) from the analysis of the enol-(R)-1 crystal (thickness, 2.3 μm) on the (001) face using the CCD-HAUP and G-HAUP systems. We made successive approximations for determining the values of q and δY. From Eq. (12), the optical rotatory power (ORP) is defined as follows:

ORP=πλCB=πλ2kLB=Δkd.
(15)

Additionally, the ORP dispersion (ORD) spectra is related to the absorption spectrum of the optical active sample by the Drude expression25 as follows:

ORP=jAjλ2λj2,
(16)

where λj is the wavelength of the transition from the ground state to the excited state j, and Aj is a constant depending on the absorption strength of this transition. Therefore, we assumed that k is dependent on λ from Eqs. (15) and (16) as follows:

k=dΔjAjλ2λj2.
(17)

By assuming that k is dependent on λ as follows:

k=A330Δ(λ23302),
(18)

where A330 = A330 × d, and k′ = 0 above 400 nm; the optimum values of q, δY, and N and the constant A330 are obtained by least-squares fittings using Eqs. (6) and (9) and the p value.

The systematic error parameters q and δY determined by least-squares fittings are given below:

q=1.04×104
δΥ=6.15×105.

The values of k and k′ were calculated by introducing the obtained systematic errors, A330, Δ, and E to Eqs. (6) and (9), as shown in Figure S4 of the supplementary material. Finally CB and CD are determined from Eqs. (12) and (13), as shown in Figure 6.

FIG. 6.

Optical anisotropic and chiroptical spectra of enol-(R)-1 crystal on the (001) face: (a) LB, (b) LD, (c) CB, and (d) CD. These properties were measured using the G-HAUP (black open circle) and CCD-HAUP (black solid circle).

FIG. 6.

Optical anisotropic and chiroptical spectra of enol-(R)-1 crystal on the (001) face: (a) LB, (b) LD, (c) CB, and (d) CD. These properties were measured using the G-HAUP (black open circle) and CCD-HAUP (black solid circle).

Close modal

Figure 6 shows the LB, LD, CB, and CD spectra of an enol-(R)-1 crystal on the (001) face at 20°C measured using the G-HAUP and CCD-HAUP systems. The LB and CB values obtained using the CCD-HAUP are consistent with those obtained using the G-HAUP above 400 nm. Oscillations were observed in the LB and LD, probably due to multiple reflections between the (001) and (00–1) parallel planes.26 The real LB and LD values should be nearly constant above 400 nm (0.02 and 0, respectively) by taking into account multiple reflections, which are in good agreement with the previous report of the G-HAUP.13 Above 400 nm, the sign of CB is positive and the value (0.09 × 10−4 at 600 nm) is comparable with the data measured using G-HAUP (0.06 × 10−4 at 600 nm). The measurement time of the CCD-HAUP was 1.5 h, while that of the G-HAUP was 24 h. The results show that we can rapidly and simultaneously measure the LB and CB spectra of an enol-(R)-1 crystal on the (001) face above 400 nm using our CCD-HAUP system.

In the UV regime, below 400 nm, the LB, LD, CB and CD values obtained using the present CCD-HAUP were different from those obtained using the G-HAUP (Figure 6). In comparison with the previous report,13 the LB, LD, CB and CD values obtained using the G-HAUP are correct in the UV region. Although LD and CD are observed below 400 nm using the CCD-HAUP, these values are not so accurate; the transmitted light intensity of the CCD-HAUP without a sample is very weak below 400 nm (Figure 7), less than 5% of the maximum incident light intensity at 580 nm. Hence, the LB, LD, CB and CD spectra below 400 nm could not be measured accurately using the CCD-HAUP system. On the other hand, in the UV range below 400 nm, the transmitted light intensity of the G-HAUP without a sample is more than 15% of the maximum incident light intensity at 470 nm. Hence, the LB, LD, CB, and CD spectra below 400 nm obtained by the G-HAUP are accurate even with large absorption at around 330 nm.13 

FIG. 7.

