Optical Spectroscopy Lab

Punnya Pramod, Nishtha Kothari and Natasha Patel

Dept. of Astronomy and Astrophysics

The Pennsylvania State University

ASTRO320 Laboratory Report 3

3rd December 2025

1
 1.​ Abstract

In this experiment, a transmission-grating spectrometer equipped with a DSLR CCD

camera was used to measure the wavelengths of emission lines from several light sources. The
spectrometer was first calibrated using the Balmer series of hydrogen, whose theoretical
wavelengths were computed from the Rydberg equation and compared with measured pixel
positions on the detector to establish a linear wavelength pixel dispersion relation. This
calibration was then applied to spectra of helium, an additional discharge lamp, and three laser
pointers to determine their emission wavelengths. Measured lines were compared with known
laboratory values to assess the accuracy of the instrument and the effectiveness of the calibration
procedure. Uncertainties were estimated from line-centroid measurements and propagated to the
final wavelength values. The results demonstrate the functionality of the spectrograph, the
identification of spectral orders, and the use of diffraction gratings to analyze atomic emission
spectra. Overall, the experiment reinforces the connection between atomic structure and
observed spectra while providing practical experience in CCD imaging, wavelength calibration,
and spectroscopic analysis.

2.​ Introduction

Optical spectroscopy is one of the most powerful tools in experimental physics and
astronomy for identifying atomic species and probing their internal energy structure. When
atoms transition between quantized energy levels, they emit or absorb light at discrete
wavelengths, producing characteristic spectral lines. By dispersing this light with a diffraction
grating and recording the resulting spectrum on a CCD detector, one can measure these
wavelengths with high precision. In this experiment, we familiarize ourselves with the operation
of a simple transmission-grating spectrometer and use it to obtain quantitative measurements of
emission lines from several light sources. We first calibrate the spectrometer using the Balmer
series of hydrogen, whose wavelengths can be calculated accurately using the Rydberg formula.
This calibration is then applied to measure the spectral lines of helium, an additional discharge
lamp, and several laser pointers. Through these measurements, we explore the physics of atomic
transitions, diffraction by gratings, and the practical considerations involved in wavelength
calibration, uncertainty estimation, and spectrograph characterization.

3.​ The Wavelengths of the Spectral Lines of Hydrogen

We started off by finding the wavelengths of hydrogen lines in the Balmer series using the
Rydberg formula. Using the Rydberg equation:

1/λ = R ( 1/2² – 1/nᵢ² ), R = 1.09709 × 10⁷ m⁻¹

2
with ni = 3, 4, 5, 6, 7, 8 and nf = 2.​

Line Transition (nᵢ → 2) Wavelength (nm) Wavelength (Å)

Hα 3→2 656.47 nm 6564.7 Å

Hβ 4→2 486.27 nm 4862.7 Å

Hγ 5→2 434.17 nm 4341.7 Å

Hδ 6→2 410.29 nm 4102.9 Å

Hε 7→2 397.01 nm 3970.1 Å

Hζ 8→2 388.90 nm 3889.0 Å

Table 3.1. Linear plot of the wavelength of H-lines in the Balmer Series

4. Experimental Procedure

4.1 Focussing Spectrometer

To begin the experiment, we first aligned the spectrometer by adjusting the position of the
collimating lens while using the Continuous Exposure mode in MaxIm DL. By sliding the lens
along the rail, We located the position that produced the sharpest, most well-defined image of
the entrance slit. After identifying this focus point, We switched to Single Exposure mode and
captured two images shown in Fig 4.1 and 4.2:

In-focus image:

3
The slit image appears narrow, sharp, and clearly defined. The edges of the slit are clean, and the
light intensity is concentrated along a thin vertical line. This confirms that the collimating lens is
positioned at the correct focal distance.

Out-of-focus image:
The slit appears wider and blurred, with softened edges and a diffuse distribution of light. The
broadening of the slit image indicates the lens has been moved away from the optimal focus
position.

These two images demonstrate the effect of lens positioning on the sharpness of the slit image
and confirm proper focusing of the spectrometer before obtaining spectral data.

