How to use geom_tile to plot factorial variable (photoperiod) behind continuous numeric data (x = ti-CodePudding

I am working with time-series data, where at every timepoint, I have many measurements for each of up to 16 subjects, spanning multiple days. I currently have the data organized as tidily as I can - here is a snippet.

> df
# A tibble: 23,844 x 40
   DateTime             Time exp_day Photoperiod    ZT Sex   Group Treatment Cohort Animal
   <dttm>              <dbl>   <dbl>       <dbl> <dbl> <chr> <chr> <chr>      <dbl>  <dbl>
 1 2021-10-18 12:47:00  0          1           1     6 M     mHFD  chABC          1    264
 2 2021-10-18 12:50:00  0.05       1           1     6 M     mHFD  chABC          1    264
 3 2021-10-18 12:53:00  0.1        1           1     6 M     mHFD  chABC          1    264
 4 2021-10-18 12:56:00  0.15       1           1     6 M     mHFD  chABC          1    264
 5 2021-10-18 12:59:00  0.2        1           1     6 M     mHFD  chABC          1    264
 6 2021-10-18 13:02:00  0.25       1           1     7 M     mHFD  chABC          1    264
 7 2021-10-18 13:05:00  0.3        1           1     7 M     mHFD  chABC          1    264
 8 2021-10-18 13:08:00  0.35       1           1     7 M     mHFD  chABC          1    264
 9 2021-10-18 13:11:00  0.4        1           1     7 M     mHFD  chABC          1    264
10 2021-10-18 13:14:00  0.45       1           1     7 M     mHFD  chABC          1    264
... with 23,834 more rows, and 30 more variables

The DateTimes are duplicated for each animal - so while I have 23,844 total observations, there are 1987 for each subject:

> n_distinct(df$DateTime)
[1] 1987
> n_distinct(df$Animal)
[1] 12

What I would like to do is plot time on the x axis, and variables on the y axis, but have the light cycle (light vs. dark) plotted as shaded or light bars behind the data, as shown below.

Plot goals

I found