Light intensity spectra transmitted through P and A using with the G-HAUP (black line) and CCD-HAUP (red line).

FIG. 7.

Light intensity spectra transmitted through P and A using with the G-HAUP (black line) and CCD-HAUP (red line).

Close modal

Next, we attempted to measure the LB, LD, CB, and CD spectra of an enol-(S)-1 crystal on the (001) face at 20°C using the CCD-HAUP system before, during, and after UV irradiation. An UV-light emitting diode (LED) light (365 nm; UV-400, Keyence) provided continuous irradiation from a direction almost vertical to the HAUP light path at low power (5 mW cm−2) to minimize the incidence of UV light on the CCD spectrometer.13 Crystal bending by UV irradiation was prevented by fixing the crystals to the Cu plate with silicone grease.13 Upon UV irradiation of the enol-(S)-1 crystal, strain and stress are generated by changes in unit cell sizes due to photoisomerization to trans-keto-(S)-1 near the crystal surface. Despite the total prevention of bending motion by fixing the crystal to a thin Cu plate with silicon grease for the HAUP measurement, internal stress should arise even under weak UV irradiation (5 mW cm-2), and thereby the observed LB might include the birefringence caused by stress. However, according to our previous report, it is suggested that the stress birefringence along the c axis might be very small.13 In the CCD-HAUP measurement, we assumed that the thickness of the enol-(S)-1 crystal did not decrease upon UV irradiation within 1.5 h.

Figure S5 of the supplementary material shows the raw experimental results (Δ, E, H″21, θ0) from the analysis of the enol-(S)-1 crystal (thickness, 5.0 μm) on the (001) face using the CCD-HAUP and G-HAUP systems. Before and after UV irradiation, the data analysis of the enol-(S)-1 crystal was identical to that of the enol-(R)-1 crystal, as described above. The systematic error parameters q and δY determined by least-squares fittings before and after UV irradiation are as follows:

q=4.49×104
δΥ=9.41×104,

and

q=2.61×104
δΥ=3.83×104,

respectively.

Under continuous UV irradiation, assuming that k is dependent on λ:

k=A330Δ(λ23302)+A460Δ(λ24602),
(19)

where A460 = A460 × d, and k′ = 0 above 550 nm; the optimum values of q, δY, and N and the constants A330 and A460 were obtained by least-squares fittings using Eqs. (6) and (9) and the p value.

The systematic error parameters q and δY determined by least-squares fittings are as follows:

q=2.51×104
δΥ=6.90×104.

The values of k and k′ were calculated by introducing the obtained systematic errors, A330, A460, Δ, and E to Eqs. (6) and (9), as shown in Figure S6 of the supplementary material. Finally CB and CD were determined from Eqs. (12) and (13).

Figure 8 shows the LB, LD, CB, and CD spectra of both the enantiomeric enol-(S)-1 and enol-(R)-1 crystals on the (001) face before UV irradiation. Although small oscillationswere observed in LB and LD, probably due to multiple reflections between the (001) and (00–1) parallel planes, the real LB and LD values were nearly constant (0.02 and 0, respectively), consistent with previous reports.13 The LB and LD spectra between the S and R enantiomeric crystals were coincident (Figures 8a and b), because optical anisotropic properties are not related to the chirality of the crystals. The sign of the CB of enol-(S)-1 crystal was negative, which is in a mirror relationship with the enol-(R)-1 crystal (Figure 8c). The CB value of the enol-(S)-1 crystal was in good agreement with the previous data (–0.1 × 10−4 at 600 nm) measured using G-HAUP.13 The baseline was shifted slightly to the positive (approximately +0.5 × 10−5) in the CB spectra, which we consider to be caused by the inaccuracy of systematic error (q and δY) evaluations. The obtained LB, LD, CB, and CD values of enol-(S)-1 crystal before UV irradiation were comparable to those obtained using the G-HAUP above 400 nm (Figure S7 of the supplementary material).

FIG. 8.