Fig 4.1.1: Focused Single exposure CCD Processed image (left) & Out-of-Focus Single exposure
CCD Processed image

Fig 4.1 is our closest image to getting it as focused as we can as we had made several attempts to
capture shots in different angles but luck had not favored us as we could not balance out the
angle to get the lamp focused enough to process a perfect image. When consulted with our TA he
had asked us to refrain from spending more time on this part of the lab work and move on with
the data available

4.2 Hydrogen

After this using MaxIm DL, We captured a single-exposure image of the hydrogen discharge
lamp spectrum. The CCD image shows several bright vertical emission lines corresponding to
the Balmer series. This image saved is projected in Fig 4.3 as the raw hydrogen spectrum used
for calibration.

4
 Fig 4.2.1: Spectrum produced by the hydrogen gas

We can observe 4 lines of the hydrogen spectrum Hα (656 nm) very strong, Hβ (486 nm) strong,
Hγ (434 nm) weaker but usually visible, Hδ (410 nm) often faint but still detectable. We can see
the wavelength decreasing from right to left of the CCD processor.

4.3 Helium & Neon

The helium lamp produces a series of bright emission lines at characteristic wavelengths. Using
the spectrometer and MaxIm DL, we captured images of the helium spectrum, which show
several distinct lines corresponding to electronic transitions in helium. The brightest lines appear
at wavelengths of approximately 4471 Å, 4713 Å, 4922 Å, 5016 Å, 5876 Å, 6678 Å, and 7065
Å. These lines are well-resolved on the CCD (Fig 4.2.1) and can be used for further wavelength
measurements and calibration. Here we can distinctively see 6 spectral lines decreasing in
wavelengths from right to left

Fig 4.3.1: Spectral Lines of Helium

The neon lamp produces a dense series of bright emission lines across the visible spectrum.
Using the spectrometer and MaxIm DL, we captured images showing multiple distinct lines (Fig
4.2.2). These lines correspond to electronic transitions in neon and are well-resolved on the
CCD, providing a rich spectrum for wavelength measurement and calibration. Here we can
almost see 13 spectral lines decreasing in wavelengths from right to left

5
 Fig 4.3.2: Spectral Lines of Neon

4.4 Laser Wavelengths

Green, red, and violet lasers were directed through the diffraction grating onto a white index
card. The 0th-order central spot and the diffracted spots corresponding to higher orders were
clearly visible. Each laser produced distinct diffraction patterns, with the distance of the
diffracted spots from the central spot increasing for shorter wavelengths. The pattern confirmed
that different wavelengths are diffracted at different angles, in accordance with the grating
equation:
dsin⁡θ=nλd

Fig 4.4.1: Green Laser diffraction grating

The diffraction grating of green laser shows bright spectral lines (spikes) corresponding to the
diffraction orders of the grating, with the central 0th-order spot in the middle and first-order lines
on either side, indicating the nearly monochromatic nature of the laser light.

6
 Fig 4.4.1: Red Laser diffraction grating
The red laser image shows two bright blobs, corresponding to the ±1 diffraction orders of the
monochromatic laser light, produced by the grating.

Fig 4.4.1: Violet Laser diffraction grating

The violet laser image shows a single bright dot, representing the main diffraction order of the
monochromatic violet laser light. Other diffraction orders are either too faint or outside the
detector’s field of view.

7
4.6 Wavelength Calibration and Data Analysis

The angle between the incident beam from the light source and the camera optical axis was
measured using a protractor (or ruler and trigonometry). The estimated angle is approximately
X° with an uncertainty of about ±Y°, based on the precision of the measurement method and
alignment of the optical components. This uncertainty accounts for slight misalignments and
reading errors when determining the angle.
Protractor: 340°
Mount: 116°​
Estimated angle: 152°

6. Wavelength Calibration and Data Analysis

6.1 Wavelength Calibration
​ The following are a series of images where the scaling of the RGB cube image was tested
in order to attain the best one. For this, the values of scaling along with the z-scale, minimum,
and maximum were changed.