Anisotropic optical and chiroptical spectra of enol-(S)-1 (black open circle) and enol-(R)-1 (black solid circle) crystals on the (001) face at 293 K: (a) LB, (b) LD, (c) CB, and (d) CD. These properties were measured using the CCD-HAUP.

FIG. 8.

Anisotropic optical and chiroptical spectra of enol-(S)-1 (black open circle) and enol-(R)-1 (black solid circle) crystals on the (001) face at 293 K: (a) LB, (b) LD, (c) CB, and (d) CD. These properties were measured using the CCD-HAUP.

Close modal

Figures 9a and b show the LB and LD spectra of an enol-(S)-1 crystal on the (001) face at 20°C measured using the CCD-HAUP before, during, and after UV irradiation. Under continuous UV irradiation at 365 nm, the negative δLD peak corresponding to the photostationary state of the product trans-keto-(S)-1 appeared at around 460 nm (Figure 9d). The δLD value (−0.18 × 10−2 ± 0.02 × 10−2 at 460 nm), measured using CCD-HAUP, was comparable to the LD value (−0.15 × 10−2 ± 0.02 × 10−2 at 460 nm) measured using G-HAUP.13 The δLB spectra exhibited a slight anomalous dispersion of negative and positive peaks at around 500 and 420 nm, with a change in sign at the weak δLD peak (Figure 9c). The LB and LD spectra of the enol-(S)-1 crystal under continuous UV irradiation were consistent with those obtained using G-HAUP above 400 nm (Figures S8a and b of the supplementary material). These results confirm that the LB and LD spectra satisfy the Kramers–Kronig relationship, showing that the LD and LB spectra under continuous weak UV light irradiation were successfully measured by the CCD-HAUP system. After the removal of UV irradiation, the LD peak and anomalous dispersion of LB disappeared, returning to the initial spectra before UV irradiation.

FIG. 9.

(a) LB and (b) LD spectra of enol-(S)-1 crystal on the (001) face measured with the CCD-HAUP before, during, and after UV light irradiation (black open circle, red solid circle, and light blue open triangle, respectively). (c) δLB and (d) δLD present the differences between before and during UV irradiation. The curve line of δLB is served as an eye guide, and that of δLD is fitted by Gaussian functions.

FIG. 9.

(a) LB and (b) LD spectra of enol-(S)-1 crystal on the (001) face measured with the CCD-HAUP before, during, and after UV light irradiation (black open circle, red solid circle, and light blue open triangle, respectively). (c) δLB and (d) δLD present the differences between before and during UV irradiation. The curve line of δLB is served as an eye guide, and that of δLD is fitted by Gaussian functions.

Close modal

Figure 10 shows the CB and CD spectra of the enol-(S)-1 crystal on the (001) face at 20°C measured using the CCD-HAUP before, during, and after UV irradiation. Under continuous UV irradiation, a new small negative CD peak appeared at 460 nm due to the formation of trans-keto-(S)-1 crystals (Figure 10b). The CB spectrum also exhibited slight anomalous dispersions at around the new CD peak (Figure 10a), as with the LB and LD spectra. The Kramers–Kronig relationship also roughly holds between the CB and CD spectra. The CD value at 675 nm is not identical before UV irradiation and after stopping the UV irradiation (Figure 10b). As described above, the obtained systematic errors before UV irradiation and after stopping UV irradiation are slightly different. This small difference might cause the CD value difference, especially at 675 nm.

FIG. 10.

CB (a) and CD (b) spectra of enol-(S)-1 crystal on the (001) face. These properties were measured with the CCD-HAUP before, during, and after continuous UV light irradiation at 365 nm (black open circle, red solid circle, and light blue open triangle, respectively). The red solid curve lines are CB and CD under continuous UV irradiation, which are served as an eye guide and fitted by Gaussian functions, respectively.

FIG. 10.

CB (a) and CD (b) spectra of enol-(S)-1 crystal on the (001) face. These properties were measured with the CCD-HAUP before, during, and after continuous UV light irradiation at 365 nm (black open circle, red solid circle, and light blue open triangle, respectively). The red solid curve lines are CB and CD under continuous UV irradiation, which are served as an eye guide and fitted by Gaussian functions, respectively.

Close modal

The CD value under continuous UV irradiation at 460 nm measured using the CCD-HAUP was −0.20 × 10−4 ± 0.02 × 10−4, which is roughly comparable to that measured using the G-HAUP (−0.44 × 10−4 ± 0.07 × 10−4).13 The deterioration of the surface condition of the sample occurred in the G-HAUP measurement, which may have affected the accuracy of the measurement, especially in CD and CB due to the very small order of 10–4.13 Therefore this deterioration in the G-HAUP measurement might cause the different CD value in comparison with that measured using the CCD-HAUP (Figure S8d of the supplementary material). The short measurement time of the CCD-HAUP might contribute to the small standard error of the CD value. In contrast, the deterioration of the surface smoothness of the crystal sample by prolonged UV irradiation for three days might affect the large standard error of the CD value obtained by the G-HAUP. The CB values of the enol-(S)-1 crystal under continuous UV irradiation were similar to those obtained using G-HAUP above 400 nm (Figure S8c of the supplementary material). After stopping the UV irradiation, the CD peak and the anomalous dispersion of CB disappeared, returning to the initial spectra before UV irradiation.

In the G-HAUP measurement, we had to take into account the decrease in thickness of the specimen by prolonged UV irradiation for three days to obtain accurate LB, LD, CB, and CD spectra.13 The specimen thickness at each wavelength was corrected by dividing the total change in thickness by the number of G-HAUP measurements. In this study, the LB, LD, CB and CD values were successfully measured using the CCD-HAUP under the assumption that the thickness of the enol-(S)-1 crystal did not decrease upon UV irradiation within 1.5 h. The LB, LD, CB and CD obtained by this assumption were almost coincident before and after UV irradiation (Figures 9 and 10). Thus, the problem of decreasing specimen thickness under continuous UV irradiation during HAUP measurements was overcome by the use of the CCD-HAUP.

We have developed a fast, high-accuracy universal polarimeter using a CCD spectrometer (CCD-HAUP), to conduct optical anisotropic (LB and LD), as well as chiroptical (CB and CD), measurements of single crystals without any pretreatment, in the visible region. This measurement system has the unique feature that white transmitted light across all wavelengths can be measured simultaneously using a CCD spectrometer instead of a monochromator and a photomultiplier. The CCD-HAUP drastically reduced the measurement time for a dataset to only 1.5 h, from the 24 h required for a G-HAUP system. The LB, LD, CB, and CD spectra of single crystals of α-quartz and enantiomeric salicylidenephenylethylamines before, during, and after UV irradiation were comparable to those obtained using G-HAUP. The developed system is very effective for samples susceptible to degradation induced by external stimuli, such as light and heat. The future work is improvement of the system to extend the measurable wavelength to the UV region.

See supplementary material for the definition of polarizer, specimen, and analyzer expressed in Jones matrix, and experimental results of HAUP analysis.

This work was supported by the JSPS Scientific Research in the Challenging Exploratory Research, and the grant-in-aid from Mitsubishi Materials Corporation. A. T. acknowledges the Leading Graduate Program in Science and Engineering, Waseda University from MEXT, Japan.