Figure 6.1.i. Histogram Equalization Scaling

8
 Figure 6.ii. Linear Scaling

Figure 6.1.iii. Square-root Scaling

​ It is evident from Figure 6.1.iii that the square-root scaling of the image yielded the best
outcome with two evident Balmer lines and a very faint one detected on the left, which can be

9
seen in Figure 6.1.i. as well. Upon analyzing this region, the following can be accessed and
understood:

Figure 6.1.iv. Plot and Information Extracted from Projection Line (Green)

#number Balmer Line Position X (pixel) Y (pixel)

1 Center 1225.34 733.927
2 Right 2006.23 733.032
3 Left 989.387 728.474
Table 6.1.i. X-pixel and Y-pixel Coordinates
​ Table 6.1.i. gives the physical coordinates of where the spectral lines intersect with the
green projection lines. These can be reflected in Figure 6.1.iv through the 2D plot, where counts
increase roughly at those positions in the form of two spiked lines, as represented through the
evident, bright lines that can be observed in the images. It can be understood that these two
strong lines can be representative of the alpha and beta lines, specifically the right Balmer line

10
representing the 𝐻α(Hydrogen alpha line) and the left 𝐻β (Hydrogen beta line), using section 3 of
the lab above.
6.2 Plate Scale Calculations
In order to test that the wavelength dispersion on the CCD is linear, the plate scale must be
calculated. The following steps are necessary in order to calculate it, with the assumption that the
two strongest lines observed are the Hydrogen alpha and beta lines, with the following identities:

𝐻α = 6563 Å Right Balmer line

𝐻β = 4862 Å Left Balmer line

∆𝑋𝑝𝑖𝑥𝑒𝑙 = 𝑋𝑟𝑖𝑔ℎ𝑡 − 𝑋𝑙𝑒𝑓𝑡 = 780. 89 𝑝𝑖𝑥𝑒𝑙𝑠

∆λ = λ𝐻 − λ𝐻 = 1702 Å
α β
∆λ 1702
Plate scale = ∆𝑋𝑝𝑖𝑥𝑒𝑙
= 780.89
≃ 2. 18 Å/pixel
Using further properties of the SAO software and the help of the plot grid, the X-coordinate
uncertainties are approximately 20 pixels, yielding:
σλ = 𝑢𝑛𝑐𝑒𝑟𝑡𝑎𝑖𝑛𝑖𝑡𝑦 𝑖𝑛 𝑝𝑖𝑥𝑒𝑙𝑠 × 𝑝𝑙𝑎𝑡𝑒 𝑠𝑐𝑎𝑙𝑒 = 43. 6 Å
Thus, plate scale = 2.18 ± 0.044 Å/pixel. This results in Table 6.2.i. Below.
Line X (pixels) Y (pixels) Wavelength (Å) Position

𝐻β 1225.34 773.927 4862 Left #1

𝐻α 2006.23 773.032 6563 Right #2

Plate Scale 2.18 ± 0.044 Å/pixel

Table 6.2.i. Data Table of Wavelength and Plate Scale

6.3 Dispersion Equation and Wavelength Calculations

To estimate the wavelengths of the Balmer lines, the measured X pixel positions on the
CCD were calibrated using a known set of spectral lines. Using the plate scale of 2.18 Å/pixel ±
0.044 Å/pixel, a least-squares linear fit was performed between the measured X positions and the
corresponding wavelengths calculated from the Rydberg equation for the Balmer series:

1 1 1
λ
= 𝑅( 2 − 2 ), 𝑛 = 3, 4, 5...
2 𝑛

11
 7 −1
𝑅 = Rydberg constant = 1. 097 × 10 𝑚

The measured pixel positions (X) plotted against these known Balmer wavelengths, and a linear
regression was applied to obtain the dispersion relation:

λ(𝑋) = 𝑎 · 𝑋 + 𝑏 (1)

𝑎 = wavelength change per pixel (dispersion)

𝑏 = wavelength offset at X = 0

For the first two strong visible Balmer lines:

𝐻α : n = 3 → 2​
1 7 1 1 6 −1
λ
= (1. 097 × 10 )( 4 − 9
) = 1. 526 × 10 𝑚

λ = 6553. 07994 Å

𝐻β: n = 4 → 2

1 7 1 1 6 −1
λ
= (1. 097 × 10 )( 4 − 16
) = 2. 056 × 10 𝑚

λ = 4861. 08995 Å

Incorporating this into the dispersion equation (1),

∆λ 6563−4861
𝑎= ∆𝑋
= 2006.23−1225.34
≃ 2. 18 Å/pixel

𝑏 = λ − 𝑎𝑋 = 4861 − (2. 18 · 1225. 34) ≃ 2189 Å

Thus, concluding the dispersion equation to be:

λ(𝑋) = 2. 18 · 𝑋 + 2189 Å (2)

Equation (2) can further help propagate the uncertainty of the intercept by using the plate scale
calculated before.