1.
T. M.
Lowry
,
Optical Rotatory Power
(
Dover Publications
,
New York
,
1964
).
2.
J. F.
Nye
,
Physical Properties of Crystals
(
Oxford University Express
,
1985
).
3.
J.
Kobayashi
and
Y.
Uesu
,
J. Appl. Cryst.
16
,
204
(
1983
).
4.
T.
Asahi
,
H.
Utsumi
,
Y.
Itagaki
,
I.
Kagomiya
, and
Z.
Kobayashi
,
Acta Crystallogr., Sect. A: Found. Crystallogr.
52
,
766
(
1996
).
5.
H.
Koshima
,
M.
Nagano
, and
T.
Asahi
,
J. Am. Chem. Soc.
127
,
2455
(
2005
).
6.
J.
Kobayashi
,
T.
Asahi
,
M.
Sakurai
,
M.
Takahashi
,
K.
Okubo
, and
Y.
Enomoto
,
Phys. Rev. B
53
,
11784
(
1996
).
7.
J. R. L.
Moxon
and
A. R.
Renshaw
,
J. Phys.: Condens. Matter
2
,
6807
(
1990
).
8.
J. R. L.
Moxon
,
A. R.
Renshaw
, and
I. J.
Tebbutt
,
J. Phys. D: Appl. Phys.
24
,
1187
(
1991
).
9.
E.
Dijkstra
,
H.
Meekes
, and
M.
Kremers
,
J. Phys. D: Appl. Phys.
24
,
1861
(
1991
).
10.
M.
Kremers
and
H.
Meekes
,
J. Phys. D: Appl. Phys.
28
,
1212
(
1995
).
11.
M.
Tanaka
,
N.
Nakamura
,
H.
Koshima
, and
T.
Asahi
,
J. Phys. D: Appl. Phys.
45
,
175303
(
2012
).
12.
K.
Nakagawa
,
H.
Harper-Lovelady
,
Y.
Tanaka
,
M.
Tanaka
,
M.
Yamato
, and
T.
Asahi
,
Chem. Commun.
50
,
15086
(
2014
).
13.
A.
Takanabe
,
M.
Tanaka
,
K.
Johmoto
,
H.
Uekusa
,
T.
Mori
,
H.
Koshima
, and
T.
Asahi
,
J. Am. Chem. Soc.
45
,
15066
(
2016
).
14.
J.
Kobayashi
,
H.
Kumomi
, and
K.
Saito
,
J. Appl. Crystallogr.
19
,
377
(
1986
).
15.
J.
Kobayashi
,
T.
Asahi
,
S.
Takahashi
, and
A. M.
Glazer
,
J. Appl. Crystallogr.
21
,
479
(
1988
).
16.
T.
Asahi
,
T.
Osaka
, and
J.
Kobayashi
,
Proc. SPIE
4467
,
20
(
2001
).
17.
T.
Asahi
,
T.
Osaka
,
S. C.
Abrahams
,
R.
Miyazaki
,
H.
Asai
,
S.
Nanamatsu
, and
J.
Kobayashi
,
Proc. SPIE
5218
,
223
(
2003
).
18.
R.
Matsuki
,
T.
Asahi
,
J.
Kobayashi
, and
H.
Asai
,
Chirality
16
,
286
(
2004
).
19.
20.
A.
Takanabe
,
M.
Tanaka
,
A.
Taniguchi
,
H.
Yamanaka
, and
T.
Asahi
,
J. Phys. D: Appl. Phys.
47
,
285402
(
2014
).
21.
T.
Asahi
,
M.
Nakamura
,
J.
Kobayashi
,
F.
Toda
, and
H.
Miyamoto
,
J. Am. Chem. Soc.
119
,
3665
(
1997
).
22.
T.
Kawato
,
H.
Koyama
,
H.
Kanatomi
, and
H.
Shigemizu
,
Chem. Lett.
26
,
401
(
1997
).
23.
H.
Koshima
,
R.
Matsuo
,
M.
Matsudomi
,
Y.
Uemura
, and
M.
Shiro
,
Cryst. Growth Des.
13
,
4330
(
2013
).
24.
H. E.
Smith
,
S. L.
Cook
, and
M. E.
Warren
, Jr.
,
J. Org. Chem.
29
,
2265
(
1964
).
25.
L. D.
Barron
,
Molecular Light Scattering and Optical Activity
(
Cambridge University Press
,
Cambridge
,
2009
).
26.
C.
Hernández-Rodríguez
and
P.
Gómez-Garrido
,
J. Phys. D: Appl. Phys.
33
,
2985
(
2000
).

Supplementary Material