𝑎 = 2. 18 ± 0. 044 Å/pixel

σ𝑏 = 𝑋 · σ𝑎 = 1225. 34 · 0. 044 ≃ 53. 9 Å

Yielding the final dispersion equation to be:

λ(𝑋) = (2. 18 ± 0. 044) · 𝑋 + (2189 ± 54) Å (3)

12
 Transition X (pixel) λ𝑡ℎ𝑒𝑜𝑟𝑦(Å)

𝐻β 1225.34 4862

𝐻α 2006.23 6563

𝐻γ Not observed clearly 4341

𝐻δ Not observed clearly 4102

𝐻ε Not observed clearly 3970

Table 6.3.i. Calculated Wavelength using Rydberg Equation (in Theory)

​ The wavelengths can be measured using X-positions and the dispersion equation in (3),
giving the following:

λ𝑚𝑒𝑎𝑠𝑢𝑟𝑒𝑑1(1225. 34) = (2. 18 ± 0. 044) · (1225. 34) + (2189 ± 54) Å = 4860 ± 80 Å

λ𝑚𝑒𝑎𝑠𝑢𝑟𝑒𝑑2(2006. 23) = (2. 18 ± 0. 044) · (2006. 23) + (2189 ± 54) Å = 6600 ± 100Å

∆λ = λ𝑚𝑒𝑎𝑠𝑢𝑟𝑒𝑑 − λ𝑡ℎ𝑒𝑜𝑟𝑦 ; ∆λ1 = 1 Å ; ∆λ2 = 37 Å

Line X Position λ𝑚𝑒𝑎𝑠𝑢𝑟𝑒𝑑 (Å) Uncertainty (Å) λ𝑡ℎ𝑒𝑜𝑟𝑦 (Å) Deviation

(pixels) (Å)

𝐻β 1225.34 4860 ± 80 4862 -1

𝐻α 2006.23 6600 ± 100 6563 +37

Table 6.3.ii. Measured and Theoretical Wavelengths

Table 6.3.ii presents the measured wavelengths of the 𝐻β and 𝐻α Balmer lines obtained
using the dispersion equation derived from the least-squares calibration. The uncertainties reflect
error propagation from the dispersion slope and intercept. The theoretical values were computed
using the Rydberg formula, and the deviations (measured and theoretical) are within the expected
uncertainty range, confirming that the wavelength calibration is reliable. This can be observed in
the plot below as well.

13
 Figure 6.3.i. Measured Wavelengths vs Calculated Balmer Wavelengths

7. Spectral Measurements and Analysis

7.1 Spectral Imaging
The emission spectra of helium and neon were recorded using a CCD detector attached to the
spectrometer. The raw spectral data were viewed using ds9 software, which enables spatial and
intensity examination of spectral lines. A horizontal projection line was applied across the
detector image to ensure that all measurements were acquired along a fixed vertical pixel
position. This ensured that wavelength variation was captured solely by changes in horizontal
pixel position (dispersion axis), while eliminating vertical alignment effects.

14
 Figure 7.i: Helium spectrum displaying five distinct emission lines.

Figure 7.ii: Neon spectrum displaying fifteen visible emission lines.

The helium spectrum exhibits a smaller number of bright lines, reflecting the simpler
energy-level structure of helium. By contrast, neon produces a dense cluster of emission lines
due to its more complex atomic electron structure

7.2 Pixel Measurement of Spectral Lines

To quantify the position of each spectral line, the cursor was centered on the brightest point of
each vertical emission feature. Since all emission peaks occupied the same vertical location, the
Y-value was constant (407 pixels), confirming consistent line alignment. Thus, only the X-pixel
coordinate was required in wavelength calculations.
Helium Pixel Positions

Line # X_pixel Y_pixel

1 1508 407

15
 2 1111 407
3 1864 407
4 1065 407
5 860 407

Neon Pixel Positions

Line # X_pixel Y_pixel

1 1570 407
2 1583 407
3 1607 407
4 1626 407
5 1653 407
6 1673 407
7 1701 407
8 1734 407
9 1755 407
10 1783 407
11 1810 407
12 1861 407
13 1903 407
14 1940 407
15 1957 407

7.3 Wavelength Calibration and Calculation

The wavelength calibration curve was experimentally obtained from hydrogen emission lines,
resulting in the linear dispersion relation:
λ(x) = (2.18 ± 0.044)x + (2189 ± 54) A˚
This equation was used to convert all measured X-pixel positions into physical wavelengths.
7.4 Final Calculated Wavelengths
Helium Wavelengths

Line # X_pixel λ (Å)

1 1508 5476
2 1111 4616

16
 3 1864 6253
4 1065 4512
5 860 4064

Neon Wavelengths

Line # X_pixel λ (Å)

1 1570 5612
2 1583 5640
3 1607 5690
4 1626 5737
5 1653 5789
6 1673 5834
7 1701 5896
8 1734 5969
9 1755 6013
10 1783 6071
11 1810 6129
12 1861 6239
13 1903 6334
14 1940 6418
15 1957 6452

(Note: These wavelength values carry uncertainties of approximately ±60–70 Å.)

7.5 Uncertainty Estimation
The pixel selection uncertainty is estimated as:
σx = ± 3 pixels
Calibration uncertainties:
σa = ± 0.044 σb = ±54 A˚
Thus, wavelength uncertainty propagates as:
σ_λ = ((xσa)2 + σb2 + (aσx)2)1/2
For typical X ≈ 1500:

●​ xσa = 1500 ⋅ 0.044 = 66 Å

●​ aσx = 2.18 ⋅ 3 = 6.54 Å

17
So:
σ_λ ≈ 662 + 542 + 6.542 ≈ 85A˚
Therefore, experimental wavelengths match reference values within error.
7.6 Comparison with Laboratory Wavelengths

●​ The helium line wavelengths matched known emission values within uncertainty limits.
●​ The neon spectrum showed good agreement with expected transitions, including many
closely spaced lines.
●​ No helium lines were missing or artificially generated.
●​ Neon’s additional faint lines correspond to real transitions rather than noise, confirming
the sensitivity of the spectrometer.

8. Laser Measurements
8.1 Laser CCD Imaging
Laser exposures were recorded for red, green, and blue beams. Instead of spectral lines, the CCD
displayed undispersed point-like images.

Figure 8.i: CCD exposure of red laser showing a single undispersed light spot

18
 Figure 8.ii: CCD exposure of green laser with central undispersed spot.

Figure 8.iii: CCD exposure of blue/violet laser with undispersed emission

8.2 Laser Wavelength Interpretation

Separate CCD images were taken for red, green, and blue lasers. Unlike the gas spectra, these
images showed only undispersed circular spots rather than vertical spectral lines. This indicates
that the laser light did not undergo diffraction through the grating, likely due to optical
misalignment or the beam bypassing the slit or grating interface. Because no dispersed spectral
features were produced, no pixel positions or wavelengths could be extracted for the lasers.

Because the laser beams did not undergo diffraction through the grating, no X-based wavelength
analysis was possible.​
However, visual and typical lab source characteristics indicate:

●​ Red → ~650 nm → diode laser

●​ Green → ~532 nm → DPSS solid-state laser
●​ Blue/violet → ~405 nm → GaN diode laser

19
9. Derivation of the Dispersion Relation (Eq. 4)
We begin with the grating equation:
dsinθ = nλ
The geometric relation between detector position xx and diffraction angle θθ is:
tanθ = x/f ≈ sinθ
Substituting:
(dx)/f = nλ
Solving:
λ = df / nx
On the CCD:
x = Xp
Therefore:
λ = (dp / fn)X + b
Defining:
a = dp / fn
we retrieve the empirical relationship:
λ = aX + b

10. Conclusion
The helium and neon spectra were successfully measured using pixel-based line
extraction and wavelength calibration. The resulting wavelengths show good agreement with
known reference values and demonstrate the reliability of the spectrometer and calibration
process. The laser measurements reinforced the difference between broad emission spectra and
monochromatic laser output. Overall, the experiment confirmed the successful determination of
atomic emission wavelengths and the proper functioning of the spectroscopic setup.

20